1

Dissertation report on

Vidyasagar University

Report submitted by
ANKITA MONDAL
ROLL NO:
Registration No.
BENGAL INSTITUTE OF BUSINESS STUDIES

THIS REPORT IS SUBMITTED FOR THE PARTIAL FULFILMENT OF

MASTERS OF BUSINESS ADMINISTRATION FROM VIDYASAGAR
UNIVERSITY
2

Dissertation report on

Report submitted by

NAME –

Roll No –

Registration number –
Institute – Bengal Institute of Business Studies
3

Declaration
I hereby declare that the study entitled is being submitted by
me as my final year dissertation project in the partial fulfilment of the requirement for the award of
masters of Business Administration. The study is based on secondary source of data as well as on
primary data/information source.

The material borrowed from similar titles other sources and incorporated in the dissertation has been
duly acknowledged.

The matter embodied in this project report has not been submitted to any other university or
institution for the award of degree. This project is my original work and it has not been presented
earlier in this manner. This information is purely of academic interest.

Signature Date:

(Bengal Institute of Business Studies)

4

Certificate
To whomever it may concern

This is to certify that, Reg. No. of the year: 2022-2023 of; Roll No: ,a
Bonafede student of this college has completed his dissertation as per guidelines. He has done this
completely on his own and it is an original work and not copied from any resource.

Thanking you.

Yours truly,

Assistant Professor

Bengal Institute of Business Studies.

5

Acknowledgement
I would like to express my thanks of gratitude to my guide as well as for his valuable mentoring and
inputs who gave me the golden opportunity to do this wonderful project on the topic entitled
which also helped me in doing a lot of Research and I came to know about so many new things. I am
really thankful to them.

Secondly, I would like to thank my parents and friends who helped me a lot in finishing this project
within the limited time.

I am making this project not only for marks but also to increase my knowledge on consumer
behaviour towards Branding in current scenario.

Thanks again to all who helped me.

Signature of Student
6

Abstract
This paper titled ‘A details analysis of dairy market in India’ reviews the evolution of the dairy
sector in India and the stellar role played by dairy cooperatives since the launch of operation flood
which has made India the largest producer of milk in the world. The paper also provides the role
private sector is likely to play in future. It also provides policy recommendations for ensuring
sustainable growth of the sector in the coming decade. The paper explains the existing challenges
faced by the dairy sector and provides recommendations to the Government for ensuring sustainable
growth of the sector, where the private sector is likely to play a critical role. The paper stresses the
need for data collection on the milk processing capacities created by the private sector. At present,
reliable data is available only about the processing capacity of cooperative sector. This data would be
critical for proper implementation of a dedicated fund recently announced by the government for
expanding role of private sector milk processing capacity. The paper also discusses the export
potential of Indian dairy products and suggests policy measures to increase the export footprint of
dairy products. Dairy sector plays a critical role in providing livelihood opportunities to millions of
people, largely women, in rural areas. It has an important role to play if the target of doubling
farmer’s income has to be achieved in near future.

Viewed through the eyes of Western dairy observers, dairying in India is not only unique, but also
difficult to
comprehend.
Dairying in India is an unusual combination of economic endeavors. Cows and buffalo are a source
of family
milk, employment, and income from milk and dung production, but not from meat production. Dung
is used exten-
sively as fuel, and what manure is not used for fuel is used for fertilizer. Because of their sacred
status in Hindu
theology, bull calves and non-productive dairy cows are not a food source and therefore are a
liability, not the asset
that they represent to dairy farmers in most other countries. Indian buffalo—more prevalent than
dairy cows—are
not sacred, but because of dietary preferences, buffalo meat is seldom eaten. Consequently, cull
buffalo are no
more valuable or less costly than cull cows.
Dairying in India is largely a by-product of crop production, resulting in low milk yields. Most of the
feed for
dairy animals comes from crop residues. An often-heard adage is, “Feed first the people and what
remains to the
cows and buffalo.” Little green forage is fed to most of the dairy animals in India compared to other
countries, and
there is very little starch in the dairy animal diet. Much of the nutrition research conducted in India is
focused on
7

increasing the nutritive value of plant materials that are often discarded in Western countries. These
dairy animal
ration restrictions limit potential gains in milk yield.
Dairying in India is characterized by an extensive informal distribution system. While reliable
published data
verifying milk distribution are not available, there is evidence that nearly one-third of all milk
produced is con-
sumed on-farm. An estimated 70 percent of the remaining milk is sold in the unorganized sector,
either as liquid
milk or as home-produced dairy products such as dahi, paneer, butter, ghee and Indian sweets.
About one-sixth
of the Indian milk supply—approximately 16 million tons in 2005—is used for processing in
commercial dairy
plants, either as packaged fluid milk or manufactured products. Dairy products are perceived by
Indian consum-
ers as superior goods and are easy to sell. There would appear to be significant potential for increased
demand as
population and—especially—income grows. The volume of milk directed into the formal sector
seems woefully
inadequate to supply what will be a rapidly-growing market in higher-income areas.
Dairying in India is viewed from a national policy perspective less as a contributor to Gross
national Product
(GNP) than as a rural development initiative to provide a minimal income to tens of millions of
rural residents.
Consequently, there is strong resistance among national dairy leaders to increase dairy herd size,
since this could
reduce rural employment. There is equally strong resistance to subsidized dairy imports, which
would compete for
the relatively small market for domestically-produced manufactured dairy products from the
commercial (formal)
sector. The Indian dairy research and outreach structure has focused on generating a little bit
more milk from a
whole lot of cows. This is consistent with the national political view of dairying as a critical source
of subsistence
income for rural residents and a source of nutrition for the rural poor. That view may be
appropriate, since rural
areas have seen few benefits from the recent large rates of growth in Indiaʼs GNP. At the same time,
it raises ques-
tions about the ability of India to generate enough marketable surplus to remain self-sufficient in
dairy products
when consumers demand more and better-quality products.
Since dairying in India contributes to both social and economic goals, it would be presumptuous to
draw conclu-
sions about the dairy sector based only on economic criteria. Nonetheless, we offer the following
observations:
National goals of food self-sufficiency dating to Indian independence in 1947 have been
overachieved for wheat
and rice. High price supports frequently generate surplus production. Arguably, too much scarce
land is devoted
8

to wheat and rice and at the same time, green fodder for dairy animals is in very short supply.
Significant gains in
Viewed through the eyes of Western dairy observers, dairying in India is not only unique, but also
difficult to comprehend. Dairying in India is an unusual combination of economic endeavours. Cows
and buffalo are a source of family milk, employment, and income from milk and dung production,
but not from meat production. Dung is used extensively as fuel, and what manure is not used for fuel
is used for fertilizer. Because of their sacred status in Hindu theology, bull calves and non-productive
dairy cows are not a food source and therefore are a liability, not the asset that they represent to dairy
farmers in most other countries. Indian buffalo—more prevalent than dairy cows—are not sacred,
but because of dietary preferences, buffalo meat is seldom eaten. Consequently, cull buffalo are no
more valuable or less costly than cull cows. Dairying in India is largely a by-product of crop
production, resulting in low milk yields. Most of the feed for dairy animals comes from crop
residues. An often-heard adage is, “Feed first the people and what remains to the cows and buffalo.”
Little green forage is fed to most of the dairy animals in India compared to other countries, and there
is very little starch in the dairy animal diet. Much of the nutrition research conducted in India is
focused on increasing the nutritive value of plant materials that are often discarded in Western
countries. These dairy animal ration restrictions limit potential gains in milk yield. Dairying in India
is characterized by an extensive informal distribution system. While reliable published data verifying
milk distribution are not available, there is evidence that nearly one-third of all milk produced is
consumed on farm. An estimated 70 percent of the remaining milk is sold in the unorganized sector,
either as liquid milk or as home-produced dairy products such as dahi, paneer, butter, ghee and
Indian sweets. About one-sixth of the Indian milk supply—approximately 16 million tons in 2005—
is used for processing in commercial dairy plants, either as packaged fluid milk or manufactured
products. Dairy products are perceived by Indian consumers as superior goods and are easy to sell.
There would appear to be significant potential for increased demand as population and—especially—
income grows. The volume of milk directed into the formal sector seems woefully inadequate to
supply what will be a rapidly-growing market in higher-income areas. Dairying in India is viewed
from a national policy perspective less as a contributor to Gross national Product (GNP) than as a
rural development initiative to provide a minimal income to tens of millions of rural residents.
Consequently, there is strong resistance among national dairy leaders to increase dairy herd size,
since this could reduce rural employment. There is equally strong resistance to subsidized dairy
imports, which would compete for the relatively small market for domestically-produced
manufactured dairy products from the commercial (formal) sector. The Indian dairy research and
outreach structure has focused on generating a little bit more milk from a whole lot of cows. This
9

is consistent with the national political view of dairying as a critical source of subsistence income for
rural residents and a source of nutrition for the rural poor. That view may be appropriate, since rural
areas have seen few benefits from the recent large rates of growth in India’s GNP. At the same time,
it raises questions about the ability of India to generate enough marketable surplus to remain self-
sufficient in dairy products when consumers demand more and better-quality products. Since
dairying in India contributes to both social and economic goals, it would be presumptuous to draw
conclusions about the dairy sector based only on economic criteria. National goals of food self-
sufficiency dating to Indian independence in 1947 have been overachieved for wheat and rice. High
price supports frequently generate surplus production. Arguably, too much scarce land is devoted to
wheat and rice and at the same time, green fodder for dairy animals is in very short supply.
Significant gains in milk yields could be made with expanded production of forage and feed grains.
Altering the incentive structure to decrease area planted to crops in surplus and increase area devoted
to crops to support dairy production would seem to be a prudent policy choice. Rural development
efforts have focused on marginal gains in income, relying on multiplier effects combined with a high
marginal propensity to consume. Small gains in farmer income from added milk production translate
into small gains for merchants, service providers, landlords and others; small gains that translate into
larger aggregate gains in income and employment at the village level. However, the added income
appears to be doing little to alleviate poverty in rural areas and the quality of jobs created is not high.
This raises the question of whether there is a better way to allow rural areas to gain a larger share of
India’s robust overall economic growth. In particular, we question the lack of government support
for commercial dairying, which has the potential to create well-paying jobs in rural areas, not
necessarily at the expense of small-holder dairy farmers. India has shown impressive growth in milk
production and related gains in per capita availability of milk despite large population increases, but
most of the increased milk production has been consumed on the farms where it is produced or
absorbed by the informal sector. As India’s economy grows, it would appear logical to direct efforts
not only to increasing milk production, but also to increasing the proportion of marketing within the
formal sector. India has stressed self-sufficiency in its basic food supply and uses world markets
primarily as a balancing wheel for food grains and as a source of edible oils. Government policies
have not considered the comparative advantage India appears to have in fluid milk. There may be
benefits to channelling domestically produced milk to fluid and importing some manufactured dairy
products to meet growing demand.

The Basic intent of the study is to understand the current scenario of dairy industry in India and
various issues of the stakeholders of the industry. This is a review of research articles related to dairy
10

industry. 24 research articles referring data for different states of India are randomly selected.14 out
of 24 papers were based on primary data collected from dairy farmers (members/ non-members of
cooperative societies). The major issues revealed were related to lack of fodder and concentrates,
scarcity of veterinary and diagnostic services, lack of information and technological awareness.
These were common issues faced by dairy farmers in most of the regions of India. Some of the issues
were area specific like in Assam milk and milk products were not the components of daily
consumption basket of people in the state which is not the scenario in other states of the country. The
reaming 10 studies were either related to cooperative and private dairy plants, cooperative
federations or general dairy scenario of the relevant state. Procurement cost was found to be the
major component of total costs followed by processing cost. These studies evaluated the economic
viability of concerned dairy units.

Viewed through the eyes of

Western dairy observers,
dairying in India is not only
unique, but also difficult to
comprehend.
Dairying in India is an
unusual combination of
economic endeavors. Cows
and buffalo are a source of
family
milk, employment, and
income from milk and dung
11

production, but not from

meat production. Dung is
used exten-
sively as fuel, and what
manure is not used for fuel is
used for fertilizer. Because of
their sacred status in Hindu
theology, bull calves and
non-productive dairy cows
are not a food source and
therefore are a liability, not
the asset
that they represent to dairy
farmers in most other
countries. Indian buffalo—
more prevalent than dairy
cows—are
12

not sacred, but because of

dietary preferences, buffalo
meat is seldom eaten.
Consequently, cull buffalo
are no
more valuable or less costly
than cull cows.
Dairying in India is largely a
by-product of crop
production, resulting in low
milk yields. Most of the feed
for
dairy animals comes from
crop residues. An often-
heard adage is, “Feed first
the people and what remains
to the
13

cows and buffalo.” Little

green forage is fed to most
of the dairy animals in India
compared to other countries,
and
there is very little starch in
the dairy animal diet. Much
of the nutrition research
conducted in India is focused
on
increasing the nutritive value
of plant materials that are
often discarded in Western
countries. These dairy
animal
ration restrictions limit
potential gains in milk yield.
14

Dairying in India is
characterized by an
extensive informal
distribution system. While
reliable published data
verifying milk distribution
are not available, there is
evidence that nearly one-
third of all milk produced is
con-
sumed on-farm. An
estimated 70 percent of the
remaining milk is sold in the
unorganized sector, either as
liquid
milk or as home-produced
dairy products such as dahi,
paneer, butter, ghee and
15

Indian sweets. About one-

sixth
of the Indian milk supply—
approximately 16 million
tons in 2005—is used for
processing in commercial
dairy
plants, either as packaged
fluid milk or manufactured
products. Dairy products are
perceived by Indian
consum-
ers as superior goods and
are easy to sell. There would
appear to be significant
potential for increased
demand as
16

population and—especially—
income grows. The volume
of milk directed into the
formal sector seems
woefully
inadequate to supply what
will be a rapidly-growing
market in higher-income
areas.
Dairying in India is viewed
from a national policy
perspective less as a
contributor to Gross national
Product
(GNP) than as a rural
development initiative to
provide a minimal income to
17

tens of millions of rural

residents.
Consequently, there is
strong resistance among
national dairy leaders to
increase dairy herd size,
since this could
reduce rural employment.
There is equally strong
resistance to subsidized
dairy imports, which would
compete for
the relatively small market
for domestically-produced
manufactured dairy products
from the commercial
(formal)
18

sector. The Indian dairy

research and outreach
structure has focused on
generating a little bit more
milk from a
whole lot of cows. This is
consistent with the national
political view of dairying as a
critical source of subsistence
income for rural residents
and a source of nutrition for
the rural poor. That view
may be appropriate, since
rural
areas have seen few benefits
from the recent large rates
of growth in Indiaʼs GNP. At
19

the same time, it raises

ques-
tions about the ability of
India to generate enough
marketable surplus to
remain self-sufficient in dairy
products
when consumers demand
more and better-quality
products.
Since dairying in India
contributes to both social
and economic goals, it would
be presumptuous to draw
conclu-
sions about the dairy sector
based only on economic
criteria. Nonetheless, we
20

offer the following

observations:
National goals of food self-
sufficiency dating to Indian
independence in 1947 have
been overachieved for wheat
and rice. High price
supports frequently generate
surplus production.
Arguably, too much scarce
land is devoted
to wheat and rice and at the
same time, green fodder for
dairy animals is in very short
supply. Significant gains in
Viewed through the eyes of
Western dairy observers,
21

dairying in India is not only

unique, but also difficult to
comprehend.
Dairying in India is an
unusual combination of
economic endeavors. Cows
and buffalo are a source of
family
milk, employment, and
income from milk and dung
production, but not from
meat production. Dung is
used exten-
sively as fuel, and what
manure is not used for fuel is
used for fertilizer. Because of
their sacred status in Hindu
22

theology, bull calves and

non-productive dairy cows
are not a food source and
therefore are a liability, not
the asset
that they represent to dairy
farmers in most other
countries. Indian buffalo—
more prevalent than dairy
cows—are
not sacred, but because of
dietary preferences, buffalo
meat is seldom eaten.
Consequently, cull buffalo
are no
more valuable or less costly
than cull cows.
23

Dairying in India is largely a

by-product of crop
production, resulting in low
milk yields. Most of the feed
for
dairy animals comes from
crop residues. An often-
heard adage is, “Feed first
the people and what remains
to the
cows and buffalo.” Little
green forage is fed to most
of the dairy animals in India
compared to other countries,
and
there is very little starch in
the dairy animal diet. Much
of the nutrition research
24

conducted in India is focused

on
increasing the nutritive value
of plant materials that are
often discarded in Western
countries. These dairy
animal
ration restrictions limit
potential gains in milk yield.
Dairying in India is
characterized by an
extensive informal
distribution system. While
reliable published data
verifying milk distribution
are not available, there is
evidence that nearly one-
25

third of all milk produced is

con-
sumed on-farm. An
estimated 70 percent of the
remaining milk is sold in the
unorganized sector, either as
liquid
milk or as home-produced
dairy products such as dahi,
paneer, butter, ghee and
Indian sweets. About one-
sixth
of the Indian milk supply—
approximately 16 million
tons in 2005—is used for
processing in commercial
dairy
26

plants, either as packaged

fluid milk or manufactured
products. Dairy products are
perceived by Indian
consum-
ers as superior goods and
are easy to sell. There would
appear to be significant
potential for increased
demand as
population and—especially—
income grows. The volume
of milk directed into the
formal sector seems
woefully
inadequate to supply what
will be a rapidly-growing
27

market in higher-income
areas.
Dairying in India is viewed
from a national policy
perspective less as a
contributor to Gross national
Product
(GNP) than as a rural
development initiative to
provide a minimal income to
tens of millions of rural
residents.
Consequently, there is
strong resistance among
national dairy leaders to
increase dairy herd size,
since this could
28

reduce rural employment.

There is equally strong
resistance to subsidized
dairy imports, which would
compete for
the relatively small market
for domestically-produced
manufactured dairy products
from the commercial
(formal)
sector. The Indian dairy
research and outreach
structure has focused on
generating a little bit more
milk from a
whole lot of cows. This is
consistent with the national
29

political view of dairying as a

critical source of subsistence
income for rural residents
and a source of nutrition for
the rural poor. That view
may be appropriate, since
rural
areas have seen few benefits
from the recent large rates
of growth in Indiaʼs GNP. At
the same time, it raises
ques-
tions about the ability of
India to generate enough
marketable surplus to
remain self-sufficient in dairy
products
30

when consumers demand

more and better-quality
products.
Since dairying in India
contributes to both social
and economic goals, it would
be presumptuous to draw
conclu-
sions about the dairy sector
based only on economic
criteria. Nonetheless, we
offer the following
observations:
National goals of food self-
sufficiency dating to Indian
independence in 1947 have
been overachieved for wheat
31

and rice. High price

supports frequently generate
surplus production.
Arguably, too much scarce
land is devoted
to wheat and rice and at the
same time, green fodder for
dairy animals is in very short
supply. Significant gai

OLIVE OIL
32

INTRODUCTION
Olive oil is a liquid fat obtained from olives ,a traditional tree crop of the Mediterranean Basin, produced by pressing
whole olives and extracting the oil. It is commonly used in cooking: for frying foods or as a salad dressing. It can be
found in some cosmetics, pharmaceuticals, soaps, and fuels for traditional oil lamps. It also has additional uses in some
religions. The olive is one of three core food plants in Mediterranean cuisine; the other two are wheat and grapes. Olive
trees have been grown around the Mediterranean since the 8th millennium BC.

Spain accounts for almost half of global olive oil production; other major producers
are Portugal, Italy, Tunisia, Greece, Morocco and Turkey.[4] Per capita consumption is highest in Greece, followed by
Italy and Spain.

VARIENT OF OLIVE OIL

Extra Virgin Olive Oil - seasoning or salad

Virgin Olive Oil – cooking, sauteing, baking , salad too

Refined Olive Oil - only for cooking

Pure Olive Oil – cooking, body massage and preparing herbal oils

Olive Pomace Oil – cleaning purposes, it is not good for eating sometimes it is used for high heat cooking.

Extra-Virgin Type of Olive Oil

33

The extra-virgin olive oil type is certainly the best type among all. It is extracted by cold-
pressing method to prevent the natural content from being altered by exposures to high
temperature. This oil even has lower acidity content compared to its standard virgin
counterpart. It is considered the best simply because it retains most of the natural flavor
and aroma. This is as well the most commonly used type of olive oil for health purposes.
You can even consume or drink extra virgin olive oil directly without any other ingredient
mixed with it because the health benefits of olive oil are extensive.

Virgin-Olive Oil
The virgin type of olive oil is an unrefined form. It is extracted by cold-pressing method, and
it brings more natural aroma and flavors compared to the aforementioned types above. It
contains slightly high acidity content between 1 to 4 percent. This specific type of olive oil
is temperature resistant as well. Compared to the extra virgin type, this one has milder
taste. This is suitable for low-heat cooking purposes.

Refined Olive Oil

This type of olive oil is extracted from black and ripe olive fruits. Among other types, it has
the least amount of virgin olive oil which is known for its great benefits in preventing colon
cancer. The lite type is sometimes marketed as better than the regular one, but it is still
considered low quality and therefore is only good for cooking purposes, especially one that
involves high temperature. It still contains the same fat and calories as other types.

Pure Olive Oil

Although the name clearly suggests that it is pure, this type of oil is not at all pure. In fact,
it is a blend extra virgin or virgin type olive oil with a refined form of olive oil. The blending
method is often used when the extraction quality is not as good as expected. In order to
improve the quality, the refined oil has to be mixed with a better quality one, either extra
virgin or virgin type of olive oil. The added oil will give higher amount of vitamin E and
better flavor.

Olive Pomace Oil

The pomace type is the lowest quality olive oil. It is extracted from the residue that remains
after the fruit is pressed. After the actual fruits are pressed, they leave residue containing
oil and water. Once the water and oil are removed, the residue still contains certain amount
of oil and can be extracted. Since it is extracted merely from residue, virgin oil is mixed to
help to improve the quality. Due to this problem, it is the least expensive in the market and
suitable only for high-heat cooking.

HISTORY

Olive oil has long been a common ingredient in Mediterranean cuisine, including ancient
Greek and Roman cuisine. Wild olives, which originated in Asia Minor, were collected
by Neolithic people as early as the 8th millennium BC. [5][citation not found][6][better source needed] Besides
food, olive oil has been used for religious rituals, medicines, as a fuel in oil lamps, soap-
34

making, and skincare application.[citation needed] The Spartans and other Greeks used oil to rub
themselves while exercising in the gymnasia. From its beginnings early in the 7th century
BC, the cosmetic use of olive oil quickly spread to all of the Hellenic city-states, together
with athletes training in the nude, and lasted close to a thousand years despite its great
expense.[7][8] Olive oil was also popular as a form of birth control; Aristotle in his History of
Animals recommends applying a mixture of olive oil combined with either oil of cedar,
ointment of lead, or ointment of frankincense to the cervix to prevent pregnancy.[9]

REVIEW LITERATURE
Olive oil: a review

Influence of olive processing on virgin olive oil quality

… olive oil quality depends on different factors such as olive cultivar, olive tree cultivation and
the operations of olive … technological operations of olive processing on oil yields and quality …

Active components and clinical applications of olive oil.

… olive oil consumption needed to achieve health benefits. Research indicates
annual olive oil … an olive oil-rich diet to an olive oil-poor diet, the level of olive oil consumed …

Olive oil and health: summary of the II international conference

on olive oil and health consensus report, Jaén and Córdoba
(Spain) 2008
… Olive oil (OO) is the most representative food of the traditional Mediterranean Diet (MedDiet). …
This is because of the customary high intake of olive oil (OO), the characteristic culinary fat …

Save Cite Cited by 652 Related articles All 13 versions

Related searches
 virgin olive oil
 olive oil fatty
 olive oil phenolic
 olive oil mill
 olive oil polyphenols
 olive oil cooking
 olive oil adulteration
 olive oil wastewater
[PDF] olive-ltd.com.au
A review of waste management options in olive oil production
In this article, treatment and disposal alternatives of olive oil mill wastes and technical …
This literature also cited that the anaerobic treatment of the olive mill residual solids alone …
35

International conference on the healthy effect of virgin olive oil

… the ingestion of fried fats may be successfully buffered when using virgin olive oil [15]. In
summary, virgin olive oil consumption leads to the preservation of mitochondrial function and its …

Chemistry and health of olive oil phenolics

… olive oil phenolics. Moreover, a number of factors have the ability to affect phenolic concentrations
in virgin olive oil, … essential health promoting benefits of olive oil phenolic compounds. …

Olive oil and human cancer: an assessment of the evidence

… the literature. All papers from 1966 through 1995 were identified using either the term ‘‘olive
oil’… , or Spain) where a substantial part of the monounsaturated fat is provided by olive oil. All …

Save Cite Cited by 216 Related articles All 9 versions

Olive oil as a functional food: epidemiology and nutritional

approaches
… review focuses on olive oil and presents the most recent evidence available concerning its
therapeutic properties. The European Union has embraced olive oil as the oil … literature review…

Major phenolic compounds in olive oil: metabolism and health

effects
… The total phenol content of olive oil has been reported numerous times in the literature,
however there are inconsistencies with the concentrations obtained.

Integrated approach for the eco design of a new

process through the life cycle analysis of olive
.
oil: Total use of olive by-products

The disposal of waste is one of the main problems faced by the majority of the olive growing agribusiness:
olive oil by-products require specific management. In this work, the result of life cycle analysis shows that
the major problem is related to the management of mill waste water. The formation of compost from the
depleted pomace is bound to the emission of greenhouse gases. Therefore, the traditional way of waste
recovery is no longer the best alternative. In this context, we propose a complete process for total use of
olive by-products during the process of olive oil extraction. The project presents a pulp olive oil rich in
natural antioxidants, stone olive oil for cosmetic application and an olive paste with a smoky taste for food
consumption. The proposed system contributes to lower the environmental impact, improve the quality of
the oil, and promote the development of new and useful by-products. Practical applications: A new facility
composed of a destoner and a two phase extraction system is proposed to improve the working capacity and
contribute to lower the environmental impact during all the process of olive oil production. The proposed
new system produces, on the one hand, an olive pulp that gives an extra virgin olive oil rich in natural
antioxidants and an olive paste with a smoky taste for food consumption, and on the other hand, stones that
gives an olive oil for cosmetic application and a raw material to produce energy
36

1 Introduction The demographic explosion and the strong economic growth that has occurred since the
industrial revolution have profoundly modified the environment. This trend marked by the exhaustion and
the exploitation of natural resources has given rise to the development of strategies to mitigate or avoid the
production of harmful emissions. The use of cleaner production methods can limit these effects. In this work,
emphasis is placed on the life cycle of the olive oil. Although it is considered as a key product in
Mediterranean food production it is also associated with several adverse effects on the environment. This
involves the consumption of large amounts of resources and the production of toxic gaseous emissions and
natural environmentally harmful byproducts [1, 2] thus full discussions of the energy supply and greenhouse
gas (GHG) emissions are needed. The tool selected for calculating the environmental load is life cycle analysis
(LCA), one of the most commonly used environmental tools for management and decision-making. Detailed
investigations [2, 3] have shown that particular plantation characteristics and farming practices have
potentially the greatest negative environmental impacts, particularly in the form of soil erosion, run-off to
water bodies, degradation of habitats and landscapes, and exploitation of scarce water resources.
Inappropriate weed-control and soilmanagement practices, combined with the inherently high
Correspondence: Dr. Cinzia Benincasa, Council for Agricultural Research and Economics Olive Growing and
Olive Oil Industry Research Centre, via Li Rocchi, 87036 Rende (CS), Italy E-mail: [email protected]
Abbreviations: GHG, greenhouse gas; LCA, life cycle analysis Eur. J. Lipid Sci. Technol. 2017, 119, 1700009
1700009 (1 of 9) 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.ejlst.com risk of erosion in
many olive farming areas, is leading to desertification on a wide scale in some of the main producing regions,
as well as considerable run-off of soils and agrochemicals into water bodies, not beneficial for organic
farming. Indicatively, some authors [4–7] compared conventional and organic extra virgin olive oil
production systems. The study has resulted in the knowledge that biological system has a better
environmental profile. In addition to the impact of olive cultivation and the use of mineral fertilizers, which
give rise to the most significant environmental costs [8], showed that solid waste and wastewater generated
during production make significant challenges for the environmental sustainability of the industry. In this
context, other authors [3, 9–14] have focused their attention on the environmental impacts of olive
production by using a LCA and proposed some solutions such as the use of by-products as a source of energy
(fuel) or fertilizer. Hanandeh [15] has analyzed five alternatives for the management of this waste stream.
The study showed that, although composting can achieve significant environmental benefits using the olive
waste to produce energy, the production of pellets for the use in hot water boilers is the most likely option
for environmental benefits. Although a number of papers have been already published, up to the author’s
knowledge, this is the first study dedicated on the analysis of life cycle of the olive oil in Tunisia. The
objectives of the work focus mainly on the industrial processing of the olives and the by-products
management. More particularly, methods and elements that could reduce the negative impacts of this
sector in the environment were considered. For this purpose, in this paper, the following points were
investigated: (1) data collection relative to the exploitation of natural resources, energy consumption,
emissions and the resulting products of the various components of the production systems; (2) evaluation of
the environmental impact associated with each extraction system; (3) proposition of solutions based on the
eco-indicators (lower energy consumption, reduced maintenance end of life optimized, better management
of by-products) and development of strategies to reduce and minimize the impacts. In this way, the current
situation can be assessed and best management strategies can be chosen to optimize the use of energy,
protect the environment, and ensure sustainable development. This paper is divided into three sections:
Section 1 describes the assessment of life cycle of olive oil, Section 2 proposes an integrated approach for
recovery of waste, and Section 3 characterizes the new products. The use of a destoner in the process of
olive oil extraction will be highlighted since it improves the working capacity, the quality of the oil produced,
and the management of by-products [16]. 2 Materials and methods 2.1 Plant material The experimental
study is based on the production of olive oils of Chemlali cultivar. This variety, cultivated in the Center and
South of Tunisia, occupies 60% of the olive cultivation area. The olive tree is vigorous and characterized by
37

an abundant pollen production. The average yield of oil is 20% and contributes for more than 30% to the
national production. 2.2 Framework of the study and collection of data In this work, energy and economic
flows of olive production and its GHG emissions were assessed using a methodological framework according
to the ISO standard [17]. The scope of the present study consisted of four steps which are mainly olive
production, industrial processing (extraction), management of by-products, and transport. According to
Pishgar Komleh and coworkers [18], the most time consuming step of LCA studies implementation is the
collection of data to construct life-cycle inventory. Data are collected from the questionnaires carried out on
site in various oil mills. Another questionnaire was conducted at an olive residues center located in Sfax.
Additional data and results were collected from laboratories, official reports, and bibliographical references.
The LCA software tool used in our study was Quantis SUITE 2.0 that generates multiple indicators, such as
carbon, quality of ecosystems, use of resources, health,water footprint,and costs. Scalable and based on
validated methodologies, this software includes internationally recognized databases (Eco invent, balance
carbon ADEME, etc.) and standards in force in the various countries. The impacts are studied according to
the method of impact 2002þ. Different effects are studied utilizing the life cycle impact assessment
methodology IMPACT 2002þ. Through scientific models, it allows to join and classify a large number of
inventoried data also for the most simple environmental impact analysis. 2.2.1 Functional unit Functional
unit is an essential notion of Eco design. This unit allows the reduction in environmental impacts at a “slice of
result” identical for all the studied products belonging to the same category. Once the functional unit is set,
the search for ecological improvements can begin. In our case the functional unit is 1 L of extra virgin olive
oil. 2.2.2 Borders and stages of the system Our LCA study focused primarily on three objectives: olive
cultivation, transport of olives to the oil mill by mean of 10 tons capacity trucks, and extra virgin olive oil
extraction.

capacity, cleaning, grinding, kneading, decanting, separation, and storage of the oil produced. Two and
three-phase systems were considered. Product management has been included in the study, as well as the
management of waste water coming mainly from washing the olives and the equipments used. Packaging,
retail, and transport abroad were not included in this context. 3 Results and discussion Results were
processed using the Quantis suite.2.0 software. For each figure, the degrees of impact of each data stream
are expressed relatively to the impact categories mentioned in the impact method 2002þ. Median axis of
each graph was considered as an analytical marker. Percentages located below this axis have negative
values, but knowledge about such positive effects is explained by the substitution of a natural resource by a
by-product which is recoverable. Percentages located above represent positive values, but lead to negative
impacts due to the use of natural resources and the pollution emitted into the air, water, and soil. 3.1
Assessment of environmental impact of olive oil 3.1.1 Overview: Analysis of life cycle of olive oil In order to
evaluate the olive oil production, different types of management were considered: extensive farming, three
phases olive oil extraction system, sale of dry pomace for oil extraction unit, and natural evaporation of
wastewater. The results were tested by Quantis Suite 2.0 and illustrated in Fig. 1. As it can be noticed
through the figure, olive oil extraction is a process that have negative effects on the environment, human
health, and climate change. Despite that, trees represent a vital source since they are true air filters and
olive planting is an excellent means of controlling soil erosion. However, olives transport, oil extraction, and
waste management present an insignificant part of the environmental impacts in respect to those associated
to the agricultural practices. Gaseous emissions, mainly related to the use of chemical mineral fertilizers and
pesticides, and consumption of a large part of natural resources during the irrigation phase are the main
negative impacts associated to agricultural practices, while the emissions of toxic substances, produced
during the burning of fossil fuels for electricity production, are associated to olives transport, oil extraction,
and waste management. In fact, the use of pomace as fuel causes the emission of particles detrimental for
human health and climate change. Moreover, the negative impacts associated to the recovery of olive
residue are mainly due to the chemicals, phosphoric acid, and hexane, employed for the extraction and
38

refining of pomace olive oil. These results were confirmed by the work of Avraamides and Fatta [8]. The
authors showed that 3.9 kg of fossil carbon dioxide were releases from the whole system to produce a liter
of olive oil of which 3.6 kg (93.8%) were coming from processes related to the agricultural phase.
Furthermore, the authors showed that the production of a liter of oil was accompanied by the production of
32.2 g of nitrogen oxides and 13.7 g of SO2. The agricultural phase contributed for 99.1 and 88.3%,
respectively. Moreover, 1.6 g of SO2 were Figure 1. Influence of lifecycle of olive oil on the environmental
impacts. Eur. J. Lipid Sci. Technol. 2017, 119, 1700009 LCA of olive oil: Total use of olive by-products 1700009
(3 of 9) 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.ejlst.com produced to generate
electricity required for the extraction process. These observations highlighted the negative impacts induced
by the agricultural practices, largely related to the use of chemical fertilizers, synthetic pesticides, and
irrigation. Many researchers have adopted strategies to reduce these impacts thus emphasizing the benefits
of (1) organic farming, which is based on the use of ecological processes, biodiversity and cycles adapted to
local conditions; (2) implementation of good irrigation practices aimed at ensuring good crop yields and
minimizes environmental risks through the use of treated wastewater as an alternative source; (3) using
semi extensive lines. 3.1.2 Comparison between olive oil extraction systems The olive oil sector generates
nuisances for the environment, which requires the restructuring and modernization of olive cultivation, and
the use of appropriate technologies for olive oil extraction. Such technologies help to produce quality oil at a
lower cost with less negative impacts on the environment. Figure 2 shows that three-phase system was the
most impacting process (63%) compared to two-phase system. LCA data showed that these impacts were
mainly due to the emissions of toxic substances during the use of pomace to heat the boiler. Moreover, the
lack of good hygiene practices when cleaning equipment could be a source of these impacts; in fact, hygiene
and maintenance of the mill are needed to preserve the quality of oil and thus protect human health. Not to
mention the effluents produced during the washing caused impacts since they represented a hazard to
aquatic life. The environmental impacts of the production of olive oil using LCA was, also, undertaken by
Souza and coworkers [19] who studied and determined the emission of GHG caused by different fossil fuels.
In this context Notarnicola et al. [4] have shown that during the pressing of olives, emissions of nitrogen and
phosphate can occur causing harmful effects to the environment. From all these results, it can be concluded
that two-phase system generates the least impact on the environment for the three categories of impacts:
human health, ecosystem, and climate change. This was, also, confirmed by the results of Gurbuz and
coworkers [20]. The authors demonstrated that by substituting the three-phase system with a two-phase
one the costs to produce an olive oil of better quality could be achieved. In fact, the resulting olive oil was
richer in polyphenols, a determining factor for the economy of olive oil sector [21]. Despite the fact that the
negative impacts are minor using a two-phase system, the problem of the disposal of olive oil mills wastes is
still to be solved. In Tunisia, the simplest solution is in the lagooning of the wastes and so alternative ways to
improve the current situation are needed. However, environmental analysis performed on the production n
of a liter of extra virgin olive oil in Tunisia has revealed that (1) two-phase system was the most ecological
system; (2) management of wet pomace post-separation pulp-stone decreased the negative impacts and
generated new positive impacts. 3.2 Description of the new process of valorization In this session, a new
process of olive oil extraction is presented with the aim of developing a cleaner industry. The facility is
composed of a destoner and a two phase extraction system producing a pulp olive oil rich in natural
antioxidants, a stone olive oil for cosmetic application, and a smoky taste olive paste for food consumption
(Fig. 3). Fresh olives (100 kg) were carefully placed in an oven for 20 min at 35°C. This procedure turned the
olives in a dark color. Preheating the olives by means of evaporation is a procedure which consumes energy
causing the emission of particles harmful to the environment. However, this step is necessary to facilitate
the separation of the two components of the olives (pulp-stone) and decrease the time for the lactic
fermentation of the pulp. This prototype is on laboratory scale and we are prefacing one where this step can
be eliminated. The paste formed after the destoning is then mixed at 30°C for 35 min and centrifuged to
obtain a pure olive pulp. At this stage no water has been added. The deoiled pulp produced has undergone a
39

natural debittering by lactic fermentation at room temperature with 10% of brine for 28 days. The
fermentation process has been detailed studied in Section 3.3.2. Drying procedures of oil by-products were
not reported so far. The resulting olive paste with negligible oil content Figure 2. Environmental impact
related to the extraction oil mode. 1700009 (4 of 9) Z. Guermazi et al. Eur. J. Lipid Sci. Technol. 2017, 119,
1700009 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.ejlst.com and a smoky taste was
appreciated by consumers. The remaining pulp waste is a good food adjuvant to use in the preparation of
products such as olive taste pasta/bread. In order to evaluate the environmental impacts resulting from the
projected system, Quantis software was utilized and the data are summarized in Table 1. The results showed
that the new system used has a significant positive impact (P < 0.001) on the quality of the ecosystem,
consumption of water and water stress with respect to the three- and two-phase system (scenarios 1 and 2,
respectively). These results may encourage the producers to use this proposed process. 3.3 Characterization
of new products 3.3.1 Stone olive oil The knowledge of olive oil stone characteristics is quite limited. This
lack of data seems to be a handicap for the valorization of these oleaginous by-products that might pose a
potential for many applications. Few papers have dealt with stone olive oil that contains a low percentage of
saturated fats and a good ratio of unsaturated and saturated fatty acids. However, the content of
tocopherols resulted to be quite low [22, 23]. A sign that the topic is undoubtedly of great interest is that,
nowadays, several scientific studies are focused on this subject since many companies are moving in this
direction. The results produced in this study may contribute to the characterization of the good nutritional
values of this product as well as improve the Tunisian economy by producing this oil to be used for a
different purpose. The oils were extracted with n-hexane in a Soxhlet apparatus for 8 h. Physico-chemical
characteristics were determined according to conventional standard methods and results reported in Table
2. Acid and peroxide values (4 mg KOH/g and 22.87 meqO2/kg, respectively) exceeded the Figure 3. An
overview of the projected situation with different recovery pathways. Table 1. Influence of olive oil
extraction system and management of waste in on the different environmental impacts using Quantis 2.0
Software Olive oil extraction Management of mill wastewater Management of pomace Code Scenario 1 3-
Phase system Natural evaporation Extraction of pomace olive oil S1 Scenario 2 2-Phase system Natural
evaporation Oil of second extraction S2 Scenario 3 Proposed system Negligible amount Development of new
products: pulp olive oil, stone olive oil, and olive paste S3 – Climate change (kg CO2-eq) Human health
(DALY) Ecosystem quality (PDF.m2.y) Consumed water output (m3) Water stress indicator (m3-eq) Olive oil
extraction system S1 0.0893 2.602E-08 0.0045 0.00018 0.00016 S2 0.1135 2.567E-08 0.0062 0.00026
0.00024 S3 0.0981 2.448E-08 0.0094 0.00009 0.00011 n.s. Management of waste S1 1.8306 11.164E-07
0.4535 0.00479 0.0037 S2 1.7385 12.982E-07 0.5673 0.00844 0.0059 S3 0.4275 10.842E-07 0.3481 0.00024
0.0018 n.s. Significance at P < 0.05; significance at P < 0.01; n.s. not significant. Eur. J. Lipid Sci. Technol.
2017, 119, 1700009 LCA of olive oil: Total use of olive by-products 1700009 (5 of 9) 2017 WILEY-VCH Verlag
GmbH & Co. KGaA, Weinheim www.ejlst.com maximum limit established by COI for the extra virgin olive oil
category (0.8 mg KOH/g and 20 meqO2/kg); tocopherol and chlorophyll concentration were 0.234 mg/g and
3.07 mg/ kg, respectively. The iodine value was equal to 119.7, reflecting its unsaturated character. The fatty
acid profile indicated a significant proportion in mono and polyunsaturated fatty acids with a value of oleic
and linoleic acid of 54.12 and 21.24%, respectively. Palmitic and stearic acids were 14.78 and 3.12%,
respectively. All these results were in agreement with those found by Moghaddam and coworkers [23]. 3.3.2
Olive paste In the traditional process, for the preparation of dough spreads, olives were initially debittered
before being destoned and grinded. The idea was to enhance directly the pulp after the oil extraction. A pulp
exhausted of stone gives a very interesting value of natural antioxidants, pigments, vitamins and a significant
percentage of oil (1–2%): this could constitute a very interesting product similar to “tapenade.” This
procedure makes possible the obtainment of a new product useful for human consumption and eliminates
the transport and recovery of waste that consumes a lot of energy and releases a large amount of toxic
emissions into the environment. Several researchers have focused their attention in finding the best
conditions to debitter whole olives, but, to our knowledge, data on thedirect debittering of olive pasteafter
40

oil extractionare not available. The pulp obtained had a very bitter taste [24], therefore, to make it more
acceptable to the consumers, a process of debittering was necessary. The physico-chemical characteristics of
the paste that had undergone a natural fermentation process are summarized in Table 3 where paste 0
refers to the paste before fermentation; paste 1 refers to the paste fermented for 14 days with 8% of brine;
paste 2 refers to the paste fermented for 28 days and 28 days with 10% of brine; paste 3 refers to the paste
fermented for 42 days with 12% of brine. The results showed that there were significant differences
between pastes obtained at different fermentation conditions. The majority of consumers enjoyed paste 2
which presented a medium bitterness with fruity taste. Using this new facility we demonstrated that the oil
obtained was better than the traditional one: the fruity aspect was increased by 60% and the spicy taste by
54% [16, 25, 26]. Working with a destoned product increases the oil extraction capacity and the oil produced
is of a better quality. Furthermore, this systemmakes it possible the disposal of a quarter of solid waste
before the extraction process [27–30]. 3.3.3 Pulp olive oil Producers believe that oils obtained from
destoned olives are of better quality than oils extracted from the whole fruits, but more information in this
regard are needed [31–36]. It is established that the two type of oils show some differences. Concerning the
sensory profiles, oils obtained from whole fruit are more aromatic than oils obtained from destoned ones.
However, the stone contains a series of enzymes that trigger a number of different metabolic processes
when they are in contact with the oil during the crushing causing a reduction of the oil shelf life. All these
considerations should be taken into account when an olive oil is chosen. Moreover, the technology of the
majority of modern oil mills has not been designed for working a raw material free from stones. In fact, the
use of pulp core separators occur in the last stage of the oil production in order to recover the stones from
the pomace. A prior separation of the stone from the pulp before the olive oil extraction process makes it
possible to extract the oil with the production of a pulp rich in phenolic compounds [16] for human
consumption. Mansour and coworkers [25] showed that fruit and pulp oils are similar in fatty acid
compositions. Our previous experimental study indicates that stone removal before crushing affects to a
greater extent the quality of the oil (Tables 4 and 5). Indeed the use of stoned olives decreases the acidity by
20%. The specific results of extinctions K232 and K270 confirm that using stoned olives improves the stability
of oil since less oxidized compounds were detected (Table 4). This study showed that the obtained oil had a
better sensory note and a slightly higher antioxidant property (Table 5) than the conventional one and
consequently a higher commercial value in the market [16].

Analysis and trends for Life Cycle Assessment of

olive oil production

Analysis and trends for Life Cycle Assessment of olive oil production

Gabriela Espadas-Aldana a ([email protected]), Claire Vialle a ([email protected]),

Jean-Pierre Belaud b ([email protected]), Carlos VacaGarcia a ([email protected]),
Caroline Sablayrolles a ([email protected]) a Laboratoire de Chimie Agro-industrielle, LCA,
Université de Toulouse, INRA, Toulouse, France b Laboratoire de Génie Chimique, LGC, Université de
Toulouse, CNRS, Toulouse, France Postal address a Laboratoire de Chimie Agro-industrielle, ENSIACET, 4
allée Emile Monso, 31030 Toulouse Cedex 4 France b Laboratoire de Génie Chimique, ENSIACET, 4 allée
Emile Monso, 31030 Toulouse Cedex 4 France Corresponding author: Gabriela Espadas-Aldana, Laboratoire
de Chimie Agro-industrielle, LCA, Université de Toulouse, INRA, Toulouse, France - ENSIACET 4 allée Emile
Monso, 31030 Toulouse Cedex 4 France. E-mail: [email protected] © 2019 published by
Elsevier. This manuscript is made available under the Elsevier user license https://www.elsevier.com/open-
41

access/userlicense/1.0/ Version of Record:

https://www.sciencedirect.com/science/article/pii/S2352550919300533
Manuscript_e0a998f399d4ed3a8a3490309cf8ee7e 2 Abstract For over 10 years, the olive oil sector has been
largely studied through Life Cycle Assessment (LCA). Thus, a review of the numerous results and the different
evaluated scenarios is relevant for future LCA studies on this important agro-economic sector of the
Mediterranean region. The herein present work, based on scientific and technical literature aims to compare
system boundaries, functional units, life cycle inventories, allocation, impact assessments, and
interpretations from the currently existing LCAs of olive oil. Furthermore, an analysis of the assessment of
biogenic carbon has been carried out among the reviewed papers. As a result, 23 relevant LCA studies on
olive oil production or olives cultivation (for olive oil production) were identified. Analysis of the literature
converged on an unequivocal environmental hotspot, the agricultural phase; which represents the most
impactful phase of the olive oil life cycle, due in particular to fertilisation, pesticides treatment, and
irrigation. Waste management and distribution also appear to represent a crucial issue. The comparison
work made on climate change impact assessment is on the order of magnitude of 460 kg CO2-eq/ton of olive
and 1.6 kg CO2-eq/L of olive oil. This study highlights the complexity of carrying out an LCA on the olive oil
sector. At last, best practices and methodological recommendations were matured. HIGHLIGHTS: • Review
of life cycle assessment research on olive oil production • Critical analysis of 23 publications between 2008
and 2018 • Best practices and methodological recommendations • Climate change impact of 460 kg CO2-
eq/ton of olive and 1.6 kg CO2-eq/L of olive oil KEYWORDS: Agro-industry – olive oil – environmental
assessment – LCA – critical review 1 1. Introduction Given its organoleptic properties, olive oil consumption
trends have been increasing worldwide. Consumption went from 2.590x106 tons (t) in the 2000/2001 crop
year (a period from 1 October to 30 September) to 2.726x106 t in 2016/2017, an increase of 5.2% in the last
seventeen years. Over the same time span, worldwide production has constantly risen (Fig.1), despite
continuous fluctuations from year to year, linked to the uncertainty of climate and alternate-year bearing. It
is estimated for 2017/2018 a worldwide consumption of 3.008x106 t, a rise of 10.4% in only one year (IOC,
2018a). Performance figures for the 2017/2018 crop year point to a worldwide production growth by 29.4%
compared to the last crop year. IOC data for November 2018 estimate world production at 3.314x106 t (IOC,
2018b). Fig. 1 - Olive oil production vs consumption 1.0 1.5 2.0 2.5 3.0 3.5 x10 6 tons olive oil Crop year
WORLD WIDE CONSUMPTION WORLD WIDE PRODUCTION EU CONSUMPTION EU PRODUCTION 2 The
European Union (EU) account for more than 93% of all the olive oil production in the world. Three-quarters
of world production is concentrated in only three Mediterranean countries: Spain, Italy and Greece, each
accounting of 41%, 18% and 12% respectively (IOC, 2018a). The production of olive oil represents a strategic
sector for EU countries as it faces emerging competition with the arrival of new producing countries outside
the EU. The major competitors include Argentina, Australia, Mexico, Chile, New Zealand, South Africa and
the United States of America (Salomone and Ioppolo, 2012). These new producers use intensive and highly
mechanised methods that increase yields and reduce operational costs. Concerning the consumption in the
EU, this has been decreasing over the years (Fig. 1), no one at the IOC has been capable to explain these
trends. Despite the economic importance of this food product in the mentioned Mediterranean countries,
olive oil production is associated with several negative effects on the environment which causes depletion of
resources, land degradation, air emissions and waste generation (Salomone et al., 2015). Moreover, the olive
oil production sector faces crucial issues concerning waste management. Eighty percent of the olives mass is
composed of olive pulp and stones. Thus, the extraction process yields 4 times more waste than oil. The
composition of the waste products depends on extraction technologies; which include press, 2-phase or 3-
phase systems (Cinar and Alma, 2008). Additionally, in olive oil-producing countries (typically
Mediterranean), climate is sunny, warm and dry; the overexploitation of the relatively thin and fragile
organic soil layer might lead to soil erosion. This problem is exacerbated by irrigation, the use of pesticides
and mineral fertilisers. Also, it can be attributed to waste management, as a large proportion of organic
waste from olive mills are applied directly on the same soils. According to Roig et al. (2006) the negative
42

effects are associated to its high mineral salt content, on the other hand, the carbon content and soil fertility
might be improved with this organic matter application thanks to its high nutrient concentration (like
potassium). Given that, the Mediterranean region is subject to particular levels of dryness, water is scarce
and represents an issue of significant importance. In such areas, irrigation remains a sensitive topic that is
environmental and socially controversial. Environmental assessments carried out for water-consuming
systems in said arid regions need to take into account irrigation processes.

Detection of Extra Virgin Olive Oil Adulteration

with Lampante Olive Oil and Refined Olive Oil
Using Nuclear Magnetic Resonance Spectroscopy
and Multivariate Statistical Analysis:
High-field 31P NMR (202.2 MHz) spectroscopy was applied to the analysis of 59 samples from three grades
of olive oils, 34 extra virgin olive oils from various regions of Greece, and from different olive varieties,
namely, 13 samples of refined olive oils and 12 samples of lampante olive oils. Classification of the three
grades of olive oils was achieved by two multivariate statistical methods applied to five variables, the latter
being determined upon analysis of the respective 31P NMR spectra and selected on the basis of one-way
ANOVA. The hierarchical clustering statistical procedure was able to classify in a satisfactory manner the
three olive oil groups. Subsequent application of discriminant analysis to the five selected variables of oils
allowed the grouping of 59 samples according to their quality with no error. Different artificial mixtures of
extra virgin olive oil-refined olive oil and extra virgin olive oil-lampante olive oil were prepared and analyzed
by 31P NMR spectroscopy. Subsequent discriminant analysis of the data allowed detection of extra virgin
olive oil adulteration as low as 5% w/w for refined and lampante olive oils. Further application of the
classification/prediction model allowed the estimation of the percent concentration of refined olive oil in six
commercial blended olive oils composed of refined and virgin olive oils purchased from supermarkets.

INTRODUCTION Olive oil is a fine product with high nutritional value and significant health benefits (1-3). It is
known for its superior organoleptic characteristics (aroma and taste) and its remarkable antioxidant
properties. The cultivation of olive trees, the harvesting of the olive fruits, and the extraction of olive oil are
hard and time-consuming tasks, which add to its relatively high commercial price. Therefore, attempts to
adulterate this commodity with less expensive materials, such as seed oils and/or olive oils of lower quality
(refined olive oil), are by no means rare. Needless to say, this practice deteriorates its quality and nutritional
value and causes major economic losses. European Mediterranean countries, which are major suppliers of
olive oil in the world market, adopted common legislations to protect olive oil growers and consumers from
olive oil fraud. The current European Union legislation (4) classifies olive oil into various categories reflecting
its quality. These include various grades of virgin olive oils, in which extra virgin olive oil (EVOO) is considered
to be the oil of the highest quality. Its free acidity expressed as oleic acid must be 2.0%) and poor
organoleptic properties. Refined olive oil (ROO) is an oil of lower grade obtained usually from virgin olive oil
mechanically extracted from damaged olive fruits or from olives stored in unsuitable conditions and using
refining methods that do not lead to alterations of the initial glyceridic structure. Its free acidity does not
exceed 0.3%. Addition of ROO to EVOO is expected to deteriorate the antioxidant properties and
organoleptic characteristics of EVOO (1-3). The presence of foreign oils in EVOO is usually reflected on
measured physical and chemical parameters, such as K232-270, peroxide value, and ratio of trans-2-hexenal
43

to hexanal, as well as on organoleptic assessments (4). Adulteration of EVOO with LOO is not a common
practice and is detected quite easily by conventional methods. However, LOO distinction from EVOO is
important because LOO cannot be consumed without refining (6), and the present methodology is suitable
for this purpose. Authentication and detection of olive oil adulteration is based on chromatographic
techniques including high-resolution gas chromatography and high-performance liquid chromatography
(see, for instance, ref 5), which is the current basis of olive oil adulteration standards. Chromatography is the
most accurate analytical technique with low detection limits, although it is time-consuming because it
requires several steps to complete quantification. This methodology is useful for routine work for * Author
to whom correspondence should be addressed (fax 30 2 81 0 393 601; e-mail [email protected]). 2810
J. Agric. Food Chem. 2005, 53, 2810−2816 10.1021/jf040279t CCC: $30.25 © 2005 American Chemical
Society Published on Web 03/24/2005 an assessment of the olive oil quality according to EU regulations
(ECN 42, spectroscopic constants K232-270, triglycerides, peroxide value, etc.). In the past decade, there has
been an increasing interest in the classification of edible oils including olive oil as an alternative means to
examine authentication and to detect possible adulteration of extra virgin olive oils with seed oils and/or
olive oils of lower grade. Classification of various grades of olive oil has been carried out in several instances
by using a variety of analytical techniques and chemometric procedures. Among these are classical methods
based on gas chromatography and high-performance liquid chromatography (7, 8), mass spectrometry (9),
isotope mass ratio analysis (10), metal oxide sensors (6), vibrational spectroscopy (11-16), and nuclear
magnetic resonance (NMR) spectroscopy (17-20). In a recent publication (21) we proposed a methodology
based on NMR spectroscopy in combination with multivariate statistical analysis to classify 13 types of edible
oils and to detect virgin olive oil adulteration with corn, soybean, sunflower, and hazelnut oils. In particular,
31P NMR spectroscopy was used to determine in a single experiment 1,2-diacylglycerols (1,2-DGs), 1,3-
diacylglycerols (1,3-DGs), total diacylglycerols (TDGs), the ratio D of 1,2-DGs over TDGs, total free sterols,
and free acidity. This method (21) is based on the derivatization of the labile hydrogens of hydroxyl and
carboxyl groups with 2-chloro4,4,5,5-tetramethyl dioxaphospholane (1) and the use of the 31P chemical
shifts of the phosphitylated compounds to identify the labile centers. The phosphorus reagent reacts rapidly
and quantitatively under mild conditions with the hydroxyl and carboxyl groups (21). The same procedure
will be used in the present study to discriminate among three different grades of olive oil, namely, EVOO,
ROO, and LOO, and to detect EVOO adulteration with these oils. MATERIALS AND METHODS Samples and
Reagents. Thirty-four EVOO harvested between 2001and 2003 were provided by local cooperatives. Twenty-
one olive oils were sampled in Crete (Peza, Kolymbari, Heraklion, Sitia), 3 oils in Messinia, and 1 in Ilia. They
were extracted from the olive variety Koroneiki, whereas the oils from Lakonia (6 oils), Lesvos (1 oil), and
Pilion (2 oils) were extracted from the varieties Athinolia, Kolovi, and a local variety, respectively. The
samples were extracted by centrifugation within 48 h after harvesting and stored immediately in brown
screw-capped bottles at -20 °C prior to spectrum acquisition and were coded as extra virgin olive oils
according to the official limits (4). A total of 25 samples of olive oil of lower quality were kindly supplied by
the olive oil company Minerva (Athens, Greece): 12 samples of LOO and 13 samples of ROO. Refined olive
oils were deodorized in the final step of the refining process. Finally, six commercial samples of blended olive
oil composed of refined olive oil and virgin olive oil (BOO) were purchased from local supermarkets. The 31P
NMR method was applied to detect olive oil adulteration. Fresh EVOO samples were mixed with LOO and
ROO samples. Two set of mixtures of 1, 3, 5, 10, 15, 20, 35, 50, 65, and 80% (w/w) and 1, 3, 5, 10, 15, 20, 35,
and 50% for ROO adulterant in EVOO and one set of 1, 3, 5, 10, 15, and 20% (w/w) of LOO in EVOO were
prepared. EVOO samples of different geographical and botanical origins and different ROO samples were
used for the preparation of the various mixtures. Finally, three EVOO samples of different geographical and
botanical origins (1 from Pilion, 1 from Sitia, and 1 from Lesvos), not included in the 34 EVOO samples used
for the present analysis, were adulterated with different ROO samples at concentrations of 60, 65, 70, 75,
and 80% (w/w). These mixtures were analyzed by 31P NMR spectroscopy and used as blank tests to estimate
the composition of BOO samples. All olive oil samples and their mixtures were analyzed twice. Pinacol,
44

triethylamine, phosphorus trichloride, protonated solvents (reagent or analytical grade), and deuterated
solvents used in the present study were purchased from Sigma-Aldrich (Athens, Greece). The derivatizing
phosphorus reagent 2-chloro-4,4,5,5-tetramethyldioxaphospholane (1) was synthesized from pinacol and
phosphorus trichloride following the method described in the literature (23). However, to increase the yield

suggested in the original method. This modification resulted in ∼45% yield of the product against 19%
of the reaction, we utilized hexane solvent instead of benzene and pyridine instead of triethylamine as

obtained with the original method. Sample Preparation for 31P NMR Spectral Analysis. A stock solution was
prepared by dissolving 0.6 mg of chromium acetylacetonate, Cr(acac)3 (0.165 µM), and 13.5 mg of
cyclohexanol (13.47 mM) in 10 mL of a mixture of pyridine and CDCl3 solvents (1.6:1.0 volume ratio) and
protected from moisture with 5A molecular sieves. Cyclohexanol was used as an internal standard for
quantification purposes. Olive oil (100-150 mg) was placed in a 5 mm NMR tube. The required volumes of

∼15 min at room temperature. Upon completion of the reaction, the solution was used to obtain the 31P
the stock solution (0.4 mL) and the reagent 1 (15 µL) were added. The reaction mixture was left to react for

NMR spectra. 13P NMR Experiments. 13P NMR experiments were conducted on a Bruker AMX500
spectrometer operating at 202.2 MHz for the phosphorus-31 nucleus at 30 ( 1 °C. The spectra were recorded
by employing the inverse gated decoupling technique in order to suppress NOE. Typical spectral parameters
for quantitative studies were as follows: 90° pulse width ) 12.5 µs, sweep width ) 10 kHz, relaxation delay )
30 s, memory size ) 16K (zero-filled to 32K). Line broadening of 1 Hz was applied, and drift correction was
performed prior to Fourier transform. Polynomial fourth-order baseline correction was performed before
integration. For each spectrum 32 transients were accumulated. All 31P chemical shifts are relative to the
product of the reaction of 1 with water (moisture contained in all samples), which gives a sharp signal in
pyridine/CDCl3 at δ 132.20. It should be noted that the presence of the paramagnetic metal center of
Cr(acac)3 in the samples lowers the relaxation times of the phosphorus nuclei, shortening thus the duration
of the measurements significantly. The relaxation delay in the 90° pulse sequence is based on the 31P spin-
lattice relaxation time of 4.9 s measured for the internal standard. Lower values were measured for other
model compounds (3.5 s for stearic acid and 2.3 s for 1,2-diolein). Therefore, the relaxation delay (5 times
the longest relaxation time) is suitable for quantitative analysis. The applicability of this method to
quantitative analysis, as well as its reproducibility and repeatability, has been tested thoroughly in previous
studies (20, 21). Statistical Analysis. One-way ANOVA was used to find the variables with the highest
discriminatory power. The results of this analysis are reported as the Fisher F ratios and p level. The F ratio
with the degrees of freedom tests whether the between and within variances are significantly different. The
p level represents a decreasing index of reliability of a result and gives the probability of error involved in
accepting a result as valid. The results reported below show that the selected variables were significantly
different for the three grades of olive oil. The unsupervised statistical method of hierarchical cluster analysis
(HCA) was applied to explore similarity (or dissimilarity) among the various olive oil samples, whereas the
supervised method discriminant analysis (DA) classified olive oils according to different grades and detected
adulteration. Discriminant function analysis is used to determine which variables (e.g., compositional
parameters of olive oil samples) discriminate between two or more naturally occurring groups (e.g., olive oil
grades). The DA method applied to a matrix of 5 independent variables and 59 olive oil samples provided
good results (see below) while being very straightforward in calculation and interpretation. Visualization of
the oil classification is achieved by plotting the individual scores for two principal discriminant functions,
whereas for the detection of EVOO adulteration the discriminant functions obtained from the DA model and
the Mahalanobis distances were used. The statistical package Statistica for Windows 5.1B (StatSoft Inc.) was
used.
45

Chemometrics Methods for Specificity, Authenticity and

Traceability Analysis of Olive Oils: Principles, Classifications
and Applications
by

Habib Messai

1
,

Muhammad Farman

,
2

Abir Sarraj-Laabidi

3
,
46

Asma Hammami-Semmar
4
and

Nabil Semmar
3,5,*

Laboratory of Biomedical Genomics and Oncogenetics, Institut Pasteur de Tunis, University of Tunis El
Manar, 1002 Tunis, Tunisia
2

Department of Chemistry, Quaid-i-Azam University, 45320 Islamabad, Pakistan

3

Laboratory of Bioinformatics, Biomathematics and Biostatistics (BIMS), Institut Pasteur de Tunis, University of
Tunis El Manar, 1002 Tunis, Tunisia
4

National Institute of Applied Sciences and Technology (INSAT), University of Carthage, 1080 Tunis, Tunisia
5

Laboratoire de Biomathématiques, Faculté des Sciences de Saint-Jérôme, Aix-Marseille Université, 13397

Marseilles, France
*

Author to whom correspondence should be addressed.

Foods 2016, 5(4), 77; https://doi.org/10.3390/foods5040077

Received: 11 October 2016 / Revised: 7 November 2016 / Accepted: 10 November 2016 / Published: 17
November 2016

(This article belongs to the Special Issue Food Identity, Authenticity and Fraud: The Full Spectrum)

Download

Browse Figures
47

Versions Notes

Abstract
Background. Olive oils (OOs) show high chemical variability due to several factors of genetic, environmental
and anthropic types. Genetic and environmental factors are responsible for natural compositions and
polymorphic diversification resulting in different varietal patterns and phenotypes. Anthropic factors, however,
are at the origin of different blends’ preparation leading to normative, labelled or adulterated commercial
products. Control of complex OO samples requires their (i) characterization by specific markers; (ii)
authentication by fingerprint patterns; and (iii) monitoring by traceability analysis. Methods. These quality
control and management aims require the use of several multivariate statistical tools: specificity highlighting
requires ordination methods; authentication checking calls for classification and pattern recognition methods;
traceability analysis implies the use of network-based approaches able to separate or extract mixed
information and memorized signals from complex matrices. Results. This chapter presents a review of
different chemometrics methods applied for the control of OO variability from metabolic and physical-chemical
measured characteristics. The different chemometrics methods are illustrated by different study cases on
monovarietal and blended OO originated from different countries. Conclusion. Chemometrics tools offer
multiple ways for quantitative evaluations and qualitative control of complex chemical variability of OO in
relation to several intrinsic and extrinsic factors.

Keywords:

chemometrical methods; olive field; blends; quality control; ordination; clustering; pattern
recognition; prediction; chromatographic profiles; spectral data

1. Introduction
Olive oils (OOs) are complex food matrices due to their highly variable compositions. Different OO
cultivars are associated with different environmental living conditions, and are characterized by different
chemical patterns resulting in different organoleptic properties [1,2,3]. Thus, chemical patterns play central
interests between influencing environment and produced quality of olive products. These interests related to
intrinsic regulation and extrinsic sensitivity of metabolic profiles helping for specificity, authenticity and
traceability analysis of OO samples and populations [4,5,6,7,8]:

Specificity, authenticity and traceability of complex OO samples can be more or less controlled through
statistical analysis of variability between metabolic profiles. Links between these three basic concepts and
metabolic variability can be organized around three questions (Figure 1):

 What are the roles and status of separated metabolites in the polymorphism of OO
samples (Figure 1a)?
 What is the usefulness of metabolic profiles or spectroscopic features for authentication
of OO samples (Figure 1b)?
 How and how much the metabolites’ levels vary the ones relatively to the others to
favour the development or formation of well-distinct OO patterns? (Figure 1c)
48

Figure 1. General interests of applications of different statistical methods for specificity (a), authenticity (b)
and traceability (c) analysis of olive oil (OO) samples. Some methods are unsupervised (Un) whereas others
are supervised (S).

The first question focuses on highlighting of high or low regulation levels of some metabolites specifically
to some phenotypes, cultivars or environmental conditions. Metabolite-phenotype specificity can be
statistically highlighted by ordination methods including principal component analysis (PCA) and
correspondence analysis (CA).

The second question refers to authenticity consisting in combining several metabolites to define
chemical fingerprints from which different OO samples can be reliably classified or predicted. Classification of
a wide set of OO samples into different authentic groups is statistically carried out by hierarchical cluster
analysis (HCA). However, affiliation of outside samples into appropriate groups implies predictive models
based on pattern recognition techniques including linear discriminant analysis (LDA), Soft Independent
Modelling of Class Analogies (SIMCA), Support Vector Machines (SVMs) and K-Nearest Neighbours (KNNs).
Beyond this qualitative aspect, OO blends can be quantitatively evaluated by predicting some target
characteristics or components from high number of input variables (e.g., spectroscopic data). This question
can be treated by partial least square (PLS) regression.

The third question implies the analysis of memorized and interactive variations within heterogeneous
matrices. This traceability question requires statistical techniques able to absorb and separate different types
of variations associated with (or at the origin of) different polymorphic or multi-aspect patterns. These
techniques include artificial neural networks (ANNs) and simplex mixture design-based approach.
49

Apart from the goal-based criterion, these different statistical methods can be classified into supervised
and not-supervised types. The first ones do not need any preliminary information and provide neutral (not
guided) results which will help to understand complex structures of studied systems. These techniques
include PCA, CA and HCA. However, supervised methods are guided by preliminary information which will
serve as reference for target final results. Supervised techniques include LDA, PLS, SIMCA, SVMs, KNN and
simplex approach. Finally, ANNs, represent a set of methods including supervised or unsupervised ones.

This review provides an illustrated presentation of traditional and modern chemometrics methods applied
in olive oil field. Its ultimate objective is to provide a guideline linking directive questions of authenticity,
specificity and traceability to appropriate chemometrics methods. Basic methodological aspects are
completed by several recent illustrative applications showing the wide interests and perspectives of
chemometrics in this food field. Applications essentially concerned chromatographic and spectroscopic data
including HPLC, GC, UV, NIR, MIR, NMR data, etc. in addition to genetic markers. Chemical data concerned
several types of metabolites including fatty acids (FAs), triacylglycerols (TGCs), sterols, phenols, volatiles, etc.

Crucial interests of chemometrics analysis of olive oil samples included geographical and varietal origins
determination, standard composition control (conformity, purity, adulteration), historic handling process
detection, proportions’ evaluations of mixed OO cultivars in binary and multivarietal blends.

These qualitative pattern recognitions and quantitative evaluations of compositions usefully help for
more precise quality control and adulteration reduction.

2. Chemometrics Analysis of Specificity

Specificity refers to the research of several elementary traits which are specific or characteristic of OO
sample leading to its distinction within a polymorphic complex system. Several OO samples associated with
different geographical origins/environmental conditions were characterized by different variation ranges and
trends of metabolites. Specificity of different samples can be highlighted by ordination methods including
principal component analysis (PCA) and correspondence analysis (CA).

2.1. Principal Component Analysis and Correspondence Analysis

2.1.1. General Principle
PCA and CA are multivariate methods helping to graphically visualize trends of individuals governed by
partial correlations between variables characterizing a polymorphic system [9,10,11,12,13]. This is carried out
by compression of the large-dimension and highly dispersed initial dataset into a small and well-structured
space. This structuration is provided by new components, called principal components (CPs), combining all
the initial variables. PCs represent orthogonal directions along which the total variation is decomposed into
structured blocks; such blocks show hierarchically decreasing coverage of total variation.

PCA and CA are carried out by analysing variability of both columns (variables) and rows (individuals)
leading to dual analysis which highlights partial links between variables and subsequent behaviours of
individuals. By this way, individuals with similar or opposite behaviours are topologically identified by particular
levels and trends between some variables. This leads to identify specificity of several groups of individuals on
the basis of high or low levels for some variables (positive or negative trends between them).

2.1.2. Methodological Steps

PCA and CA algorithms can be summarized by three steps consisting of eigenvalues, eigenvectors and
PCs calculations (Figure 2).
50

Figure 2. Different methodological (computational and visualization) steps (a–c and d, respectively) of
principal components analysis and correspondence analysis.

The number of PCs is equal to that of eigenvalues and eigenvectors, and is defined by the lowest
dimension of data matrix X. If p < n, one expects to calculate p eigenvalues λj, p associated
eigenvectors Uj followed by p PCs (PCj).

Eigenvalues λj are weighting values quantifying the relative part of total variation along different PCs
(PCj) (Figure 2a). Eigenvectors are orthogonal and define the spatial directions of PCs (Figure 2b,d). PCs
contain the new coordinates of individuals and are calculated by linear combinations between the
eigenvectors and initial data matrix X (Figure 2c). In the p linear combinations, the coordinates of
eigenvectors represent algebraic coefficients quantifying the roles of different variables to the p PCs.

Finally, the two topological spaces (called score and loading plots) are considered in parallel ways to
analyse the relative behaviours of individuals under the effect of partial correlations between variables (Figure
2d).

From the subspace defined by a given factorial plot PjPk, different trends are identified from the projection
of different points at the extremities of the different PCs. Spatial proximity between projected points means
generally similar behaviours toward the different variables. On the other hand, in the factorial plot of variables,
positive, negative or not significant correlations are indicated by close, opposite or orthogonal projection of
points, respectively (Figure 2d). For example, Figure 2d shows partial negative correlations
between X1 and X2 along PC2 and between Xj and Xp along PC1.
51

Finally, by considering both row and column plots, some individuals are specifically characterized by
some variables on the basis of their projection in a same subspace. A variable has a particularly high or low
level in an individual if the two points (individual point and variable point) are projected in a same subspace or
two opposite subspaces, respectively. These graphical analyses help to identify particular subpopulations that
can be indicative of different phenotype or polymorphic trends. For instance, Figure 2d shows four trends (Tr)
among which, Tr1 and Tr2 are characterized by high levels of variables X1 and X2, respectively. Moreover,
these trends are opposite along the second principal component CP2 because of negative correlation
between X1 and X2.

2.1.3. Application of PCA and CA in OO Field

In OO field, PCA and CA were used to highlight the best regulated metabolites or characteristic
metabolic profiles of different biological varieties leading to a better distinction between them.

CA was applied to extract typical metabolic profiles of three French virgin olive oils (VOOs) characterized
by fatty acid profiles: Aglandau, Grossane and Salonenque [14]. The three cultivars showed relatively higher
regulations toward some fatty acids (FAs): 17:0 and 17:1ω8 for A, 16:1ω7 and 18:1ω7 for G, 18:2ω6
for S (Figure 3a).

Figure 3. Different study cases (a–e) on specificity of different OO monocultivars (a–c) or OO blends (d,e) by
ordination analysis (principal component analysis (PCA) or correspondence analysis (CA)) applied on different
physical-chemical parameters (fatty acids, UV absorbances).

Among ordination analyses, PCA was applied for inter-countries varietal differentiation from FAs and squalene
chromatographic profiles containing major and minor compounds [15]. PCA helped to differentiate:

Between Tunisian (6 varieties), Algerian (6 varieties) and Moroccan VOOs (1 variety) on

one hand,

Between different Tunisian VOOs on other hand.

Tunisian and Moroccan VOOs showed clear chemical differentiations compatible with high geographical
distance between the two countries. Also, clear differentiations were highlighted between Algerian and
52

Tunisian cultivars expect for the Blanquette variety which showed some overlapping with
Tunisian Chetoui variety.

The same PCA-based work showed different trends of different VOO varieties originated from a same country
(Tunisia) (Figure 3b):

Chemchali occupied specific topological subspace due to relatively high level of

saturated FAs (20:0, 22:0).

Chemlali and Zalmati showed relatively high level of three upstream chained
metabolites: 16:0 → 16:1ω7 → 18:1ω7. These two varieties were characterized by C7-
monoiunsaturattion vs. C9-monounsaturation in Chetoui and Oueslati.

These two later varieties (Chetoui and Oueslati) were characterized by relatively high
levels of C9-monounsaturated FAs (16:1ω9, 18:1ω9, 20:1ω9) linked to sequential
metabolic elongation processes.

In another study, Semmar et al. (2016) [16] highlighted chemical differentiation

between Chetoui and Oueslati by applying correspondence analysis (CA) on FAs (Figure 3c).

Although these two varieties showed relatively higher regulation of C9-monounsaturated FAs compared
to Chemlali, Chetoui had more affinity for 16:1ω9 and 20:1ω9 than Oueslati which showed relatively higher
regulation of 18:1ω9. Moreover, CA highlighted the lowest 16:1ω7 and 18:1ω7 levels in Chetoui, two minor
FAs characterizing Chemlali (because of maximal regulations). However, Oueslati showed the lowest level for
18:2ω6 occurring at significantly higher level in Chetoui and Chemlali.

At higher complexity level, PCA was applied to characterize several blends combining VOO with other
vegetable oils. This differential characterization is needed to avoid or reduce adulteration of food oils: (1)
some oils are often used as adulterant in OO blends because they are relatively cheap; (2) virgin olive oil
(VOO) can be added in other vegetable oils for nutritional values and economic reasons [17,18].

In this field, PCA was applied on UV-spectrometry data to highlight different variation poles specific to
different extra virgin olive oils (EVOOs) showing different adulteration types and levels due to presence-
absence of corn oil, palm oil, soybean oil and sunflower oil (Figure 3d) [4]. PCA applied on a dataset of
mono-, bi- and tri-oils samples revealed that bi-component oil blends containing EVOO were clearly separated
from mono-component (pure) oils along the first principal components. This good separation was revealed to
be due to high content of EVOO-containing blends in terms of compounds absorbing at 200–325 nm. The
authors suggested that compounds could be mono-unsaturated FAs (mainly oleic acid) which are abundant in
EVOO [4,19]. The second PC separated pure and EVOO-containing blends from adulterated blends by palm
oil (PO). Loading plot of UV-variables showed that this separation was specifically due to PO-characterizing
compounds absorbing at 325–350 nm.

PCA was applied to highlight association between quantitative levels of FAs and binary blends mixing
OO with sunflower oil at the proportions of 50%–50%, 60%–40% and 40%–60% (Figure 3e) [18]. Blends
containing 60% OO were characterized by relatively high concentrations of linolenic, oleic, arachidic and
margaroleic acids. This was explained by the fact that these FAs showed higher content in OO than sunflower
oil. However, blends containing 40% OO were characterized by other FAs including behenic, linoleic, myristic
and lignoceric acids which are more concentrated in sunflower oil.

3. Chemometric Analysis of Authenticity

Authenticity refers to the definition of general patterns of samples integrating different specific
characteristics. In OO filed, this helps to classify varieties and to confirm identities of denominated samples.
53

3.1. Hierarchical Cluster Analysis (HCA)

3.1.1. General Principle
Hierarchical cluster analysis (HCA) is a classification method used to organize a whole heterogeneous
population into well-distinct and homogeneous groups (called clusters) [20]. By this way, authentic groups are
firstly defined as clusters which are more or less separated by well calculated distances: distances are
calculated between individuals the closest of which are grouped into same clusters, whereas distant ones are
separated into different clusters. In a second technical step, resulting neighbour clusters are merged by
applying a given aggregation rule among several possible one: Combination between distance kind and
aggregation rule results in several possible tree-like typologies called dendrograms (Figure 4).

Figure 4. (a–e) Different agglomeration rules used in hierarchical cluster analysis (HCA); (f) Dendrogram or
tree-like diagram provided by HCA showing different class structures characterized by different distinction and
homogeneity levels (distinctness and compactness, respectively).

In a third step, the different clusters given by dendrogram are characterized by multivariate patterns due
to differential variations and/or non-overlapping variation ranges of different experimental parameters. Finally,
dendrogram helps to identify how many distinct groups are constitutive of a whole population on one hand,
and how much these groups are close/distant the ones to the others. This usefully helps to understand the
organization of whole studied population.

3.1.2. Methodological Steps

The first methodological step in HCA consists in calculating distances or similarity levels between
individuals. Distances are calculated when data are quantitative, whereas similarity indices are calculated
54

when the system is described by qualitative variables [20,21,22,23,24]. There are several types of distances
that can be used including Manhattan, Euclidean, Chi-2, etc. [20,25].

Similarity indices are also numerous including Sorensen-Dice, Jaccard, Simpson, Tanimoto coefficients
[11,20].

After distance calculation, all the individuals will be iteratively grouped or separated into different clusters
using an aggregation or linkage rule. These rules include single-, complete-, centroid-, average- and Ward-link
clustering [20,21,22,23,24] (Figure 4):

 In single link clustering, neighbour clusters are those having the closest individuals
(Figure 4a).
 Complete link clustering proceeds by joining clusters having the less distant extreme
individuals (Figure 4b).
 In centroid link clustering, neighbour clusters are those having the closest gravity centres
(Figure 4c).
 In average link clustering, an average distance is calculated from all the distances
separating all the pairwise of points. Clusters showing the lowest average distance will be
merged (Figure 4d).
 Ward linkage is based on the calculation of ratio between two variances: variance
between clusters on variance within group. Clustering leading to the lowest reduction of this
ratio will be applied (Figure 4e).
From dendrogram, different clusters are identified and interpreted on the basis of two criteria:
distinctness between clusters and compactness within clusters (Figure 4f). Distinctness defines the distance
separating a cluster from the rest (Figure 4f). Higher distinctness indicates a more distinct (differentiated)
cluster. Compactness defines the highest distance existing within a cluster (Figure 4f). Lower is the
compactness more homogeneous is the cluster.

3.1.3. Application of HCA in OO Field

HCA was applied to classify many populations of olives using different metabolic and physical-chemical
variables.

HCA was applied on phenol, triacylglycerol and sterol contents to statistically highlight the key role of
these metabolites in chemical differentiation of five Tunisian minor olive cultivars [26].

Also, HCA based on Manhattan distance and complete linkage was used to classify and calculate
separations between five Italian OO cultivars on the basis of their FA contents [27]: Both dendrogram and
PCA scores plot showed that Coratina and Oliarola cultivars were the most distant leading to their strong
differentiation. The three other EVOOs (Simone, Olivastro, Leccino) showed intermediate states with some
co-occurrence clusters in HCA or overlapping subspaces in PCA.

HCA was also applied on 31P NMR data using Euclidean distance and single linkage to explore
similarity and dissimilarity between 59 samples representing three OO groups: 34 Greek extra virgin OOs
(EVOO), 13 refined OOs (ROO) and 12 lampante OOs (LOO) [28]. Higher homogeneity (identity) was
highlighted for EVOO followed by ROO then LOO.

In another HCA-based work, OO samples were reliably differentiated from non OO ones on the basis of
ATR-FTIR spectra [17]. In this work, 111 pure oil samples were considered including 41 edible vegetable oils
and 70 OOs originated from European and American countries (Italy, France, Spain, US, Mexico). The 111
samples were characterized by 2584 wavenumbers covering a spectral region from 650 cm−1 to 3588 cm−1.
Using a variance-weighted distance and Ward’s aggregation method, a dendrogram was constructed to
highlight three clusters separating OO samples from two other oil categories (Figure 5a): a first and second
clusters represented flax seed oils and most of non-OO samples, respectively. They were frankly separated
from a third cluster made by the 70 OO samples in addition to four non-olive samples (one high oleic
sunflower, one safflower, two peanut oils). This suggested that some non OO have more similar physical-
chemical profiles with OO than other oils. More generally, the HCA results showed that information contained
in infrared spectra can be reliably used for OO distinction from other edible oils like canola, corn, soybean,
sunflower oils among others. Spectral bands of triglycerides containing unsaturated FAs provided the main
discrimination basis.
55

Figure 5. Studies cases on authenticity olive oils (a) and olive trees (b) highlighted by cluster analysis using
spectroscopic (a) and genetic (b) variables.

Beyond authenticity analysis applied at the scale of separated OO and non-OO samples, the same
authors applied HCA to carry out classification of pure OO and binary samples mixing OO with another
vegetable oil at variable percentages varying from 10% to 90% [17]. Co-occurring vegetable oils and OOs
were concerned with eleven European and American origins. Blends were characterized by infrared (IR) data.
Dendrogram obtained with Ward’s method highlighted clear separation between pure OO, non-OO and mixed
OO-vegetable oil samples.

Beyond physical-chemical and metabolic markers, HCA was applied on genetic markers (RAPD and
ISSR) for inter- and intra-cultivar classification of Portuguese Olive trees [29]. Inter-cultivar dendrogram was
obtained using Dice coefficient. It highlighted different similarity-dissimilarity degrees between the eleven
studied olive cultivars (Figure 5b): Galega, Madural and Blanqueta cultivars were more dissimilar from all the
others, whereas Negrinha and Azeiteira were the most similar between them. The Galega cultivar was then
subjected to intra-cultivar HCA using Jaccard index. Resulting dendrogram highlighted five clusters which
corresponded to five agro-ecological regions of OO production.

4. Chemometric Analysis of Authenticity and Traceability

Beyond authenticity which focuses on identification and origin determination of samples, traceability
refers to the quantitative evaluation and qualitative discrimination of components in complex matrices. In OO
field, traceability concerned adulteration detection of labelled products and composition evaluations of
heterogeneous or multi-varietal blends. A common authenticity-traceability problem concerned the prediction
56

of geographical origins of mono-varietal samples. It is generally treated by linear discriminant analysis among
other pattern recognition methods.

Pattern recognition methods include LDA, quadratic discriminant analysis (QDA), stepwise discriminant
analysis (SDA), Soft Independent Modelling of Class Analogies (SIMCA) and combining approach based on
partial least-square regression and DA (PLS-DA) [30]. For a given profile, these methods calculate several
scores corresponding to all the possible cultivars; then the maximal score is retained to attribute the unknown
profile to appropriate affiliation group.

4.1. Linear Discriminant Analysis

4.1.1. General Principle
Linear discriminant analysis (LDA) belongs to a set of statistical methods called pattern recognition
methods. These methods are applied to predict class membership (e.g., cultivars, geographical origins, etc.)
of unknown samples (individuals) from quantitative profiles made by several measured variables (e.g.,
metabolic concentrations). This goal is reached by constructing a statistical model combining p quantitative
variables Xj (jp1)(�1�) to distinguish between q patterns (q classes).

4.1.2. Methodological Steps of LDA

Attribution of quantitative individual profiles to appropriate classes requires several methodological
steps: patterns’ space is initially constructed, then transformed to reduced discriminant space in which
individuals will be finally classified [31].

Pattern space construction is carried out by characterizing each group (class) by its centroid position and
dispersion range: for a q-group system characterized by p quantitative variables Xj, centroid of group k is
spatially located by the average vector: Ck=(X¯¯¯1k,…,X¯¯¯jk,…,X¯¯¯pk)��=(�¯1�,…,�¯��,
…,�¯��). Internal dispersion of a group is given by its variance. For the simplistic example of q = 2 groups
(A and B) and p = 2 variables X1, X2, spatial locations of the two centroids are defined by the points of
coordinates (x¯1A,x¯2A)(�¯1�,�¯2�) and (x¯1B,x¯2B)(�¯1�,�¯2�) (Figure 6a). Distances between
centroids, and distances of individuals to their centroids are used to define two types of variances called
between- and within-classes variances, respectively.
57

Figure 6. (a) Separation of individual points into well distinct groups by a discriminant function combining
discriminant variables Xj; (b) Comparative illustration of discriminant and not-discriminant axes; linear
coefficients in discriminant function are determined by maximizing the ratio of variance between groups on
variance within group.

In a second step, the initial pattern space will be transformed by means of linear combinations of
the p variables Xj (jp1)(�1�) [31,32]. Each of the linear combinations provides a linear discriminant function
(DF) which represents a separating axis between the q groups in the new discriminant space (Figure 6a).

The coefficients wj of the different variables Xj in the different linear combinations are optimized by
maximizing the ratio of between-group variance on within-group variance. This constrains the q groups to be
as distinct as possible within a reduced space showing minimum loss of differentiation between the groups
(Figure 6b).

In a third step, DFs are used to calculate q classification scores (CSs) for each individual. The highest
score among the q ones will define the appropriate affiliation class of the considered individual. The q CSs
can also be converted into q probabilities among which the maximal value will be considered to attribute the
new individual to the most plausible group [33].

4.1.3. Applications of LDA in OO Field

LDA was widely applied for chemical authentication and traceability analysis of different OO classes
including mono-varietal VOOs or EVOOs, registered designations of origins (RDOs), geographical area- and
harvesting years, etc. Authentications were based on various discriminant variables including FAs, sterols,
triacylglycerols, phenols, volatiles [33,34,35,36,37,38,39]:
58

Using NMR data of phenolic and FAs compositional parameters, Petrakis et al. (2008) [37] applied
discriminant analysis for pattern recognition of different Greek EVOOs taking into account different spatio-
temporal scales: (i) three geographical divisions (Crete, Peloponnesus, Zakynthos) including (ii) six sites of
origins (Sitia, Heraklion, Chania, Messinia, Lakonia, Zakynthos) from which (iii) 131 samples were obtained at
different harvest years. Some metabolites seemed to play important role in discrimination of EVOOs
associated with geographical location: the three geographical sites of Crete division were highly discriminated,
with frank separation from Zakynthos division (one site) and intermediate state for Peloponnesus division (two
sites) (Figure 7).

Figure 7. Chemical discrimination of different Greek extra virgin olive oils (EVOOs) originated from different
geographical sites of three divisions (Crete, Peloponneus, Zakynthos) using NMR-analysed fatty acids and
phenolic compounds.

The high discrimination of Cretean sites was due their higher total concentrations of total and free
hydroxytyrosol and total tyrosol. In addition to phenolic compounds, linoleic (LA) and oleic (OA) acids provided
significant markers of geographic origin. Traceability performance was linked to two genes OeFDA2 and
OeFDA6 encoding the key enzymes for LA and OA pathway. These genes seem to have survived the
extensive gene flow. In Mediterranean olive trees, the gene flow seemed to have obscured the phylogenetic
signal but not the geographic one.

LDA was also applied on 1H NMR spectra of phenolic extracts to reliably separate between Italian OOs
according to both cultivar types and geographical origins [35]. In another work, LDA concerned mass-
spectrometry and UV data to differentiate between three geographical nominations of Italian (Ligurian)
EVOOs [40].

Beyond authenticity of separated OO samples, LDA was applied for authenticity analysis of OO blends.
A work was carried out on normalized Raman spectral in order to discriminate between pure OO, pure
soybean oil (SO) and OO-SO blends submitted at different temperatures (20–90 °C) [41]. Pure OO and OO-
SO blends (i.e., adulterated OO) under 90 °C were frankly well-discriminated by LDA compared to those
under 20 °C (Figure 8); this highlighted interactive discriminant effect linked to temperature elevation. This
temperature-dependent effect was associated with varied spectral features of the 1265 cm −1 band which was
related to the degree of unsaturation.
59

Figure 8. Physical-chemical discrimination of pure OO and adulterated OO oil under temperature of 90 °C

due to variation of spectral feature of 1265 cm−1 manifesting under temperature elevation.

OO blends authenticity analysis was also carried out by LDA to predict five French VOO registered
designation of origins (RDO) made by combining different cultivars as primary or secondary components.
RDOs consisted of Nyons (Y, n = 126 samples), Vallée des Beaux (VB, n = 98), Aix-en-Provence (PA, n =
99), Haute Provence (HP, n = 85) and Nice (C, n = 131) (Figure 9) [34]. The five RDOs were chemically
characterized by FAs and triacylglycerols (TGC) contents (14 FAS and 19 TGCs). Application of LDA on the
whole chemical dataset of 539 samples belonging to five RDOs showed that PA and VB were the closest
origins because of minimal Mahalanobis distance separating them. However, Y was the most distant RDO
flowed by C leading to higher authenticity and identicalness. Although all the variables were useful for pattern
recognition, some ones revealed to be more particularly influent in RDOs’ discrimination because of higher
standardized discriminant coefficients which concerned:

 Triolein (OOO), linoleoyl-dioleoyl-glycerol (LOO), palmitoyl-dioleoyl-glycerol (POO),

dilinoleoyl-oleol-glycerol (LOL) for TGCs
 Monounsaturated FAs sum, oleic, palmitic, stearic and hypogeic acids for FAs.

Figure 9. Chemical discrimination of five French VOO registered designation of origins (RDOs) made by
mixing different OO cultivars and characterized by 14 fatty acids (FAs) and 19 triacylglycerols (TGC) among
which four FAs and four TGCs were the most discriminant. Legend: OOO, triolein; LOO, linoleoyl-dioleoyl-
glycerol; POO, palmitoyl-dioleoyl-glycerol; LOL, dilinoleoyl-oleol-glycerol.
60

At higher complexity level, LDA was applied on FA profiles to predict multivarietal quantitative
compositions of OO blends varying by different weights (proportions) of co-occurring French and Tunisian
cultivars [14,16]: such a problematic goes beyond qualitative discrimination of separated cultivars and RDO
(blend) types because it makes to evaluate percentages of several co-occurring OOs in same blends from
chemical profiles of blends. Application of LDA for prediction of multivarietal OO blends concerned French and
Tunisian samples and it will be presented in Section 4.7.3 [14,16].

4.2. Partial Least Square Regression

4.2.1. General Principle
PLS can be applied for differential sorting of many input variables according to their pertinence levels for
the prediction of the output variable Y. This helps for extraction of a small number of reliable predictors
of Y among a huge mass of many candidate input variables X. This is particularly interesting for datasets
containing few individuals and many variables [42,43].

Moreover, PLS is a favourite method of treatment with possibly correlated predictors which are
combined and condensed with all the initial variables into new composite variables called latent variables
(LVs) (or components or factors).

Like the principal components (in PCA or CA), LVs are newly constructed variables which weight and
condense the initial variables to handle with their highly diversified and/or correlative aspects. However, in
PLS regression (PLSR) differs from PC regression (PCR) by the fact it incorporates information on
both X and Y in the definition of LVs:

In PLSR, LVs are calculated such that the covariance between Y and X is maximal. By this
way, LVs are built on the basis of both the input (X) and output (Y) variables.

In PCR, one focuses on maximization of the variance of X in different spaces. Only the
input variables X are used for LVs construction.

4.2.2. Methodology of PLS Regression

Initially, PLS proceeds to decomposition of predictors’ matrix X into orthogonal scores T and loading P:
X = T·P

This makes to regress the output variable Y on the first columns of the scores instead of X. Therefore,
prediction of the output variable Y by PLSR is carried out iteratively by adding a single LV at each step. By
this way, several enclosing PLSR models are obtained to predict the output variable Y through linear functions
of the first k LVs. The best model can be concluded from a minimal prediction error obtained after cross-
validation.

The final PLS model has the double advantage to:

Predict the output variable Y from a small number of LVs.

Help for identification of the most contributive output variables (e.g., spectral
wavenumbers).

Thus, PLSR offers an efficient chemometrical tool for variable selection leading to extract the most
pertinent predictors of complex and highly variable system such as adulterated OO blends. Before quantitative
evaluation of adulteration level (proportion), LDA can be applied to determine the nature of adulterant.

4.2.3. Application of PLS Regression in OO Field

Olive oils can be adulterated by adding some cheap vegetable oils at different proportions in the
commercial blends. Such adulterations are qualitatively and quantitatively analysed by means of
61

spectroscopic features including near-infrared (NIR), mid-infrared (MIR) and Raman data. These techniques
provide highly condensed information on the mixture matrices through large numbers of recorded
wavenumbers (hundreds or thousands of variables).

Many chemometrical works were carried out by treating IR data by PLSR in order to predict the
adulteration level of commercial EVOOs by other vegetable oils added at variable proportions. The list of
vegetable oils concerned with adulteration includes: canola, hazelnut, pomace, sunflower, corn, soybean,
walnut, etc. [6,44,45,46,47,48,49].

4.3. PLS-DA Approach

4.3.1. General Principle
PLS-DA refers to a method combining partial least square (regression) and linear discriminant analysis.
Like LDA, it aims to optimize separation between different groups characterized by several quantitative
variables [50]. It differs from LDA in that the q groups are initially represented in q separated indicator columns
(0, 1), whereas in LDA all the groups are represented by one column containing q labels (Figure 5).
Therefore, optimization of separation between groups is carried out by linking the (1, 0)-binary matrix Y to
quantitative matrix X containing p discriminant variables.

4.3.2. Methodological Steps of PLS-DA

Initially, q class membership columns are prepared by taping 1 or 0 if individuals belong or not to
corresponding classes, respectively. Then, optimization of separation between the q groups is performed by
maximizing the covariance between the p discriminant variables Xj and the q candidate groups (Figure 10).

Figure 10. Difference between linear discriminant analysis (LDA) and partial least square regression-
discriminant analysis (PLS-DA) based on different initial labelling of groups.

Space reduction is carried out by calculating integrative variables, called latent variables (LVs), from
linear combinations of X (LVX) and Y (LVY), respectively. Coefficients of linear combinations are calculated
under the maximization constraint of covariance of LVX and LVY. This leads to define a linear classifier which
62

has proved to be statistically equivalent to that of LDA, but with noise reduction and variable selection
advantage [51].

Finally, PLS-DA model predicts continuous values of Y variables which are comprised between 0 and 1.
Values closer to 0 indicate that the individual does not belong to corresponding classes; however, closer the
value to 1 higher the chance of individual to belong to the concerned class. Therefore, individuals are
assigned to the classes which have the maximum score in the Y-vector. Alternatively, threshold values
(between 0 and 1) can be defined for the different classes using Bayes theorem [51].

4.3.3. Application of PLS-DA in OO Field

PLS-DA models were applied on spectroscopic data (MIR, NIR, FTIR, 1H NMR) to predict authenticity
and traceability of different EVOOs, protected designation of origins (PDO), registered designation of origin
(RDO) and geographically-linked samples [52,53,54,55,56].

Bevilacqua et al. (2012) [52] applied PLS-DA on MIR and NIR data for pattern recognition of
geographical origins of Italian EVOOs shared between Sabina and not Sabina (confounded other sites)
(Figure 11).

Figure 11. Prediction of geographical origin of Italian EVOOs by PLS-DA using mid- (MIR) (a) and near- (NIR)
(b) infrared data.

Before PLS-DA, MIR and NIR data were subjected to different pre-treatment for baseline corrections
consisting of: (a) linear; (b) quadratic (Q); (c) multiplicative scatter (MS); (d) detrending (D); (e) first derivative
(1d); (f) second derivative (2d); (g) MS with 1d; (h) MS + 2d; (i) MS + Q; (j) MS + Dt. NIR and MIR datasets
issued from pre-treatments were modelled using calibration (Cbr) and cross-validation (CrV) techniques.

Using MIR data, the best predictive PLS-DA model was built by MS pre-treatment (followed by QB). This
model predicted Sabina at 100% and 92.3% in Cbr and CrV vs. 95% for other origin in both Cbr and CrV.
Moreover, outside samples were given to PLS-DA model (for final validation) which predicted Sabina and
other origin at 85.7% and 86.7%, respectively. Analysis of discrimination ability of MIR variables showed that
C=O double bond stretching (around 1700 cm−1) and C–H bending (within 650–750 cm−1 range) contributed the
most to the PLS-DA model (Figure 11a). They were followed by significant (but lower) contributions of C–H
stretching (2800–3100 cm−1) and C-O (single bond stretching) (1100 cm−1).
63

Using NIR data, four PLS-DA models showed high predictive aspect both in Cbr and CrV. They
concerned MS, Dt, 1d and MS + D which predicted Sabina and other origin at 100% and 95.5%, respectively.
Final validation based on outside sample was concluded by predictive ability of: (i) 100% for both Sabina and
other origin with 1d-based model; and (ii) 100% Sabina vs. 93.3% other origin with the three other models.
Discriminant variables included all the spectral regions (Figure 11b):

 4500–5000 cm−1 due to combination bands C=C and C–H stretching variation of cis
unsaturated FAs.
 5650 and 6000 cm−1 due to combination bands and first overcome of C–H of methylene of
aliphatic groups of oil.
 7074 and 7180 cm−1 linked to C–H band of methylene.
Apart from DA, PLS was combined with linear regression (PLSR) to predict proportions (0% to 5%) of
EVOO and Palm oil (PO) in three vegetable oil blends based on corn oil (CO), soybean oil (SO) and sunflower
oil (SFO), respectively [4]. EVOO and PO represent desired and undesired (adulteration) oils in vegetable oil
blends. Blends were initially characterized by UV spectra containing many wavelength-absorbance variables.
The high number of these variables was reduced by applying PCA which provided principal components used
as latent variables in PLSR. PLSR models predicted EVOO proportions in EVOO-vegetable oil blends with
determination coefficient R2 = 0.86, 1.00 and 0.98 in EVOO-CO, EVOO-SO and EVOO-SFO blends,
respectively. For adulterant PO in EVOO-vegetable oils, PLSR predicted its proportions with R2 = 0.85 (in
EVOO-CO), 1.00 (in EVOO-SO) and 0.99 (in EVOO-SFO).

Geographic origins and compositions of virgin

olive oils determinated by chemometric analysis
of NIR spectra.
1. Introduction Nowadays, one of the major problems in the agricultural–food industry is to set
down objective tools in order to determine the origin of raw materials as well as finished
products in order to ensure their traceability. Virgin olive oils come from very many olive
varieties which are generally related to a soil that confers specific sensory [1] and chemical
properties. In order to produce high qualities of virgin olive oils and to identify clearly those

origin (RDO), which controls all the production. The French ∗ Corresponding author. Tel.: +33
oils, the European community creates some quality-labels such as the registered designation of

491289036; fax: +33 491636509. E-mail address: [email protected] (O. Galtier).

production concerns seven RDOs (“Aix-en-Provence”, “Corse”, “Haute-Provence”, “Nice”,
“Nˆımes”, “Nyons” and “Vallee´ des Baux”). The virgin olive oil is one of the more expensive
vegetable oils. Moreover, the price of virgin olive oils varied in large proportion (1 to 10).
Because of high cost, notably with the RDO label, virgin olive oils could be the subject of
fraudulent practices which consist of mixing them with various amounts of seed oils, refined
olive oils at low cost or olive pomace oils. The search for the origin and the authenticity of olive
oils has been the object of numerous studies in the past few years using the extremely varied
physical-chemical determinations that are associated with a chemometric treatment. The
studies could be classified in two main categories. In the first category, the samples are
chemically treated in order to determinate the composition in different constituents: fatty
acids and triacylglycerols [2–4], sterols [5] and aroma [6,7]. The second 0003-2670/$ – see
front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2007.02.033 O. Galtier
et al. / Analytica Chimica Acta 595 (2007) 136–144 137 category is based on spectroscopic
study on samples without preliminary treatment: 1H and 13C NMR [8,9], Fourier transform
infrared spectroscopy (FTIR) [10,11] and fluorescence spectroscopy [12–16]. Spectroscopic
techniques associated to chemometrics method present an alternative to chromatography
64

[17]. Molecular spectroscopy techniques, such as near infrared (NIR), have been shown to be
useful for oil and fat classifications[18,19] and for study on lipid oxidation [20]. NIR has been
used for quality control of olive oils [21] and as alternative method for quantitative
determination of peroxide value and acidity [22]. NIR has been used also to determine total
levels of chlorophyll and carotenoid pigments in virgin olive oils [23]. Measurement of
adulteration of virgin olive oils by different vegetable oils or by olive pomace oils [24–27] was
based on a chemometric analysis of the NIR spectra. Authentication of Spanish virgin olive oils
of very close geographical origins [28] and geographical classification of extra virgin olive oils
from the eastern Mediterranean [29] have been looked carefully by chemometric analysis of
the NIR spectra. The aim of this study was to develop by FT-NIR spectroscopy associated to
chemometric treatment, a direct and rapid test method that quantified fatty acid and
triacylglycerol contents in virgin olive oils in order to discriminate those virgin olive oils
according to their French RDOs. 2. Partial least square regression (PLS) theory This supervised
analysis was based on the relation between the signal intensity and the modification of the
sample [30]. Interference and overlapping of the spectral information may be overcome by
using powerful multicomponent analysis such as PLS [31,32]. This method allows a
sophisticated statistical approach using the full spectral region rather than unique and isolated
analytical bands. The algorithm is based on the ability to mathematically correlate spectral data
to a property matrix of interest while simultaneously accounting for all other significant
spectral factors that perturb the spectrum. It is thus a multivariate regression method that uses
the full spectral region selected and is based on the use of latent variables. The model was
build by block cross validation method of during the calibration developments. Each block
contains 10 samples. The evaluation of the errors in the calibration is estimated by computing
the standard error of calibration (SEC) after comparing the real concentration with the
computed one for each component. The formula for the standard error of calibration is: SEC =
N i=1 (Ci − C ′ i ) 2 N − 1 − p (1) where Ci is the known value, C ′ i is the calculated value, N the
number of samples and p is the number of independent variables in the regression optimized
by cross validation. The standard error of prediction (SEP) gives an estimation of the prediction
performance during the step of validation of the calibration equation SEP = M i=1 (Ci − C ′ i ) 2
M (2) where Ci is the known value, C ′ i is the value calculated by the calibration equation, and
M is the number of samples in the prediction set. Another useful parameter is the relative error
of prediction (REP) that shows the predictive ability of the model and calculated from equation
REP = SEP y¯ × 100 (3) where ¯y was mean of the observed values of crop variables. The
predictive ability of the model should also be expressed by the bias and the square of
correlation coefficient (R 2 ) also called determination coefficient, usually called Q 2 in
prediction. The regression coefficients are the numerical coefficients which express the link
between the predictor variations and the response variations. The bias is systematic difference
between predicted and measured values. The bias is computed as average value of the
residuals. The residual is the measure of the variation which is not taken into account by the
model. The residual for a given sample and a given variable is computed as the difference
between observed value and fitted (projected or predicted) value of the variable on the
sample. PLS regression is not previously suited to pattern recognition problems, such as
classification purposes. However, this technique can be adapted for classification [33–36] giving
rise to the partial least square discriminant analysis method (PLS-DA). PLS1-DA is carried out
using an exclusive binary coding scheme with one bit per class, providing a quintuplet {a; b; c;
d; e} if one wants to discriminate between five classes. Each number represents a “membership
value” for each class, e.g., a response encoded {0; 1; 0; 0; 0} means that the sample belongs to
class 2. During the calibration process, the PLS1-DA method is trained to compute the five
65

“membership values”, one for each class; the sample is then assigned to the class showing the
highest membership value [37]. For the codification of olive origin, the five origins
corresponding to the RDOs are arbitrary classified in the order “Aix-en-Provence”, “Haute-
Provence”, “Nice”, “Nyons” and “Vallee des Baux”. Thus, for each sample, the origin may be ´
represented by a five-dimensional output vector with 1 at the positions corresponding to
geographic origin and 0 at the other positions. For instance, the sample numbered 2, which is
“Vallee´ des Baux” origin, will be codified by the vector {0; 1; 0; 0; 0}. Five models are
computed, one for each origin. 2.1. Software The chemometric applications are performed by
the UNSCRAMBLER software version 9.2 from CAMO (Computer Aided Modelling, Trondheim,
Norway). 138 O. Galtier et al. / Analytica Chimica Acta 595 (2007) 136–144 3. Materials and
methods 3.1. Virgin olive oil samples Commercial virgin olive oil samples (n = 135) were
obtained from the French Inter-Professional Olive Oil Association (AFIDOL), Aix-en-Provence,
France and from the Direction Gen´ erale de la Concurrence, de la Consommation et de ´ la
Repression des Fraudes (DGCCRF) laboratory, Marseille, ´ France. Sampling was carried out
during two successive crops (2002/2003-2003/2004). 125 virgin olive oil samples came from
five French RDOs (Fig. 1). 102 samples were used in the calibration set and 23 samples in the
prediction set which were not used to calculate models. The samples used in the prediction set
were randomly selected. Table 1 shows the different constituted cultivars of RDOs. The RDOs
are principally made up of primary and secondary cultivars and also local and old varieties.
“HauteProvence” (n = 24), “Nice” (n = 23) and “Nyons” (n = 24) are made up of one unique
principal cultivar. “Aix-en-Provence” (n = 27) and “Vallee des Baux” ( ´ n = 27) have up to three
or four principal cultivars of which at least two do not have their proportions specified. 10
virgin olive oil samples were the commercial ones without any RDOs. 3.2. Fatty acid
determination Olive oil in n-heptane (0.12 g/2 mL) was transmethylated with a cold solution of
KOH (2 M) (1 mL). Fatty acid methyl esters (FAME) were analyzed on a Perkin-Elmer
Autosystem 9000 gas chromatograph (GC) equipped with a split/splitless injector (t = 250 ◦C)
and flame ionization detector (FID) (t = 250 ◦C). A silica capillary column (60 m × 0.25 mm i.d.,
0.25m film thickness) coated with DB WAX (polyethylene glycol, JW) was used. The inlet
pressure of the hydrogen as carrier gas was 154 kPa with a ratio: 70. The oven temperature
program was as follows: 13 min at 200 ◦C, from 200 to 230 ◦C at 6 ◦C min−1 , 17 min at 230 ◦C.
Fig. 1. Geographical areas of five French registered designations of origin (RDO). (RDO 1: “Aix-
en-Provence”; RDO 2: “Haute-Provence”; RDO 3: “Nice”; RDO 4: “Nyons”; RDO 5: “Vallee des
Baux”). ´ 3.3. Triacylglycerol determination Triacylglycerols were analyzed by a HPLC system
composed of a Merck liquid chromatograph Model LaChrom equipped with a Merck RP-18
Supersphere 100 column (250 mm × 4 mm i.d., 4m particule size, temperature 28 ◦C) coupled
with a Merck refractometric L-7490 detector. A sample loop of 100 L capacity was used in
which 10L were injected. Propionitrile (Carlo Erba, Milan) were the mobile phase with a flow
rate linear gradient (0.5 to 1 mL min−1 ) for 47 min. Triacylglycerols in olive oils were separated
according to equivalent carbon number (ECN), defined as CN-2n, where CN is the total acyl
carbon number and n is the number of double bonds of fatty acids. 3.4. Near infrared
spectroscopy FT-NIR spectra were recorded with a Nicolet Antaris spectrometer interfaced to a
personal computer. Oil samples were filled into a 2 mm pathlength quartz cell directly sampled
from Table 1 Varieties from five French registered designations of origin (RDOs) RDOs Aix-en-
Provence Haute-Provence Nice Nyons Vallee des Baux ´ Primary varieties Aglandaua Aglandau
Cailletier Tanche Aglandau Cayanne Grossane Salonenque Salonenque Verdale BRb Secondary
varieties Bouteillan Bouteillan – – Picholine Grossane Picholine Picholine Tanche Verdale BRb
Ribiere, Sabine Boube Araban – Yes ` Saurine, Sigoise Colombale Blanquetier Local and old
varieties Triparde Estoublaisse Blavet Filayre Nostral Grapie Ribeyrou ´ a Synonym of Berugette.
66

´ b Verdale des Bouches du Rhone. ˆ O. Galtier et al. / Analytica Chimica Acta 595 (2007) 136–
144 139 the bottle without any chemical treatment. All the spectra were computed at 4 cm−1
resolution between 4500 and 10000 cm−1 thanks to the software result integration 2.1 Thermo
Nicolet. Co-addition of symmetrical interferograms on 10 scans was performed for each
spectrum. A reference spectrum was recorded as before each sample measurement on an
empty tube. 3.5. Nomenclature Fatty acids: 14:0, myristic acid (tetradecanoic acid); 16:0,
palmitic acid (hexadecanoic acid); 16:1n − 9, hypogeic acid (7-hexadecenoic acid); 16:1n − 7,
palmitoleic acid (9- hexadecenoic acid); 17:0, margaric acid (heptadecanoic acid); 17:1n − 8,
margaroleic acid (9-heptadecenoic acid); 18:1n − 9, oleic acid (9-octadecenoic acid); 18:1n − 7,
cisvaccenic acid (11-octadecenoic acid); 18:2n − 6, linoleic acid (9,12-octadecadienoic acid);
18:3n − 3, linolenic acid (9,12,15- octadecatrienoic acid); 20:0, arachidic acid (eicosanoic acid);
20:1n − 9, gondoic acid (11-eicosenoic acid) (2); 22:0, behenic acid (docosanoic acid); 24:0,
lignoceric acid (tetracosanoic acid). Triacylglycerols: The triacylglycerols (TAG) are designated
by letters corresponding to abbreviated names of fatty acid carbon chains that are fixed on the
glycerol. The abbreviations of fatty acids names are: P, palmitoyl; Po, palmitoleyl; S, stearoyl; O,
oleoyl; L, linoleoyl; Ln, linolenoyl and A, arachidoyl. 4. Results and discussion 4.1. Quantification
of fatty acid and triacylglycerol compositions of virgin olive oils Spectra obtained for all samples
of virgin olive oils were similar to naked eye. Typical spectra of two RDO virgin olive oil samples
are presented on Fig. 2. The band assessments were realized according to literature [19,38].
Fatty acid and triacylglycerol compositions of virgin olive oils were determined by Fig. 2. Near-
infrared spectra of two RDO virgin olive oil samples. (A) Combination of the CH stretching
vibration with other vibrational modes. (B) First overtone of CH stretching vibration (methyl,
methylene and ethylene groups). (C) Combination of the CH stretching vibration. (D) Second
overtone of CH stretching vibration (methyl, methylene and ethylene groups). chemometric
analysis of NIR spectra using PLS1 algorithms. The prediction models were achieved using
different pretreatments. The best prediction results were obtained using either raw spectral
data in the 4500–7500 cm−1 range or full spectrum. The reference data on fatty acid and
triacylglycerol compositions (n = 125) were done by chromatographic methods. Table 2 shows
rate ranges of five components of virgin olive oils coming from the five French RDOs. Tables 3
and 4 give the results for 34 compounds obtained by NIR spectra.

Analytical Methods Detection of

adulteration of extra-virgin olive oil by
chemometric analysis of mid-infrared
spectral data.

left to right for rapeseed adulteration. Also, Coomans’ plot of corn– sunflower adulteration
reveals discrimination of adulterated samples from non-adulterated oil samples. In a previous
study, analysis of different varieties of olive oils in Turkey with mid-IR spectra and PCA resulted
in tolerable discrimination between Ayvalik and Memecik olive varieties (Gurdeniz, Ozen, &
Tokatli, 2008). Consequently, the model constructed here appears to be more efficient in that it
can group significantly different olive oil varieties together and discriminate adulterated olive
oil samples. There are other studies involving the use of sunflower–olive oil and corn–olive oil
67

binary blends to detect olive oil adulteration with FT-IR. Tay et al. (2002) were able to
discriminate olive oil samples from sunflower adulterated samples at levels between 2% and
10%. Also, results of FT-IR analysis indicated that the detection limit for olive oil adulteration
was 9% if the adulterant is only corn oil while it was lower (6%) if the adulterant is sunflower oil
(Vlachos et al., 2006). However, each of these studies includes detection of adulteration from
binary mixtures. As far as the authors are aware, no published work has focused on the
detection of ternary olive oil adulteration except the study of Özdemir and Öztürk (2007)
whose study focused on quantification of binary and ternary adulteration of olive oil with
sunflower and corn oil using NIR in conjunction with genetic inverse least square. Overall, SEP
ranged between 2.49% and 2.88% (vol/vol) for the binary mixtures of olive and sunflower oil
whereas it was between 1.42 and 6.38% (vol/vol) for the ternary mixtures of live, sunflower
and corn oil. Quantification of adulterant oil content in adulterated oil samples was performed
using PLS algorithm. The same data set except 2% adulterant level was used in PLS analysis. The
samples of all the adulterated and pure olive oils were randomly divided into a calibration and
a validation set. OSC was applied to remove systematic variation in X (spectral data) that is not
related with Y (per cent adulteration). In addition, wavelength compression was selected to
increase the efficiency and speed. Fig. 3 displays actual versus predicted percentages of
cottonseed, rapeseed oils and corn–sunflower binary oil mixture in olive oil for calibration and
validation sets. Difference between the actual and predicted adulterant oil concentration is
small indicating the success of PLS model. The statistics of PLS of calibration and validation sets
are given in Table 2. High R2 value and low SEP and SEC values demonstrate high predictive
ability of each model. Christy et al. (2004) employed NIR spectra with PLS to detect olive oil
adulteration with corn, sunflower and hazelnut oils and error terms were determined as 1.32
and, 0.57 for sunflower and corn oils at levels of 0–100%. In the last part of the study,
adulterated oils regardless of the type of adulterant were grouped as one adulterant class and
subjected to PCA to establish if authentic olive oil samples could be distinguished from
adulterated samples. Observations were classified as adulterated including spectra of rapeseed
and cottonseed oils and corn–sunflower binary oil mixture and pure olive oil samples. It is
important to determine if this model would be able discriminate an olive oil sample
adulterated with an unknown adulterant from pure olive oils. To construct Coomans’ plot (Fig.
4a) PCA was performed on adulterated and pure classes separately. Modeling adulterated oil
samples together means more variable chemical information making them more difficult to be
placed in the same group. Thus, 2% and 5% adulterated olive oil samples are placed so close to
pure olive oil samples according to scores plot (not shown). Therefore, the samples (2% and 5%
adulterated) were excluded from the data set. According to Coomans’ plot, most samples of
adulterated and pure samples are correctly placed in their region. However, few samples are
plotted in the region where they can not be classified as pure or adulterated. Same classified
data was also manipulated with PLS-DA and scores plot is shown in Fig. 4b. Pure and
adulterated samples are placed in right and left hemispheres, (Fig. 4b) indicating quite
successful discrimination using PLS-DA. SIMCA and PLS-DA are both supervised pattern
recognition techniques that aim to establish classification. However, PLS-DA classifies samples
into one of the given categories, even if they do not belong to any of them. On the other hand,
SIMCA visualized with Coomans’ plot is able to discriminate nonmembers (Berrueta, Alonso-
Salces, & Héberger, 2007). Therefore, visual discrimination seems to be better for PLS-DA than
Coomans’ plot. However, samples that do not belong to any group can also be displayed by
Coomans’ plot (Fig. 4a). There are studies involving the application of other analytical and
chemometric methods to discriminate olive oil from set of different adulterant oils. Availability
of chromatographic profiles with SIMCA model to distinguish between pure olive oil samples
68

and those adulterated with one of the vegetable oils (sunflower, corn, peanut and coconut oils)
was illustrated (Capote et al., 2007). Also, NIR spectra manipulated with PCA was able to
classify adulterated olive oil samples with respect to type of adulterant oil (Christy et al., 2004).
Amongst these and similar studies, mid-IR spectroscopy data was not employed before to
construct one model including adulterated olive oil samples using different oils. In fact one can
not know the type of adulterant used to adulterate an olive oil sample and employment of such
an overall model can be reasonable. In conclusion, manipulation of mid-IR spectra with
chemometrics leads to detection and quantification of olive oil adulteration. Fig. 4. (a)
Coomans’ plot and (b) scores plot of PLS-DA for corn–sunflower binary mixture, cottonseed and
rapeseed adulterated versus pure olive oil samples (d nonadulterated, N adulterated olive oil
samples). 524 G. Gurdeniz, B. Ozen / Food Chemistry 116 (2009) 519–525 Besides, detecting
potential adulterants, cottonseed and rapeseed oils, in olive oil at a level of 5%, detection limit
of binary mixture of corn–sunflower oil in olive oil is also found out as 5%. Furthermore, mid-IR
spectroscopy with chemometrics has the potential to detect adulteration of olive oil regardless
of the type of adulterant oil but at a higher adulteration detection level. Quantification studies
resulted in calibration curves with R2 values around 0.9. In the further studies, this work can be
expanded by including samples from other regions and other parts of the World and these olive
oil samples can be mixed with common adulterant oils. It is likely that the use of Mid-IR results
from an expanded vegetable oil reference data set will result in more comprehensive models to
improve the detection and quantification of adulteration of olive oils.

Chemical composition of olive oil

Olive oil is mainly a mixture of triglycerides, namely triesters of glycerol with fatty

acids. Some of the fatty acids are saturated and some others are unsaturated. Apart

from triglycerides, olive oil holds a small number of other ingredients, which are

coming from olive fruit including:

31

• Free fatty acids

• Phospholipids

• Sterols

• Aliphatic alcohols

• Phenols

• Tocopherols

• Pigments

• Volatile organic components

The synthetic arrangement of olive oil has resulted as an object of numerous scientific
69

researches. It constitutes from saponifiable (99-99. 5%), as well as from other

unsaponifiable substances, which are known as minor components (0.5-1%), which

contribute to its stability and its sappiness. Saponifiable contain substances such as

triglycerides, free fatty acids and phosphatides and unsaponifiable contain substances

such as hydrocarbons and fatty alcohols. The composition of olive oil can be used for

the appreciation of its authenticity. (33)

Chemical composition in fatty acids

The chemical composition of olive oil in fatty acids is not stable. Elements in particular:

➢ The diversity of olive

➢ The climate changes of the location

➢ The grade of full growth of olive fruit and other factors, influence the

constitution.

The most important fatty acids of olive oil are unsaturated. Between them, the one

with the highest amount is monounsaturated elaic (C 18:1). The second most

important unsaturated fatty acid is linoleic (C 18:2). Other unsaturated fatty acids

which exist in smaller amounts are linolenic (C 18:3), and arachidonic (C 20:4). From

the saturated acids, the ones that exist in the highest amount are palmitic (C 16:0) and

stearic (C 18:0).

The main glycerides of olive oil are the ones that exist in oleic acid, as they represent

70-80% of the total weight of olive oil. These triglycerides stay in a liquid state in the

temperature room, so the final olive oil also stays in a liquid state.

According to the structure of olive oil in fatty acids, scientists resulted in two types of

olive oils. The first one includes olive oils with low content in linoleic and palmitic and

high in elaic acid and the second one includes a high content of linoleic and palmitic

and low content in elaic acid.

The environment has a major role in the final composition of olive oil in fatty acids.

Factors such as harvest, variety, and origin influence the synthesis of the main fatty

32

acids of olive oil in different ways. Temperature seems to be a catalytic factor for the

unsaturated character of fatty acids of olive oil, which increases with the reduction of

the temperature and the rise of elevation. (34)

5.1.1 Phospholipids

They constitute a small ratio of the components of olive oil and the information that

exists regarding the chemical composition and the concentration in virgin olive oil is
70

not enough. The differences in the results of some researches mostly have to deal

with the analytical method which has been used. The main phospholipids which have

been observed from TLC chromatography are phosphatidylcholine,

phosphatidylethylamine, phospatidylinositol and phosphatidylserine. With the use of

liquid chromatography, phosphatidylglycerol was the main phospholipid that was

found.

Generally, it can be noticed that there is not much evidence for the nature and the

character of phospholipids in olive oil and their concentration as well. When they exist

in a sufficient concentration, they can contribute to the rise of stability of the olive oil.

More specifically, phospholipids act as secondary antioxidants because they disrupt

hydroxyl peroxides and they form complexes with preoxidant metals. It has been

found that they have a synergistic effect with tocopherols and they work as blocking

agents in the surface of olive oil.

Years later, it was found that the concentration of phospholipids in olive oil is between

35-40mg/kg. The biggest quantity of these phospholipids is coming from the nucleus of

the olive fruit. The phospholipids which exist in olive oil are lecithin and cephalin. (35)

Basic classes of unsaponifiable components of olive oil

5.1.2 Hydrocarbons

In olive oil, one part of the hydrocarbons which exist is saturated and the other part is

unsaturated. These are byproducts of the biosynthesis of fatty acids.

The two basic hydrocarbons which are necessary to exist in olive oil are squalene and

β-carotene. Squalene is the basic unsaponified component of extra virgin olive oil and

is at the level of 450mg/100g. With a high consumption of EVOO, the reception can

reach 200-400g/day, a phenomenon very well known in the Mediterranean countries.

It is known as a component with high biological value due to its prophylactic action.

Another component such as β-carotene exists in a very small amount such as

0.36mg/100g.

33

5.1.3 Carotenoids

Carotenoids that exist in olive oil are the main reason for its yellow output. Lutein,

which exists in xanthophylls, is the basic carotenoid in olive oil. Some other important

carotenoids are carotenoids (α, β, γ), and the one with the highest concentration in

olive oil is β-carotenoid with 85%, and α-carotenoid follows with 15%. The main

concentration of lutein and carotenoids in the olive oil is 0.085-0.496 mg/100gr. (36)
71

5.1.4 Chlorophylls

Olive oil contains pigments such as chlorophylls a and b. Chlorophyll a has a cyan-green

colour and chlorophyll b has a yellow-green colour, and these are the substances that

give the characteristic green colour in the olive oil. On the other hand, chlorophylls are

also the main reason for the degradation of its quality especially when there is contact

with light.

The total amount of chlorophylls in olive oil can be from 1 to 10mg/kg. With the

progress of maturity of the olive fruit and with the timing of storage of the olive oil,

there is a reduction of the content of chlorophylls.

5.1.5 Vitamins

Vitamins are very important components, in small amounts, for the development and

the growth of the organism. The inadequate amount of vitamins in the organism can

lead to many diseases. Vitamin E and Vitamin A are the ones that are present in a small

portion of olive oil and they help for the more beneficial effect of olive oil.

5.1.6 Tocopherols

Tocopherols, which are very well known as Vitamin E, are components with high

antioxidant action and their concentration in olive oil is ranged between 60-370

mg/kg. The content of tocopherols in Greek olive oil is ranged between 90-370 mg/kg.

Mainly in the olive oil, α-tocopherol contains 90% of all the components, β-tocopherol

is 10 mg/kg, γ-tocopherol is 20mg/kg and δ-tocopherol is 10mg/kg. (37)

5.1.7 Sterols

Sterols are cyclic alcohols with high molecular weight. They exist in all-natural fat

forms; they are soluble in lipids and oils and insoluble in water. Sterols are important

fatty acids that have to deal with the quality of olive oil. The types of sterols which

exist in the olive oil are four:

34

1) Common sterols

2) 4α-methylsterols

3) Triterpenoid alcohols

4) Triterpenoid dialcohols

The concentration of sterols in olive oil is between 1000-2000 mg/kg. A specific type of

sterol, β- sitosterol contains 75-90% of the total ratio. Specific sterol such as 5-

avenasterol is one of the main reasons, olive oil keeps its antioxidative actions and can

resist to high temperatures. Percentage of 10-40% of sterols exists in olive oil in the
72

form of esters with fatty acids.

5.1.8 Triterpenic acids

These acids are related to the oxidative stability of olive oil. They exist in high amounts

in leaves rather than olive fruit. Hydroxypentacyclic tritepenic acids are important

components of olive and they are biologically active. The acids, oleanolic and maslinic

are the major triterpenic acids that exist in the olive oil, as they are components of the

cortex of the olive, which are transferred to the final product. (38)

5.1.9 Phenols

In the last decades, it has been imputed from the researchers the meaning of the

presence of natural antioxidants in natural products. This is due to the correlation of

some chronic diseases, the action of reactive oxygen and nitrogen species, oxidative

stress and the attenuation of free radicals. The meaning reactive oxygen species is

used generally for the next chemical compounds, radicals and ions such as superoxide

radical (O2-), oxygen peroxide (H2O2), hydroxylic radical (-OH), alkoxy radicals (RO-)

and the radical (NO-). These drastic forms offend biomolecules and affect mechanisms

that are involved in many diseases or mechanisms which are important for gene

expression. The involution of these mechanisms in the pathology of many diseases and

in the protection against infections became the reason that antioxidants and

nutritional products have a beneficial effect on humans’ health. The density in polar

phenols has been recommended to be used as criteria for the quality of the olive oil.

The most common method of determination of total polar phenols is the

chromatographic method.

According to the variety of the olive, the climate conditions and many other factors,

the concentration of polar phenols is ranged between 50-1000 mg/kg of olive oil, but

the majority of olive oils contain 100-300 mg/kg. When the concentration is more than

300 mg/kg, olive oil presents a bitter taste. Nonetheless, high content in polar phenols

seems to act as a benefit for the duration of life of olive oil, due to the relationship that

exists between the content and the stability in total polar phenols. (39)

35

Phenolic compounds of Virgin olive oil are classified into the next categories:

1) tyrosol, hydroxytyrosol and their derivatives,

2) 4 hydrocybenzoic acids, 4 hydroxyphenolic acid, 4 hydroxycinnamic acid and their

derivatives.

3) Ligands and Flavonoids

73

From the different phenolic compounds which have been examined for their usage on

the stability of olive oil, hydroxytyrosol and caffeine are the ones which seem to be the

most potent. (40)

5.1.10 Antioxidants

Many of the substances which have been discovered, either are coming as natural

substances, either are added during the production of food, and act as antioxidants,

preventing oxidation. Vitamins C and E and β-carotene are playing a crucial role in the

protection of the humans’ body. There are two categories of antioxidants: The first one

is natural and the other one is synthetic.

5.1.11 Synthetic antioxidants

Synthetic antioxidants are phenyl compounds that are used as additives of fats, in

order to prevent or slow down their oxidation. The highest concentration of

antioxidants in food is 0.02%. Synthetic antioxidants are more active than natural and

in this category, several types exist such as:

▪ Butylated hydroxyanisole (BHA)

▪ Butylated hydroxytoluene (BHT)

▪ Butylated hydroxyquinone (BHQ)

There can be a combination of these types for a better and faster result. For example,

one combination can be BHA and BHT.

5.1.12 Natural antioxidants

The natural ones are compounds of nutritional origin with antioxidant activity. The

result of natural antioxidants depends on the origin of the plants which are coming

from and from the way of the delivery. (41)

5.1.13 Flavonoids

36

Flavonoids are a group of natural benzo-gamma-pyrrone factors, which are responsible

for the colour of olive fruit and flower, they are very important for the plants and they

constitute an important part of the diet and they are also known as aglycones.

Flavonoids are characterized as high-level antioxidants due to their ability to

deactivate free radicals, and they also cause attenuation of oxygen in a simple

situation.

5.1.14 Carotenoids

Natural carotenoids are colouring substances that exist in various plants, such as fruits,

vegetables and olive oil. They are very important for our health and they are related to
74

the protection from cancer.

Some other antioxidants which were also mentioned before are phenols and

tocopherols. (42)

Role of antioxidants

The role of antioxidants is also really important and it should be mentioned because

they:

❖ Protect the cell membranes, by destroying free radicals of oxygen

❖ Protect the cardiovascular system: They increase the resistance of vessels, they

decline the inflammatory factors, they prevent the oxidation of LDL cholesterol,

and they play a crucial role in the control of the level of blood pressure.

❖ Contain anticancer substances: They block the adhesion of dangerous enzymes

in tissues and they inactivate substances that are causing damage to healthy

tissues.

❖ Keep the skin elastic and prevent it from early aging, restricting the disruption

of collagen.

❖ Act as ant allergens

❖ Strengthen the vision (43)

6. CHAPTER

6.1 Effect of olive oil on health

37

6.1.1 Cardiovascular Diseases

Oxidative stress takes place when the levels of oxidant arrangement and antioxidants

are not equal. The result of this process is the disability of the operation of the

endothelium, and there is a rise in the volume of LDL cholesterol, which concludes in a

high risk of atherosclerosis. One of the biggest sources of passing in Nation is definitely

the presence of diseases in the cardiac system. Some examples of these diseases are

arterial hypertension, dyslipidemia or a type of inflammation in the body. (44)

The epithelial tissue damage creates modifications in vessels and it does not allow the

smooth muscle to grow properly. The consequence of this phenomenon is

atherosclerosis, which is also a risk factor for other very serious disorders such as

coronary infarction and stroke. It has been reported that many people have added

EVOO in their everyday diet because it is a sign of obstruction of cardiovascular

problems. Studies have shown their concentration on the decline of inflammatory

cytokines such as IL-6 and LDL-cholesterol, and the rise of HDL-cholesterol. The
75

abnormal function of the endothelium is a consequence of the agglomeration of

bonding particles, which leads to vascular disabilities. In addition, the utilization of

EVOO can be really helpful for the secondary superstition in patients, since it wields

power on LDL cholesterol, and expands the concentration of HDL cholesterol, which

moderates the development of the disease, and decreases any risk for the appearance

of some cardiac diseases. (45)

One of the first steps for the evolution of atherosclerosis is the decomposition of LDL

cholesterol, and when oxidation appears it is not identified by the unique receptor that

belongs to, directing to vascular abnormalities. Some phenolic substances which are

included in EVOO, behave as defensive means in opposition to the corruption of LDL-

cholesterol, and the oleic fatty acids have the proportion to decline this process by

keeping the cholesterol discharge and so reducing LDL-c. (46)

Oleic acid is also a very important compound that needs to be used for the protection

of humans’ health. The outcome of oleic acid on the cardiac tissue was remarked in

species accepting a high-cholesterol diet and a lipid fountain, exhibiting that the oleic

acid meliorated a topical inflammation. This obviously happens because oleic acid

declines the concentration of Triglycerides and LDL-c and gives a rise to HDL-c and

protects it from oxidation. (47)

The utilization of EVOO also has an action on blood pressure, as both DBP and SBP

were reduced. A decline in blood pressure was also very visible in obese people using

25 ml/day of EVOO for almost a year. (48)

For this particular disease, there was an experiment that happened to patients who

have been suffered from a heart attack. Almost 600 people were split into the

38

following categories: The first class goes after a meal rich in linoleic and oleic acid and

the second group followed a normal diet. After four years, even if there was not such a

huge difference in plasma lipids and lipoproteins, it was noticed that in the group who

followed the Mediterranean diet, there was a reduction in mortality of more than 70%.

These really important findings show that a resulting strategy for the reduction of

cardiac deaths needs to be supported by a very careful and protective diet. (49)

Recently the relationship between the food chain was evaluated. Researchers came

across a powerful, positive correlation between coronary heart disease and the

consumption of butter, milk, meat and sweets, alongside the utilization of vegetables

and olive oil has a negative result on the occasion of Coronary heart disease. (50)
76

6.1.2 Metabolic Syndrome

Previous researches have appeared that the depletion of elevated phenol of virgin

olive oil causes reduction of pre-inflammatory, pre-antioxidant and hyper

anticoagulant indicators in comparison with low-density phenols of virgin olive oil, but

it remains unclear if the biological results of the fraction of phenol are the ones which

influence the transfer level in vivo.

39
amount of olive oil is not related to an increase in weight in patients suffering from
metabolic syndrome, but acts oppositely way, it reduces the body mass. (54)
6.1.4 Maintenance of weight
It has been supported that a diet rich in fats, leads to obesity. This argument is
advocated by evidence that refers to the balance of fat in the organism, unlike proteins
and carbohydrates, it is not finished by strict metabolic control.
The goal of the study is the comparison of acidity indicators of fats postprandial, after
a rich in fats breakfast meal which is rich either in monosaturated fatty acid from extra
virgin olive oil or saturated fats by the crème.
In five hours after breakfast with olive oil, there was a significantly bigger percentage
of acidic fats from content in carbohydrates postprandial, after breakfast with
saturated fats. The thermal effect of a meal was much higher for people with a high
body mass index (BMI). (55)
6.1.5 Hyperglycemia
A diet that contains olive polyphenols seems to have a beneficial effect on
hyperglycemia. Many trials have been experimented on animals at first and after on
humans. For instance, it was observed that a complementary diet with olive oil for four
months was related to a 20% amelioration in insulin vulnerability in patients over 40.
Last but not least, the small substances of olive oil, also have a crucial part in the
reduction of hyperglycemia and many trials have been used to prove it. Substances
such as luteonin and apigenin have been used on the diet of patients for an amount of
500mg/day, and the results have been really good for those people, as they decline the
sugar blood levels. (56)
6.1.6 Stress
Many types of research defend that a big amount of employees suffer from stress. It is
known by many scientists that a physiological plane of stress is natural and beneficial
for the organism, but if this condition is taking for years it can act as a disadvantage for
the organism of the person. Some consequences of stress include insomnia, loss of
appetite or the opposite and fatigue and on important occasions, it can guide to
depression. The outcome of stress leads to some alterations in the figure of the person
40
and this has to do with the degree of cortisol, a very important hormone, in our body.
A raise in the excretion of cortisol can lead to tachycardia and high blood pressure. (57)
One way to calculate the level of stress in the body is the functional near-infrared
spectroscopy (fNIRS). Also, some studies indicate that fat recess may reduce the
production of cortisol. After many thoughts and researches, scientists came to a result
that olive oils may be an advantage in this problem, because they belong to lipid
groups. For a better result, an examination took place on some students. Scientists
wanted to decompose the outcome of oils on stress, so they analyze the elements
from the experiments by calculating the blood flow in the brain and hemoglobin and
77

oxyhemoglobin were the basic elements they measured. The absorption of these two
is on the rise because of neuronal operation. The outcome of this is that olive oil
utilization reduces the quantity of stress in the organism. (58)
6.1.7 Diabetes Mellitus
Diabetes mellitus is a disorder that is characterized by the lifted level of glycemia. The
main goal for the cure for diabetic patients is the control of the metabolism, in
essence, the achievement of the suitable body weight and this is feasible with the
adjustment of a suitable diet. The causality of β-cell operating volume is a needful
process for the evaluation of DM2. That means that there should be a reinstatement of
β-cells to cure this disease. (59)
This particular nutritional composition of the Mediterranean diet has as a result on the
one hand low density of saturated fats and cholesterol, and on the other hand,
increased solidity in complex carbohydrates and fibers. The everyday use of olive oil in
various quantities implies a high density of the diet in mono-saturated fatty acids,
which is between 15-20% of the calories, as a result, the total amount of lipids to be
between 25-35% of the whole calories. (60)
The composition of this plan satisfies all the conditions of an adequate diet for diabetic
people. The proportion of mono-saturated fatty acids and carbohydrates can be
different, according to the metabolic profile and the personal preference, with the
differentiation in the quantity of olive oil which is consumed in everyday life. If weight
loss and the limitation of calories are necessary, the amount of olive oil can be
reduced. (61)
Additionally, some researchers have shown that a diet full of rich olive oil can reduce
the risk of atherosclerosis in diabetic patients. In their work scientists investigated the
effect of nutritional polyunsaturated fatty acids at postprandial lipoprotein, as at other
41
factors which influence glucose and the metabolism of lipids, in comparison with the
outcome of the monounsaturated lipids.
The scientists concluded that a diet full of elaic acid is much better than a diet full of
linoleic acid for the patients suffering from diabetes mellitus 2 and that may decline
the risk of atherosclerosis, due to reduction of insulin and glucose level in the blood
during fasting as the level of low-density lipoproteins and most importantly
postprandial lipoproteins. (62)
6.1.8 Rheumatoid Arthritis
People who agonize from rheumatoid arthritis when they follow the traditional diet,
whose one of the basic characteristics is olive oil, present remarkable development
regarding the signs of the disease. Swedish doctors concerned for many years with the
relationship between diet and rheumatoid arthritis. The doctors submitted 25 patients
with rheumatoid arthritis on the Mediterranean diet and the other 25 patients to the
ordinary diet. The final examinations have shown that the people who followed the
diet rich in fish, olive oil, vegetables and fruits presented an important reduction of the
symptoms of the disease and improvement of functionality and vitality. (63)
6.1.9 Olive oil and peptic system
Olive oil and stomach
Charbon-Niers research, concerning the gastric behaviour of olive oil, discovered the
existence of some nutritive hierarchy between viable fats and oils, which has a
connection with their configuration in fatty acids. Olive oil becomes acceptable from
the organism, whose most significant fatty acid is the monosaturated elaic acid.
Some researchers have signalized the beneficial influence of olive oil in perchloride
gastritis and gastro-duodenal ulcer. It was noticed that the addition of olive oil in the
78

trial meal of porridge was causing a reduction of secretion of gastric acid, in

comparison with the secretion of the pure porridge. Plenty of researchers have
imputed this affection on the liberation of enterogastrin, but finally, it was declined as
theory. (64)
One of the basic features of olive oil is its protective activity. A scientist reported a rise
of alkaline secretions of cells of gastric mucosa in a pyloric cave and in the stomach
after the granting of olive oil. It was figured out that during the therapy of 102 patients
who suffered from an ulcer, which the replacement of viable fats with olive oil
challenged the reduction of alterations in 33% of the cases and the healing of the ulcer
in 55% of the cases.
42
Definitely, the composition for the reception of the olive oil in the cases of perchloride
gastritis or gastroduodenal ulcer does not replace its therapeutic confrontation.
Additionally, in recent research on normal people who had as a unique source of fat in
their diet olive oil, it was not mentioned the basic reduction of basic secretion of
gastric acid or the acid which is caused by the pentagastrin. (65)
6.1.10 Olive oil and intestine
One or two spoons of soup of olive oil every morning before breakfast seem to have a
significant influence on acute and chronic constipation. It is believed that the
mechanism which is related to the production of cholecystokinin has double
properties; on one hand, it causes contraction of the gallbladder and on the other
hand, it activates the peristaltic movement of the small intestine. (66)
6.1.11 Cancer
People who receive a big amount of olive oil in their diet are in a very good position
about preventing their body cells from damage and this damage can lead to cancer.
There is evidence that olive oil can diminish oxidative destruction in the inherited
information of cells which can lead to the evolution of cancer. (67)
6.1.12 Breast cancer
Technologists have uncovered that oleic acid, the major ingredient of olive oil, reduces
the activity of oncogene HER-2/eu which exists in 30% of patients with cancer. In
compliance with a recent study in some laboratory tests which happened in cancer
cells of breast, it was reported that oleic acid was decreasing gradually the level of a
gene which was believed that it was promoted this particular type of cancer. There
was a limitation in the action of HER-2/eu, but also improved the effectiveness of
anticancer active substance trastuzumab, which contribute to the improvement of the
situation of many women suffering from the disease. (68)
This discovery may imply that oleic acid can delay or prevent the resistance in
transtuzumab which many patients develop due to high levels of HER-2/eu.
Simultaneously, oleic acid upgrades the activity of the drug Herceptin which reduces
the action of the oncogene. The mechanism of action between the oncogene and the
drug works in a completely different way, but there is good cooperation between
them. (69)
43
6.1.13 Small intestine
The cancer of the small intestine is a very frequent type of cancer and olive oil seems
to play a significant role in its improvement. Many views believe that olive oil
influences the secretion of bile salt in rats, it has been formulated the case that it can
influence the frequency of manifestation of the disease.
A study was conducted on a population from 28 countries all over the world.
Nutritional factors were analyzed at a national level and there was a comparison with
79

the differences of cancer for each country. With a simple analysis, a correlation was
related between the high incidences of cancer and the low level of vegetables and
cereals consumption. Previous works of scientists have shown that the consequence of
cancer is less intense in Mediterranean countries due to the low consumption of meat.
Practically it can influence the secretion of cholic acid in the small intestine, and in
turn, it may affect the metabolism of polyamines in the cells of the small intestine, so
that the reduction from the physiological mucosa to cancer will appear. (70)
6.1.14 Prostate
A research team carried out work about the effect of diet on prostate cancer. Age,
total recruitment of calories, fats and diets were the key to the solution and the
prevention for this type of cancer. It was observed that diet take part in a dominant
role in the possibilities of manifestation of prostate cancer. The danger of prostate
cancer was reduced as the consumption of carotenoids and leucopein was increased.
In conclusion, vegetables, fruits and olive oil which include carotenoids are related to
the prevention of prostate cancer in men. (71)
6.1.15 Esophagus
Various studies have taken into account the age, smoking regularity and alcohol
consumption of many people and it was observed that there is a high risk of cancer in
esophagus to people who were receiving big amounts of soups, meanwhile, the
opposite was for the ones who included rice, raw vegetables and fruits. Olive oil was
also contributed to the reduction of this danger; meanwhile, the butter was enhancing
the risk. Last but not least there is much evidence that raw vegetables and olive oil
contribute to a very healthy life as they limit the possibilities of esophagus cancer to
the population. (72)
44
7. CHAPTER
7.1 Cosmetology
In ancient times, olive oil was subsisted as a source of health and beauty. It has
nutritious and detergent properties and offers hydration, deep peeling, smoothing and
shining. Ω3 and Ω6 fatty acids which are incorporated in olive oil form a thin protective
layer that holds the internal moisture of the epidermis stable. Olive oil is also rich in
antioxidant sources such as Vitamin E and polyphenols, which act against premature
ageing, helping the formation of collagen and elastin.
Olive oil whenever it is alone or in combination with other substances gives epidermis
all the important sources to be healthy. In cosmetology, olive oil is combined with
other substances such as avocado, jojoba, eggs, honey, yogurt, salt and sugar for the
formation of cosmetics and soaps which are helpful for the skin. It can also be used in
aromatherapy as a base.
Cosmetics, which include olive oil protect the cells, from oxidation and give strength to
the skin. Also, antioxidants, stimulate the cell membrane to have regeneration of the
skin.
Cosmetics are all the substances that can be used in human skin, teeth and mouth
mucosa to improve hygiene. Cosmetology can also be used in hair and the human face
to improve it but also to protect it from hostile factors. (73)
7.1.1 Combination of olive oil with substances
• Honey: It is a natural sweet substance that is produced by bees. It is a very
popular substance in many beauty products and that is because it is a source
full of vitamins and metals. It causes deep hydration in the epidermis and
affords excellent healing properties. It acts as an antioxidant and as a shield
against bacteria.
80

• Beeswax: It is produced by young working bees and is considered the second

most profitable product. It has healing, antiseptic, anti-inflammatory, soothing
45
and hydrating possessions. It is suitable for all types of skin and can be used in
many hair and lip products. It protects the skin from irritation, lets it breathe
and helps in the regeneration of cells in the membrane.
• Eggs: They are wealthy in proteins and fatty acids and constitute the best base
of face and hair masks. They are opulent in vitamins A, C and E, they rebirth the
skin and make the hair slicker. They contribute to hair loss.
• Milk: The protein milk offers to the skin many hydrating and therapeutic
properties. Their hydrating ability is a result of an excellent absorption from the
epidermis as of their ability to bind water.
• Avocado oil: It is a physical way for the hydration of skin without chemicals. It is
used in many beauty products such as creams, conditioners, masks and lotions.
Avocado is rich in products that sooth and hydrate skin plus it also contains
Vitamin E which helps for the restoration of dry hair.
• Bananas: They contain Vitamin B such as B6 and B12 which help the hair to be
smooth and they protect the elasticity of hair and set a barrier to scissors.
• Almond Oil: It contains many vitamins such as E and K. It is very fluid and does
not give the sense of oil when applied to the skin or hair and it is the best
lubricant and softener.
• NaOH: It is a crystalloid, white compound and very hygroscopic. This element
allows the transformation of oils in soap.
• KOH: It is used in the production of soft soaps as the first material. It should be
used with caution because it can irritate the eyes and skin. (74)
7.1.2 CREAMS
Figure 9: An example of cream after its preparation (9’)
https://www.webmd.com/beauty/skin-lightening-products
1) COLD CREAMS
46
From the second century, there are references for the production of creams and more
specifically cold creams. They are called this way due to the sense of cold they leave on
the epidermis since their spread, due to the evaporation of water they include. The
first cream that was produced was by bee wax, olive oil and rose water.
Below will be analyzed a very characteristic moisturizing cream with olive oil and bee
wax, which can be used in all types of skin and is very nutritious and hydrating. It is
suggested to be spread after the morning facial cleanse. Some of the identities of the
ingredients are:
• Olive oil due to its antioxidants compounds and vitamins can make the skin
more elastic, smooth and silky.
• Rose-water makes the skin more soft and fresh.
• Bee wax helps in the maintenance of the cream (even three months after its
opening in a cool place) and improves its action.
Production
A few days before the production of the cream, you posit 3-4 spoons of olive oil in an
implement and heat it with 2-3 spoons of rose water. Leave it like this for a few days.
After, melt the beeswax on a pan and when it is ready add the previous mixture. After
some time you take it off the heat and add a small amount of water to get the right
texture. After it gets cold you place it on containers and keep it in a cool place.
2) NUTRITION CREAMS
81

These creams with the help of their compounds can react to the physiological
reparation mechanisms and improve the causes that challenged the dryness of the
skin. In addition, olive oil plays a crucial role in that procedure, as it is suitable for the
care of dry skin. It causes deep hydration and is making it more smooth and shiny.
Below a very easy recipe will be analyzed by using aloe vera and olive oil and it can be
produced at home. It has a very nice and soft aroma and is very nutritious for the skin.
It is suggested to be used in the morning in night before bed.
Production
Mix the olive oil with the aloe gel for a few minutes until the mixture will be
homogenized and then add it to your skin. (75)
3) EMOLLIENT CREAMS
These types of creams make the skin softer. When olive oil is added, not only it makes
the skin softer but also gives the suitable nutrients it deserves.
Also, these types of creams can have very good results in dry and dull hair and it helps

them find their strength and smoothness. A very easy recipe which can be made by the

home is the next one:

Production

You place in a bowl two spoons of olive oil, one-yolk and one spoon of vinegar and mix

it until it becomes a homogenous mixture. Additionally, you can place it on your hair

and leave it there for thirty minutes. After that, you can wash it away.

4) MOISTURIZING CREAMS

Moisturizing of the epidermis is very crucial, so the creams that are produced are the

most important. Water is the ingredient that makes the skin smooth and shiny, so its

absence can make the skin rough and very thin.

There are three ways in which we can make our skin work.

▪ Humidification, meaning that a suitable amount of water should be controlled

to complete its content on the skin.

▪ With the restoration of compounds that are missing, making an effort to find if

there are any mistakes by which the skin gets dry.

▪ With inclination, this is the most important because in that case olive oil is

used. More importantly, with this method, the loss of water that is eliminated

from the skin is declined. For this method, olive oil is the best solution because

it is an ingredient that cannot be penetrated by the water.

5) MASSAGE CREAMS

The first massage creams are traditionally the cold creams and most importantly the

ones that are water in oil (w/o). These creams have a high content in the oily phase

and for that reason, they cannot be absorbed as fast as the others. These creams are

the best for massage and in combination with good movements the blood flows
82

throughout the body very easily and the dead cells of the skin come on the surface.

Cosmetic industries have huge success in terms of marketing with these creams and

they can be used for various reasons such as topical fat. (76)

6) HAND CREMES

Hand creams are very important because they tend to make a thin layer on the tip of

our hands to make microbes and other pollutants stick and not pass through the skin.

They also have smoothing and emollient properties as they make the skin more elastic

48

and soft and they prevent them from making the hands dry due to their constant

exhibition in the sun or cold. The addition of olive oil helps.

7) MASKS

Except for the creams olive oil can be used as an ingredient in masks whether they are

for hair, skin or face. They tend to give the skin revitalization as they also make it clean,

due to the fact that they can bring out the dead cells and clean the skin. Also, some of

them can fight against age and decline the freckles for a small period.

The use of olive oil gives them the ability to be nutritious for the skin and generally for

the areas that need to be used. The antioxidant compounds that are used are

incorporated to the enhancement of their abilities to make them more suitable.

Procedure

Melt 35gr of glycerin and then slowly add 25gr of honey and mix it immediately in

order not to get stuck. Then add 10gr of olive oil and 35gr of rose water, mix the

mixture until the mixture gets cold. (77)

Figure 10: Homemade face mask for dry skin (10’) https://www.dreamstime.com/mask-face-

natural-ingredients-care-dry-skin-vector-facial-mask-recipe-homemade-mask-dry-skin-vector-

image123453516

7.1.3 EMULSIONS

The emulsion is characterized as a system that consists of two liquid mixed phases.

There are two types of emulsions, oil in water (o/w) and water in oil (w/o). Oil in water

emulsion has water as an external phase and water in oil emulsion has oil as an

49

external phase. They contain oils because they help with the improvement of the

epidermis.

Using olive oil in emulsions we can see a significant improvement on the skin as it is

smoother and softer. Because olive oil cannot be dissolved in water, it is really
83

important to shake the emulsion before each use, to be homogenized. When the

emulsion is homogenized, the olive oil acts as a smoothing and external lubricant and

its goal is to make the skin hydrated and soft. Moreover, olive oil is a compound that

can be used as a sun protector due to its components.

Olive oil can be combined with many fluids such as juices from fruits, water, aloe, milk

and many more. Many women choose to make their emulsions at home, by mixing

olive oil and milk and then they apply it on their face or neck. (78)

7.1.4 FACE AND BODY SCRUBS

Olive oil is also used in peeling products of the face and body. Olive oil is the ingredient

that exists in these products, as it is full of vitamins, acts against wrinkles and keeps

the temperature of the skin stable. Usually face and body scrubs are used in dry skin to

peel and hydrate them Olive oil succeeds in deep hydration of the skin and in

combination with other compounds it removes the dead cells out of the organism.

Production

The first step is to mix olive oil with salt in a bowl. When these two ingredients are

united, then add a full lemon and mix it all very well. (79)

50

Figure 11: Preparation of scrubs (11’) https://simple-

veganista.com/diy-edible-sugar-scrub/

7.1.5 SUNSCREENS

Sunscreen products play a crucial role in the everyday routine of many people. It is

well known that the sun has many benefits but also harmful properties on the skin, so

the skin needs to be protected by the sun radiation. Sunscreens exist in various forms

such as emulsions, creams, lotions etc.

Olive oil can absorb sun rays, so that is the reason that it is used in the production of

most the sunscreens. Specifically, it is ideal for the production of a tan lotion, which

helps the skin to tan but also gives suitable protection due to its antioxidant

properties. Sunscreens with olive oil contain natural filters of protection from sun rays,

which do not absorb but reflect the radiation in contrast with the synthetic filter which

absorbs the sun rays in the skin.

Production

You heat 50ml of coconut oil and after you mix it with 50 ml of olive oil and 100ml of

sesame oil. After that, you pour two sachets of black tea and then mix it. The final step

is to put it in a jar and then use it at least 20 minutes before the exhibition to the sun.
84

(80)

SOAPS

Soap has a significant meaning in everyday life and is one of the first aid products. It is

produced by a chemical reaction between alkali compounds and fatty acids in fat or oil.

Chemical soaps which are used in everyday life, sometimes cause allergies on the skin

compared to the physical ones which are produced by good quality olive oil. Good

quality soap does not contain a big amount of alkali compounds.

Production

o The simplest composition of soap is called clear soap and consists of olive oil

and soda. It is considered one of the best cleansers of the skin and it does not

irritate.

o Green soap is composed of KOH, olive oil and glycerin. It can be used as a

pharmaceutical liquid soap which is used in oily skin and causes hydration and

it exists in sterile composition for the hair. (81)

Figure 13: Final product of soap (13’) https://practicalselfreliance.com/how-to-make-

soap/

52

8. CHAPTER QUESTIONNAIRE PART

Consumers

Meaning of consumers

The satisfaction of consumers is not a new business definition, but many companies

fail to approach the expectations of their clients. The most important chapter of a

company is its clients and if they are not pleased, they obvert to another product. The

good quality of the product and service is not enough for the existence of the

company, but it should maintain a good relationship with its clients to understand

them and to keep improving its work.

For the study of condiments of olive oil consumers, a questionnaire was created and it

contains 12 questions. Twenty questionnaires were sent and filled to 50 consumers.

Next are the results from the research.

8.1 Questionnaire for the consumption of olive oil

Theme: The satisfaction of olive oil consumers

Please fill in the next questionnaire, and the data of this research will be used to obtain

a complete image for the satisfaction of olive oil consumers in Greece. To be

mentioned the next questionnaire is anonymous.

85

Thank you for your cooperate

Conclusion
Olive oil is praised no end for its wide-ranging health boons. From heart health, to controlled blood
sugar, to cell function, this oil does a lot of good. Yet, you must use it wisely. Pick top-notch,
preferably extra virgin olive oil for peak gains. While it’s a beloved part of our diets, remember no
food is a cure-all. Olive oil, like all foods, should be part of a balanced diet.

The present study provides a first level of evidence on the in vivo cardiovascular benefits of olive oil
triterpenes (oleanolic and maslinic acids) in healthy humans, decreasing plasma inflammatory
biomarkers and DNA oxidation after three weeks of supplementation. Olive oil is a good food matrix
to be a vehicle for triterpenes, as they naturally contain these compounds but in low amounts.
Experimental studies on olive oil triterpenic acids, from the unsaponifiable fraction, showed a
number of benefits against CVD and their complications [12,13,14,29,30]. The fatty acid profiles
after the three interventions confirm the adherence of the subjects to the interventions since an
increase of plasma oleic acid was observed after each of them, reflecting the lipid composition of the
diet [31]. The present study shows that a daily supplementation during three weeks with 30 mL of a
FOO providing 13.4 mg/d of phenolic compounds (487 ppm) and 4.7 mg/d of oleanolic acid and 6
mg/d of maslinic acid (171 and 218 ppm, respectively) may protect DNA against oxidative stress in
vivo, by decreasing urinary 8-OHdG concentrations compared with the same olive oil but without
triterpenes. 8-OHdG is an oxidative stress biomarker produced by oxidation of the nucleoside
deoxyguanosine and is subsequently excreted directly into the urine. It was described as a sensitive
biomarker of oxidative DNA damage [32]. A recent meta-analysis reported that 8-OHdG levels are
higher in patients with CVD than in healthy controls [33]. In this regard, on one hand, in vitro
[34,35] and animal [36] studies reported that olive oil phenolic compounds protect DNA from
oxidation. On the other hand, in vitro studies reported that pentacyclic triterpenes are also able to
protect against oxidative DNA damage [37]. The 8-OHdG results of the present study is in
agreement with observations made in the EUROLIVE study, in which supplementation with olive oil
protects DNA against oxidation after three weeks of supplementation in healthy males [38]. This
effect was independent of the phenolic content and related to the monounsaturated fatty acid content
of the olive oil [38]. In addition, a protection of DNA oxidation in hyperlipidemic subjects was
observed after the intake of a VOO complemented with thyme phenolic compounds [39], and after
eight weeks of supplementation with 50 g/d of an olive oil with 592 ppm of phenols in
postmenopausal women [40]. Furthermore, 11.13 mg of olive’s phenolic compounds have
demonstrated postprandial protection against in vivo DNA oxidation in healthy males after four days
of supplementation [41]. All these positive data differ from the null effect of our VOO and OVOO,
probably due to the lower amount of phenols contained in the oils used in the present study. In
addition, physiological and environmental situations of volunteers may also influence the results.
Therefore, all these facts may indicate that a longer intervention with the OVOO may have more
impact on healthy adults and also on different pathological situations. More studies are needed in
order to validate the positive effect of triterpenic acids-rich olive oils on DNA oxidation.

F2-isoprostanes are generated in vivo by free radical-induced peroxidation of arachidonic acid [42].
One of the major F2-isoprostanes, 15-F2t-isoprostane increase in several situations associated with
oxidative stress, including atherosclerosis, diabetes, obesity, cigarette smoking, neurodegenerative
diseases, and asthma [43]. The present study shows that there was no effect on urinary 15-F2t-
isoprostane levels after supplementation with VOOs enriched in phenolic and triterpenic compounds
in healthy humans. The effect of olive oil phenolic compounds is not fully elucidated. In agreement
with our results, a null effect on urinary isoprostanes levels of olive oil phenolic compounds were
86

reported in dyslipidemic subjects [44]. Also, a previous randomized crossover study with a larger
sample size reported that lipid peroxidation decreased in a linear manner with the phenolic content of
the olive oil, particularly in markers directly associated with LDL oxidation, but not F2-isoprostanes,
in healthy subjects [45]. In vitro [46] and in vivo [9,47] experimental studies demonstrated an
antioxidant effect of triterpenic acids. Particularly, oleanolic acid showed a substantial protection
against in vitro lipid peroxidation in isolated rat liver microsomes [48], as well as in a cerebral
ischemic animal model [49]. It was proposed that oleanolic acid is not only a free radical-scavenger
acting through direct chemical reactions, but also a biological molecule which may enhance the
antioxidant defense [49]. In addition, maslinic acid is a potent inhibitor of oxidative stress and
cytokine production in stimulated murine macrophages in vitro [16]. In fact, the negative correlation
found between oleic and urinary 15-F2t-isoprostane may suggest that a longer intervention would be
able to exert a positive effect on reducing urinary 15-F2t-isoprostane concentrations. This is in
agreement with other published preclinical [50] and clinical [51] studies.

When analyzing plasma non-enzymatic antioxidant defense system, at the beginning of the study, the
concentrations of CoQ9, CoQ10, and carotenes in plasma were different between the three sequences
of olive oil consumption, while no further changes were observed after the all interventions.
According to the nutritional data registered during the study, our volunteers did not modify the diet.
Therefore, a change in plasma antioxidant molecules was not expected. Indeed, the absence of these
changes suggests that the improved oxidative status observed after the three olive oil interventions
could mainly depend on other macro- or micro- components present in the olive oils.

The present study shows that daily supplementation during three weeks with 30 mL of a FOO rich in
phenolic compounds and enriched with olive triterpenic acids improves plasma inflammatory
biomarkers, by decreasing plasma IL-8 and TNF-α concentrations, compared with the oils with less
triterpenic acids. This is the first time than an in vivo effect of triterpenic acids from olive oil is
described in humans. The OVOO improved the IL-8 concentration, but did not modify the TNF-α.
The Mediterranean diet is associated with a lower incidence of chronic inflammatory diseases [6]. A
recent review of human studies indicated that olive oil consumption reduces TNF-α and IL-6 levels
when associated with Mediterranean diet and lifestyle [52]. A systematic review concluded that olive
oil might exert beneficial effects on inflammatory biomarkers, such as C reactive protein and IL-6,
after at least four weeks of intervention in individuals adhering to the Mediterranean diet [5]. In
addition, long-term adherence to this Mediterranean diet decreased plasma concentrations of IL-6
and IL-8 related to different steps of atheroma plaque development in elderly subjects at high
cardiovascular risk [53]. The major anti-inflammatory components in olives are tyrosol,
hydroxytyrosol, oleuropein, ligstroside, verbascoside, and their derivatives [54]. In particular,
hydroxytyrosol is considered the major anti-inflammatory compound in aqueous olive extracts [55].
In agreement with our results, human studies observed a decrease of IL-8 concentrations after the
intake of 136.2 mg of oleuropein and 6.4 mg of hydroxytyrosol per day during an intervention of six
weeks in pre-hypertension males [56]. In addition, in vivo and in vitro studies suggest that
pentacyclic triterpenes exhibit anti-inflammatory effects [57,58].

We chose three weeks of intervention based on a previous study EUROLIVE, which reported the
beneficial effects on HDLc and oxidized LDL plasma in a dose-dependent manner of the phenolic
content in healthy volunteers from the North of Spain [38]. One of the limitations of the present
study is that young and healthy subjects were recruited, and, in addition, our subjects lived in the
South of Spain, where Mediterranean diet and VOO are highly consumed. Therefore, a three-week
intervention may not be long enough to cause significant changes. Further long-term studies,
particularly in patients with CVD and related comorbidities, are required to find conclusions related
87

to the bioactive compounds presents in olive oils, and to explore the mechanisms involved in these
effects of specific components of VOO on oxidative stress.

The excessive formation of reactive oxygen species (ROS) causes extensive damage to cellular
macromolecules, such as polyunsaturated lipids, proteins, and DNA. Oxidative stress processes are
associated with cardiovascular diseases (CVD) and other noncommunicable chronic diseases, such as
obesity, type-II diabetes, and cancer [1]. CVD represent the main cause of disability and death in
industrialized countries. Atherosclerosis, the major physiopathological cause of CVD, is a
multifactorial disease involving inflammatory, oxidative, proliferative, and necrotic processes into
the arterial walls. This chronic inflammatory condition activates immune competent cells in lesions
producing mainly proinflammatory cytokines [2]. The Mediterranean diet is recognized as a healthy
food pattern [3]. Olive oil, the main source of fat of the Mediterranean diet, is shown to have
antioxidant properties [4] and is associated with a lower incidence of chronic inflammatory diseases
such as CVD [4,5,6] due to the antioxidant properties of its minor compounds [7].

Besides the high content of monounsaturated oleic acid, other minor bioactive components are
described in virgin olive oil (VOO) [8]. Components from the polar fraction include mainly phenolic
compounds, while triterpenic acids stand out in the unsaponifiable one, among others. The presence
of these molecules depends on the olive ripeness, the cultivar type, the climate of the area of
cultivation, and the type of oil extraction processes [9]. Oleuropein and hydroxytyrosol are the most
important active phenols present in VOO [10]. In 2011, the European Food Safety Authority (EFSA)
released a health claim concerning the benefits of the daily intake of VOOs, containing at least 5 mg
of hydroxytyrosol and related compounds (i.e., oleuropein and tyrosol) per 20 g of VOO, decreasing
low density lipoprotein cholesterol (LDLc) oxidation [11]. This implies that only VOOs with a high
phenolic content can claim the benefits. Therefore, the necessity of optimize the VOO processing in
order to obtain bioactive compound contents is one of the current goals in terms of increasing its
nutritional value. Among minor VOO compounds, triterpenic acids also have potential health
benefits against CVD and their complications [12,13,14]. The highest content of triterpenic acids is
found in the seeds and the skin of the olives, which are usually used for pomace olive oil production.
Among olive oil triterpenic acids, oleanolic and maslinic show anticancer [15], anti-inflammatory
[16,17], antioxidant [18] and cardioprotective [19] activities in cellular and animal models [20,21].
However, to the best of our knowledge, the NUTRAOLEUM study is the first human study focused
on providing evidence of the health benefits of olive oil triterpenes. Due to this, within the frame of
this study, we aimed to evaluate the effect of VOOs enriched with bioactive compounds, such as
phenolic compounds and triterpenes, on the oxidative and inflammatory biomarkers in healthy
adults.

In addition to being used as food, olives and olive oil are of major cultural importance in the Mediterranean
area. Hundreds of cultivars exist and produce assorted kinds of olive oils, the predominant source of visible fat
in the Mediterranean diets. Relatively recent research e mostly stimulated by the observation that the
Mediterranean diet is associated with longevity and wellbeing e is being unveiling the multiple beneficial
effects that olive oil, namely the extra virgin one and its peculiar components have on human health. Low-
(poly)phenol olive oils might also provide health benefits because they help lowering the proportion of
saturated fat in the diet. Indirect effects of olive oil consumption, e.g. increased consumption of raw
vegetables also play major roles on health, as do the social environment and the culinary civilization in which
olive oil is consumed. As outlined in this paper, accrued evidence is strongly suggestive of cardioprotective
activities of olive oil and data on chemoprevention of cancer and neuroprotection are accumulating (Fig. 1).
This might

(and should) foster dissemination of the olive oil culture innon-producing countries and e in turn e increase
world-wide demand for this product. Future research and appropriate trials will further elucidate the role of
88

olive oil on human health and will provide further scientific evidence to the notion that consuming this oil as
part of a balanced diet and healthful lifestyle improves prognosis and extends life- and health spans.