Clinical Nutrition ESPEN 60 (2024) 210e216

Contents lists available at ScienceDirect

Clinical Nutrition ESPEN

journal homepage: http://www.clinicalnutritionespen.com

Original article

Vitamin D deficiency during the coronavirus disease 2019 (COVID-19)

pandemic among healthcare workers
Aoi Ito a, 1, Shohei Yamamoto a, 1, Yusuke Oshiro b, Natsumi Inamura b, Takashi Nemoto b,
Tomofumi Tan b, Maki Konishi a, Tetsuya Mizoue a, *, Nobuyoshi Aoyanagi c,
Haruhito Sugiyama d, Wataru Sugiura e, Norio Ohmagari f
a
Department of Epidemiology and Prevention, Center for Clinical Sciences, National Center for Global Health and Medicine, Tokyo, Japan
b
Department of Laboratory Testing, Center Hospital of the National Center for the Global Health and Medicine, Tokyo, Japan
c
Kohnodai Hospital of the National Center for the Global Health and Medicine, Chiba, Japan
d
Center Hospital of the National Center for the Global Health and Medicine, Tokyo, Japan
e
Center for Clinical Sciences, National Center for Global Health and Medicine, Tokyo, Japan
f
Disease Control and Prevention Center, National Center for Global Health and Medicine, Tokyo, Japan

a r t i c l e i n f o s u m m a r y

Article history: Background & aims: Vitamin D deficiency is a common nutritional problem worldwide that may have
Received 5 October 2023 worsened during the coronavirus disease 2019 (COVID-19) pandemic. The present study sought to
Accepted 2 February 2024 examine the prevalence and correlates of vitamin D deficiency among healthcare workers three years
after the start of the COVID-19 pandemic.
Keywords: Methods: Participants comprised 2543 staff members from a medical research institute, who completed
Cross-sectional study
a questionnaire and donated blood samples in June 2023. 25-hydroxyvitamin D (25[OH]D) levels were
Healthcare workers
measured using an electrochemiluminescence immunoassay. Logistic regression was used to calculate
Vitamin D
COVID-19
the odds ratio and its 95% confidence interval while adjusting for covariates.
Results: The proportions of participants with vitamin D insufficiency (25[OH]D 20e29 ng/mL) and
deficiency (25[OH]D < 20 ng/mL) were 44.9% and 45.9%, respectively. In a multivariable-adjusted model,
factors associated with a higher prevalence of vitamin D deficiency included younger age, female sex,
fewer hours of daytime outdoor physical activity during leisure time (without regular use of sunscreen),
lower intake of fatty fish, no use of vitamin D supplements, smoking, and no alcohol consumption.
Occupational factors, including shift work, were not independently associated with vitamin D deficiency.
Conclusions: Our results suggest that vitamin D insufficiency and deficiency are highly prevalent among
healthcare workers. Health education regarding lifestyle modifications for this occupational group are
warranted to improve their vitamin D status in the COVID-19 era.
© 2024 The Author(s). Published by Elsevier Ltd on behalf of European Society for Clinical Nutrition and
Metabolism. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

1. Introduction Vitamin D receptors are present in most organs, and experimental

studies have shown that this hormone is involved in the pathogen-
Vitamin D is a pro-hormone that plays a vital role not only in the esis of the aforementioned diseases [1]. Vitamin D is produced in the
maintenance of skeletal health but also in the prevention of various skin upon ultraviolet (UV) exposure and is consumed through di-
diseases, including cancer, cardiovascular diseases, diabetes, degen- etary sources and supplements [1]. Factors associated with low
erative disorders, psychiatric disorders, and infectious diseases [1]. circulating vitamin D levels include female sex [2], older age [2], high

Abbreviations: BMI, body mass index; COVID-19, coronavirus disease 2019; CRP, C-reactive protein; HTPs, heated tobacco products; NCGM, National Center for Global
Health and Medicine; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; UV, ultraviolet; 25(OH)D, 25-hydroxyvitamin D.
* Corresponding author. Department of Epidemiology and Prevention, Center for Clinical Sciences, National Center for Global Health and Medicine, 1-21-1, Toyama,
Shinjuku-ku, Tokyo, 162-8655, Japan.
E-mail addresses: [email protected] (A. Ito), [email protected] (T. Mizoue).
1
Aoi Ito and Shohei Yamamoto contributed equally to the work as co-first authors.

https://doi.org/10.1016/j.clnesp.2024.02.005
2405-4577/© 2024 The Author(s). Published by Elsevier Ltd on behalf of European Society for Clinical Nutrition and Metabolism. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
A. Ito, S. Yamamoto, Y. Oshiro et al. Clinical Nutrition ESPEN 60 (2024) 210e216

body mass index (BMI) [3], night shift work [4], smoking [5], fewer were measured using an Elecsys Vitamin D Total III electro-
hours of outdoor activity [6,7], sunscreen use [2], and low dietary chemiluminescence immunoassay kit (Roche Diagnostics, Man-
intake of vitamin D or fish [2]. nheim, Germany) on a Cobas 8000 e602 immunoassay analyzer
Despite the important role of vitamin D for the prevention of (Roche Diagnostics) at the Central Laboratory of NCGM. Based on a
numerous diseases, vitamin D deficiency is common globally [8]. previous report [1], we defined 25(OH)D concentrations of 30 ng/
This could be ascribed primarily to the modern lifestyle with a mL, 20e29 ng/mL, and <20 ng/mL as vitamin D sufficiency, insuf-
decreased chance of sun exposure, wherein individuals tend to ficiency, and deficiency, respectively.
spend longer hours indoors and avoid sun exposure than in the past.
Among the Japanese population, despite a high consumption of fish 2.3. Covariates
and mushrooms (rich sources of vitamin D), over 80e90% of adults
have circulating 25-hydroxyvitamin D (25[OH]D) lower than 30 ng/ Information regarding potential determinants of circulating
mL [9e12], a cutoff proposed for the prevention of chronic diseases vitamin D levels was provided by the administrative department
[13]. During the coronavirus disease 2019 (COVID-19) pandemic, (sex, age, occupation, and affiliation) of the NCGM, or ascertained
people were advised to stay home, which might have worsened the via a questionnaire (others). In relation to sun exposure-induced
situation. In a Spanish study among pregnant women, the preva- vitamin D production in the skin and vitamin D oral intake, we
lence of vitamin D deficiency increased from 50.8% before the inquired about hours of daytime outdoor physical activity for lei-
COVID-19 lockdowns to 78% during the lockdowns [14]. In a Russian sure, sunscreen use (only for those who were engaged in daytime
study among healthcare workers in an infectious disease hospital, outdoor activity), fatty fish intake, and vitamin D supplementation
the prevalence of vitamin D deficiency and insufficiency was 60% or drug use. Other covariates selected included sex, age, comor-
and 30%, respectively [15]. In a Japanese study conducted during the bidities, occupation, shift work, smoking, alcohol consumption, and
pandemic (March 2021) among staff of a medical research institute, BMI, all of which have been linked to vitamin D status [2e7,20]. BMI
over 90% had vitamin D deficiency (25[OH]D < 20 ng/mL) [16]. was calculated as weight in kilograms divided by the squared
In Japan, COVID-19-related restrictions on social activities have height in meters.
been relaxed since May 2023 [17]. Thereafter, people's lives appear
to have returned to normal. However, healthcare workers remain 2.4. Statistical analysis
under pressure to maintain a lifestyle with low infection risk.
Accordingly, it is important to describe vitamin D status in relation Continuous variables are expressed as median (interquartile
to modifiable factors in this occupational group to create a strategy range [IQR]) or mean (standard deviation [SD]), while categorical
to overcome this poorly recognized nutritional deficiency. Thus, the variables are expressed as numbers (percentages). Logistic regres-
objective of the present study was to assess vitamin D status and sion was performed to calculate the odds ratios (OR) and 95 percent
examine the correlates of vitamin D deficiency three years after the confidence interval (95% CI) for vitamin D deficiency (25[OH]
start of the COVID-19 pandemic among the staff of a medical D < 20 ng/mL) for each covariate (crude model). In an adjusted
research institute in Tokyo and Chiba, Japan. model, we included age (<30, 30e39, 40e49, or 50 years), sex
(male or female), number of comorbidities (0 or 1), occupation
2. Methods (physician, nurse, allied healthcare worker, researcher, administra-
tive staff, or others), working hours (8, 9e10, or 11 h/day), BMI
2.1. Study design and participants (<18.5, 18.5e24.9, or 25.0 kg/m2), smoking status (non-smoker or
current smoker), alcohol drinking (non-drinkers; occasional
Data were derived from a repeat serological study of the COVID-19 drinkers [1e3 times/month]; or regular drinkers consuming <1 or
epidemic among the staff (Toyama area in Tokyo and Kohnodai area 1 go/day [one go contains approximately 23 g of ethanol]), daytime
in Chiba [35
N]) of the National Center for Global Health and Medi- outdoor physical activity (none, <60, 60e119, or 120 min/week),
cine (NCGM) with a specific mission of infectious disease control. fatty fish intake (<1, 1, 2e3, or 4 times/week), and vitamin D sup-
Details of the study design have been reported elsewhere [18,19]. plement or drug use (non-user or user). Occupation was not
Briefly, the participants were asked to complete an online or paper included as a covariate in the analysis of the association between
questionnaire and donate venous blood for the study at the time of shift work and vitamin D deficiency because over 80% of shift
periodic health checkup. Those with symptoms indicative of infec- workers were nurses. We additionally estimated the odds of vitamin
tion were instructed not to attend the checkup. In the latest survey D deficiency by dividing smokers according to the type of tobacco
conducted in June 2023, we measured serum 25(OH)D levels and product (exclusively heated tobacco product [HTP] users, dual users
inquired concerning COVID-19 related factors and lifestyle-related of HTP and conventional tobacco, or exclusively conventional ciga-
habits such as sun exposure habits, fish intake, and use of vitamin rette tobacco users). We also assessed the association between
D supplements. The study protocol was approved by the NCGM Ethics sunscreen use and daytime outdoor physical activity. For sensitivity
Committee (approval number: NCGM-G-003598), and informed analysis, we repeated the analyses after excluding users of vitamin D
consent was obtained from each participant. Of the 3206 invited supplements.
potential participants, 2569 (80%) completed the questionnaire and
donated blood samples. We subsequently excluded 26 participants 3. Results
with missing information regarding any of the following variables:
working hours (n ¼ 7), body weight (n ¼ 12), alcohol consumption The median (IQR) age was 38 (28e49) years and 71% of the
(n ¼ 3), leisure-time outdoor physical activity (n ¼ 6), sunscreen use participants were female. Of the participants, 39.8% had a history of
(n ¼ 4), fish intake (n ¼ 3), or vitamin D supplement use (n ¼ 5), severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)
leaving 2543 participants who were included in the final analysis. infection, and 95.6% had received three or more times of COVID-19
vaccine. The most prevalent occupation was nurse (36%), and one-
2.2. Vitamin D measurements third of the participants were shift workers. 18% of the participants
engaged in jobs with a high risk of infection. Younger or female
Casual blood samples were drawn into vacuum tubes and participants were more likely to be nurses and engaged in shift
centrifuged to separate the serum. Serum 25(OH)D concentrations work, spent fewer hours on outdoor physical activity, and used
211
A. Ito, S. Yamamoto, Y. Oshiro et al. Clinical Nutrition ESPEN 60 (2024) 210e216

sunscreen more frequently (Supplementary Tables 1 and 2). Shift outdoor physical activity; ORs (95% CI) were 1 (reference), 0.95
workers were more likely to be alcohol drinkers and smokers, (0.78e1.14), 0.68 (0.52e0.89), and 0.61 (0.45e0.83) for 0, <60,
worked longer hours, had a higher BMI, consumed fatty fish, and 60e119, and 120 min per week of daytime outdoor physical ac-
engaged in outdoor physical activity, similar to that in non-shift tivity, respectively (P for trend <0.01). This inverse association was
workers (Supplementary Table 3). observed among non- or occasional sunscreen users but not
The mean (SD) of serum 25(OH)D concentrations were 21.5 (6.9) among regular sunscreen users (P for interaction ¼ 0.02, Fig. 2).
ng/mL. The distribution of serum 25(OH)D levels is shown in Fig. 1. Regarding dietary factors, odds of vitamin D deficiency were
The proportions of participants with vitamin D insufficiency decreased with increasing fatty fish consumption: ORs (95% CIs)
(20e29 ng/mL) and deficiency (<20 ng/mL) were 44.9% and 45.9%, were 1 (reference), 0.72 (0.59e0.88), 0.55 (0.44e0.69), and 0.45
respectively. Compared with participants with vitamin D suffi- (0.26e0.77) for fatty fish consumption <1, 1, 2e3, and 4 times per
ciency (30 ng/mL), those with vitamin D insufficiency and defi- week, respectively (P for trend <0.01). Vitamin D supplement use
ciency tended to be female, younger, non-users of vitamin D was associated with a lower odds of vitamin D deficiency (OR [95%
supplements, and non-alcohol drinkers; they also reported fewer CI]: 0.36 [0.27e0.46]).
hours of daytime outdoor physical activity and consumed fatty fish There was a positive association between smoking and vitamin
less frequently (Table 1). D deficiency (OR [95% CI]: 1.36 [0.99e1.88]). When analyzed by
In the multivariable-adjusted model (Table 2), odds of vitamin type of tobacco products, exclusive HTP users exhibited higher odds
D deficiency were decreased with increasing hours of daytime of vitamin D deficiency (OR [95% CI]: 1.60 [0.95e2.70]), but this
association did not reach statistical significance. Regarding occu-
pational factors, the association between shift work and vitamin D
deficiency, which was significant in the unadjusted model (OR [95%
CI]: 1.31 [1.11e1.55]), disappeared after multivariable adjustment
(OR [95% CI]: 0.95 [0.77e1.16]). In a subgroup analysis among
nurses (n ¼ 908), who comprised the majority of shift workers
(87%), shift work was not associated with vitamin D deficiency
(OR ¼ 1.05). Other occupational factors, including occupation and
working hours, were not significantly associated with vitamin D
deficiency. Sensitivity analysis excluding vitamin D supplement
users showed similar associations (Supplementary Table 4).

4. Discussion

Among the staff of a medical and research institute in Tokyo and

Chiba, vitamin D insufficiency and deficiency were common three
years after the start of the COVID-19 pandemic (June 2023). Factors
Fig. 1. Distribution of 25-hydroxyvitamin D concentrations. associated with vitamin D deficiency included younger age, female

Table 1
Participant characteristics according to serum 25-hydroxyvitamin D concentrations.a

Total Serum 25-hydroxyvitamin D

Deficient Insufficient Sufficient

(<20 ng/mL) (20e29 ng/mL) (30 ng/mL)

Participants, n 2543 1166 1141 236

Sex, male 29.1 20.0 36.0 41.1
Age (years), median (IQR) 38 (28e49) 36 (26e48) 38 (28e50) 42 (31e51)
Comorbiditiesb, 1 disease 9.2 7.4 10.2 13.1
Occupation
Physician 16.0 13.2 19.2 14.8
Nurse 35.7 38.5 33.3 33.5
Allied healthcare worker 15.5 14.6 16.5 14.8
Researcher 12.2 12.7 10.9 15.7
Administrative staff 15.3 15.7 14.6 16.1
Others 5.4 5.3 5.5 5.1
Work shift, shift worker 31.7 34.9 29.3 27.5
Working hours, 9 h/day 38.4 38.2 38.2 35.6
Body mass index (kg/m2), median (IQR) 21.2 (19.5e23.4) 21.0 (19.3e23.4) 21.4 (19.6e23.5) 21.0 (19.8e23.4)
Smoking, current smoker 7.6 7.8 7.5 6.8
Alcohol drinking, 1 go/dayc 11.6 8.0 13.7 19.1
Daytime outdoor physical activity, 120 min/week 9.9 6.9 9.8 25.0
Sunscreen used, always 25.4 27.4 23.6 24.2
Fatty fish intake, 2 times/week 23.6 17.9 26.4 38.6
Vitamin D supplement or drug, weekly user 13.0 7.8 14.1 33.5

COVID-19, coronavirus disease 2019; IQR, interquartile range; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.
a
Data are presented as percentage unless otherwise indicated.
b
Hypertension, diabetes, chronic obstructive pulmonary disease, cardiovascular disease, cancer, chronic kidney disease, immunosuppressive diseases, diseases treated with
steroids, immunosuppressive, or anti-cancer drugs, and depression.
c
A conventional unit of volume: one go (180 mL) of sake contains 23 g of ethanol.
d
Asked only for those who were engaged in daytime outdoor activity on leisure (n ¼ 1425).

212
A. Ito, S. Yamamoto, Y. Oshiro et al. Clinical Nutrition ESPEN 60 (2024) 210e216

Table 2
Multivariable-adjusted odds ratio and 95% confidence interval for vitamin D deficiency (25-hydroxyvitamin D < 20 ng/mL) according to demographic and lifestyle factors.a

n n (%) of vitamin D deficiency Crude OR (95% CI) Adjusted OR (95% CI)b

Sex
Male 741 233 (31) 1.00 (reference) 1.00 (reference)
Female 1802 933 (52) 2.34 (1.95e2.80) 2.41 (1.93e3.01)
Age, years
<30 776 401 (52) 1.00 (reference) 1.00 (reference)
30e39 595 265 (45) 0.75 (0.61e0.93) 0.81 (0.64e1.02)
40e49 558 251 (45) 0.76 (0.61e0.95) 0.85 (0.67e1.09)
50 614 249 (41) 0.64 (0.52e0.79) 0.73 (0.56e0.95)
P for trend <0.01 P for trend ¼ 0.04
Comorbiditiesc
None 2310 1080 (47) 1.00 (reference) 1.00 (reference)
1 233 86 (37) 0.67 (0.50e0.81) 0.85 (0.62e1.17)
Occupation
Administrative staff 388 183 (47) 1.00 (reference) 1.00 (reference)
Physician 408 154 (38) 0.68 (0.51e0.90) 0.79 (0.57e1.10)
Nurse 908 449 (49) 1.10 (0.86e1.39) 0.76 (0.58e1.00)
Allied healthcare worker 393 170 (43) 0.85 (0.64e1.13) 0.87 (0.64e1.19)
Researcher 309 148 (48) 1.03 (0.76e1.39) 1.31 (0.95e1.82)
Others 137 62 (45) 0.93 (0.63e1.37) 1.31 (0.86e2.01)
Work shift
Non-shift worker 1737 759 (44) 1.00 (reference) 1.00 (reference)d
Shift worker 806 407 (50) 1.31 (1.11e1.55) 0.95 (0.77e1.16)d
Working hours, h/day
8 1566 709 (45) 1.00 (reference) 1.00 (reference)
9e10 775 367 (47) 1.09 (0.91e1.29) 1.21 (0.99e1.47)
11 202 90 (45) 0.97 (0.72e1.30) 1.20 (0.87e1.66)
P for trend ¼ 0.69 P for trend ¼ 0.08
Body mass index, kg/m2
<18.5 300 161 (54) 1.42 (1.11e1.81) 1.17 (0.91e1.52)
18.5e24.9 1884 847 (45) 1.00 (reference) 1.00 (reference)
25.0 359 158 (44) 0.96 (0.77e1.21) 1.20 (0.93e1.54)
P for trend ¼ 0.02 P for trend ¼ 0.86
Smoking
Non-smoker 2350 1075 (46) 1.00 (reference) 1.00 (reference)
Current smoker 193 91 (47) 1.06 (0.79e1.42) 1.36 (0.99e1.88)
Exclusive HTP user 66 34 (52) 1.26 (0.77e2.06) 1.60 (0.95e2.70)
Dual user of HTPs and conventional cigarettes 49 23 (47) 1.05 (0.60e1.85) 1.35 (0.73e2.49)
Exclusive conventional cigarette smoker 78 34 (44) 0.92 (0.58e1.44) 1.19 (0.73e1.93)
Alcohol drinking
None-drinker 836 419 (50) 1.00 (reference) 1.00 (reference)
1e3 times/month 687 349 (51) 1.03 (0.84e1.26) 0.98 (0.79e1.22)
<1 go/day 726 305 (42) 0.72 (0.59e0.88) 0.77 (0.62e0.95)
1 go/day 294 93 (32) 0.46 (0.35e0.61) 0.55 (0.41e0.74)
P for trend <0.01 P for trend <0.01
Outdoor physical activity, min/week
None 1118 562 (50) 1.00 (reference) 1.00 (reference)
<60 850 403 (47) 0.89 (0.75e1.07) 0.95 (0.78e1.14)
60e119 323 120 (37) 0.58 (0.45e0.75) 0.68 (0.52e0.89)
120 252 81 (32) 0.47 (0.35e0.63) 0.61 (0.45e0.83)
P for trend <0.01 P for trend <0.01
Sunscreen usee
Never 343 119 (35) 1.00 (reference) e
Sometime or occasional 437 166 (38) 1.15 (0.86e1.55) e
Always 645 319 (49) 1.84 (1.41e2.41) e
P for trend <0.01
Fatty fish intake, times/week
<1 1192 634 (53) 1.00 (reference) 1.00 (reference)
1 750 323 (43) 0.67 (0.55e0.80) 0.72 (0.59e0.88)
2e3 521 187 (36) 0.49 (0.40e0.61) 0.55 (0.44e0.69)
4 80 22 (28) 0.33 (0.20e0.55) 0.45 (0.26e0.77)
P for trend <0.01 P for trend <0.01
Vitamin D supplement or drug
Non-user 2212 1075 (49) 1.00 (reference) 1.00 (reference)
User 331 91 (27) 0.40 (0.31e0.52) 0.36 (0.27e0.46)

CI, confidence interval; HTPs, heated tobacco products; OR, odds ratio.
a
Data are shown as odds ratios and 95% confidence intervals for vitamin D deficiency as predicted using a logistic regression model.
b
Adjusted for age (<30, 30e39, 40e49, or 50 years), sex (male or female), number of comorbidities (0 or 1), occupation (physician, nurse, allied healthcare worker,
researcher, administrative staff, or others), working hours (8, 9e10, or 11 h/day), body mass index (<18.5, 18.5e24.9, or 25.0 kg/m2), smoking (non-smoker or smoker),
alcohol drinking (non-drinkers; occasional drinkers [1e3 times/month]; or regular drinkers consuming <1 or 1 go/day [one go contains approximately 23 g of ethanol]),
daytime outdoor physical activity (none, <60, 60e119, or 120 min/week), fatty fish intake (<1, 1, 2e3, or 4 times/week), and vitamin D supplement or drug use (non-user or
user).
c
Hypertension, diabetes, chronic obstructive pulmonary disease, cardiovascular disease, cancer, chronic kidney disease, immunosuppressive diseases, diseases treated with
steroids, immunosuppressive or anti-cancer drugs, and depression.
d
Occupation was not included in the multivariable-adjusted model because most shift workers (87%) were nurses.
e
Asked only for those who were engaged in daytime outdoor activity on leisure (n ¼ 1425).
213
A. Ito, S. Yamamoto, Y. Oshiro et al. Clinical Nutrition ESPEN 60 (2024) 210e216

vitamin D intake [21]. Our finding of a lower prevalence of vitamin

D deficiency among fatty fish eaters is consistent with previous
Japanese studies. Nakamura et al. [10] reported that higher con-
sumption of salmon, a popular fatty fish in Japan, was associated
with a higher prevalence of vitamin D sufficiency among 9084 in-
habitants aged 40 and 74 years. Nanri et al. [9] reported that higher
serum vitamin D concentrations were associated with higher fish
consumption among civil servants in Fukuoka. Although fish con-
sumption has decreased in Japan [22], the importance of fatty fish
intake needs to be emphasized as a major dietary factor in the
prevention of vitamin D deficiency in individuals who have limited
sun exposure.
While the use of vitamin D supplements has been consistently
linked to higher vitamin D levels [2], we are not aware of any
studies showing this association in Japan, where vitamin D sup-
plement use has been uncommon (7% in female nurses as of
2008e2013) [23]. In the present study, vitamin D supplement users
Fig. 2. Interaction between sunscreen use and daytime outdoor physical activity. Data
are shown as odds of vitamin D deficiency as predicted using a logistic regression
(13% of all participants) had a significantly lower prevalence of
model, including the following variables: age (<30, 30e39, 40e49, or 50 years), sex vitamin D deficiency than non-users, even after adjusting for all
(male or female), number of comorbidities (0 or 1), occupation (physician, nurse, covariates. Although the use of vitamin D supplements is not offi-
allied healthcare worker, researcher, administrative staff, or others), working hours cially recommended to the general public, to minimize the risk of
(8, 9e10, or 11 h/day), BMI (<18.5, 18.5e24.9, or 25.0 kg/m2), smoking status
diseases associated with vitamin D deficiency, it is an effective
(non-smoker, or smoker), alcohol drinking (non-drinkers; occasional drinkers [1e3
times/month]; or regular drinkers consuming <1 or 1 go/day [one go contains option for indoor workers, including healthcare workers, with low
approximately 23 g of ethanol]), fatty fish intake (<1, 1, 2e3, or 4 times/week), and exposure to sunlight and low dietary intake of vitamin D. Our
vitamin D supplement or drug use (non-user or user). findings highlight the need for monitoring vitamin D status and
associated daily habits of healthcare workers and, if necessary,
taking actions including vitamin D supplement use.
sex, fewer hours of daytime outdoor physical activity during leisure The higher prevalence of vitamin D deficiency observed among
time, lower intake of fatty fish, no use of vitamin D supplements, smokers, albeit not statistically significant, is consistent with the
smoking, and no alcohol consumption. This study is among the few results of a meta-analysis [5]. Substances in tobacco smoke can
large-scale studies that measured circulating vitamin D levels decrease serum 25(OH)D levels via several pathways. For example,
among healthcare workers during the COVID-19 pandemic. it decreases cutaneous vitamin D production by accelerating skin
In a smaller study involving 361 healthcare workers in Tokyo aging and modulates the expression of genes involved in vitamin D
during the COVID-19 pandemic, a much higher prevalence of homeostasis [24]. In our tentative analysis by the type of tobacco
vitamin D deficiency (92%) was documented [16]. This may reflect product, exclusive HTP users (34% of smokers) showed increased
the survey season (March) when sunlight is weak and individuals odds of vitamin D deficiency, with the magnitude of association
wear thick clothing. Studies have shown that vitamin D deficiency being somewhat higher than that among cigarette smokers
was widespread in Japan before the COVID-19 pandemic. For (OR ¼ 1.60 and 1.19 for exclusive HTP users and exclusive cigarette
example, among 5518 medical checkup attendants in Tokyo from smokers, respectively). Emissions from HTPs generally contain
April 2019 to March 2020, the proportions of those with vitamin D lower concentrations of harmful ingredients than those from reg-
insufficiency and deficiency was 19.8% and 78.5%, respectively [12]. ular cigarettes; however, the difference in nicotine levels is rela-
In a study of 9084 residents in the town of Niigata, the proportions tively small between HTPs and cigarettes [25]. Animal experiments
of those with vitamin D insufficiency and deficiency were 53.6% have shown that nicotine administration reduces serum 25(OH)D
and 37.4%, respectively [10]. In a study of 1786 employees levels [26,27]. With such biological evidence, the present findings,
(including 20% shift workers) of a manufacturing company in the which suggest an association between HTP use and vitamin D
Kanto area, the proportions of those with vitamin D insufficiency deficiency, warrant confirmation.
and deficiency were 41.6% and 50.8%, respectively [9]. Despite the We found a lower prevalence of vitamin D deficiency among
differences in vitamin D measurement methods and seasons of alcohol drinkers than among abstainers. A review of various types
blood draw across studies, these data support that nearly 90% or of studies (including alcoholics) concluded that the association
more of free-living individuals in Japan manifest vitamin D insuf- between alcohol consumption and circulating vitamin D levels was
ficiency or deficiency. inconsistent [20]. Interestingly, more recent large-scale population-
Cutaneous vitamin D production after UV exposure accounts for based studies have consistently reported higher 25(OH)D concen-
approximately 90% of the systemic vitamin D status [1]. Our study trations or a lower prevalence of vitamin D deficiency among
showed that the odds of vitamin D deficiency decreased linearly alcohol users [10,28e30]. In the UK biobank study, for example,
with increasing hours of daytime outdoor physical activity, which is alcohol drinkers had significantly lower odds of vitamin D defi-
similar to previous studies [6,7]. Meanwhile, UV protection prod- ciency than abstainers (OR ¼ 0.55 and 0.66 for weekly drinkers and
ucts have gained popularity in Japan with the increasing awareness daily drinkers, respectively) [30]. Although the mechanism under-
of sun protection for health and beauty. We found an inverse as- lying this association is unclear, the present study of healthcare
sociation between daytime outdoor activity and vitamin D defi- workers adds to the evidence that alcohol consumption may be
ciency among non- or occasional sunscreen users, but not among linked to better vitamin D status.
those who always used sunscreen. These results suggest that the While a meta-analysis showed a significant inverse (albeit
excessive use of sunscreen largely suppresses sun-induced vitamin weak) correlation between BMI and 25(OH)D levels [3], we found
D production in the skin during outdoor activities. no association between BMI category and vitamin D deficiency. One
Fish is an abundant source of vitamin D [1]. A Japanese study possible explanation is that the present study population had a
showed that vitamin D from fish accounts for 90.7% of the total lower BMI (median [IQR], 21.2 [19.5e23.4] kg/m2) and included few
214
A. Ito, S. Yamamoto, Y. Oshiro et al. Clinical Nutrition ESPEN 60 (2024) 210e216

participants with a BMI of 25 kg/m2 (14%), relative to the pop- less sunlight. Finally, our study was conducted among healthcare
ulations included in the meta-analysis in which 25 of 33 study workers at a single medical facility; hence, caution should be
populations had a mean BMI 25 kg/m2 [3]. Obesity-related exercised when generalizing the findings.
vitamin D deficiency may appear in populations with a higher BMI. In conclusion, vitamin D insufficiency and deficiency were
Sex differences in vitamin D status have been well documented highly prevalent among healthcare workers in Tokyo and Chiba,
[2]. Consistent with the literature, we also found that female par- Japan. Various demographic and lifestyle factors are associated
ticipants had higher odds of vitamin D deficiency than males. The with vitamin D deficiency. Health education on lifestyle modifica-
sex gap has been ascribed to the tendency of women to avoid sun tions targeting this occupational group is warranted to improve
exposure, which is a behavior that cannot be fully adjusted for in their vitamin D status in the COVID-19 era.
the analysis. Alternatively, the difference may be caused by greater
fat deposition (vitamin D storage) among women than men [31], Funding
limiting circulating forms of vitamin D to lower levels [32].
The ability of the skin to produce vitamin D decreases with This work was supported by the NCGM COVID-19 Gift Fund
advancing age [1]. In fact, lower circulating vitamin D has been (grant number 19K059), the Japan Health Research Promotion
documented in an older population, especially those aged 65 years Bureau Research Fund (grant number 2020-B-09), and the National
or older [7,28,29,33e35]. However, in the present study, vitamin D Center for Global Health and Medicine (grant number 21A2013D).
deficiency was less common in older participants. A possible
explanation for this result is that older participants consumed fatty Data sharing plan
fish more frequently and spent more hours engaging in outdoor
physical activity than younger participants (Supplementary The datasets generated and/or analyzed in the current study are
Table 1). After adjusting for these and other covariates, the asso- available from the corresponding author upon reasonable request.
ciation was attenuated and became statistically not significant,
except for those aged 50 years or older. It should be noted that the Conflict of interest disclosures
present study included only a few participants aged 65 years or
older (3%); thus, the effect of skin aging on circulating vitamin D All authors: No reported conflicts of interest.
levels, if any, may be small. A UK study among healthcare workers
(median [IQR] age 41 [30e50]) also reported no measurable dif- Statement of authors’ contributions to manuscript
ference in age between those with vitamin D deficiency and those
without [36]. Drs. AI, SY, and TM had full access to all data in the study and
Shift work has also been linked to vitamin D deficiency [4]. We took responsibility for the integrity of the data and accuracy of the
observed higher odds of vitamin D deficiency among shift workers data analysis.
in the unadjusted model. In the present study, shift workers tended Drs. AI and SY contributed equally to this article.
to be female, younger, and alcohol drinkers, whereas there was no Concept and design: SY, AI, and TM.
measurable difference in fatty fish intake or daytime outdoor Acquisition, analysis, or interpretation of data: SY, AI, and TM.
physical activity (Supplementary Table 3). After adjusting for Drafting of the manuscript: SY, AI, and TM.
covariates, the shift workevitamin D association disappeared, Critical revision of the manuscript for important intellectual con-
suggesting that shift work per se is not associated with vitamin D tent: SY, YO, NI, TN, TT, MK, TM, NA, HS, WS, and NO.
status. As healthcare workers who spend most of their working Statistical analysis: SY, AI, and MK.
hours indoors are already at high risk of vitamin D deficiency, Administrative, technical, or material support: SY, YO, NI, TN, MK,
engaging in shift work in this occupational group may not confer an TM, HS, NA, WS, and NO.
additional risk. Supervision: TN and NO.
The present study has several strengths, including a large
sample size, a well-defined study population with similar occupa- Acknowledgments
tional backgrounds, a high participation rate (80%), and few missing
data (0.1e0.5%). Nonetheless, our study has some limitations. First, We thank Mika Shichishima for her contribution to data
the associations derived from a cross-sectional study may not collection and the staff of the Laboratory Testing Department for
necessarily indicate causality. However, it is unlikely that circu- their contribution to 25(OH)D testing.
lating vitamin D influences occupational and lifestyle factors,
including diet and physical activity. Second, we did not measure Appendix A. Supplementary data
markers of inflammation that may influence blood vitamin D. For
example, circulating vitamin D concentrations were markedly low Supplementary data to this article can be found online at
among those with a C-reactive protein (CRP) above 10 mg/L [37]. In https://doi.org/10.1016/j.clnesp.2024.02.005.
this study, however, the participants were healthy workers without
symptoms indicative of infection. Moreover, CRP levels exceeding References
10 mg/L are rare in the general population in Japan [38]. Third, we
lacked data concerning the dose and type of vitamin D supplement, [1] Holick MF. Vitamin D deficiency. N Engl J Med 2007;357(3):266e81.
[2] Mithal A, Wahl DA, Bonjour JP, Burckhardt P, Dawson-Hughes B, Eisman JA,
skin pigmentation, sun umbrella use, and intake of foods rich in et al. Global vitamin D status and determinants of hypovitaminosis D.
vitamin D other than fatty fish (e.g., eggs and mushrooms). The Osteoporos Int 2009;20(11):1807e20.
possibility of bias owing to these unmeasured as well as residual [3] Saneei P, Salehi-Abargouei A, Esmaillzadeh A. Serum 25-hydroxy vitamin D
levels in relation to body mass index: a systematic review and meta-analysis.
confounding factors cannot be ruled out. Forth, the small number of
Obes Rev 2013;14(5):393e404.
tobacco product users (n ¼ 193) in the present study did not allow [4] Coppeta L, Papa F, Magrini A. Are shiftwork and indoor work related to D3
us to further analyze the association with each tobacco product. vitamin deficiency? A systematic review of current evidences. J Environ Public
Fifth, we assessed serum vitamin D concentrations during the Health 2018;2018:8468742.
[5] Yang L, Zhao H, Liu K, Wang Y, Liu Q, Sun T, et al. Smoking behavior and
sunny season (June). The prevalence of vitamin D insufficiency and circulating vitamin D levels in adults: a meta-analysis. Food Sci Nutr
deficiency and their determinants may differ during seasons with 2021;9(10):5820e32.

215
A. Ito, S. Yamamoto, Y. Oshiro et al. Clinical Nutrition ESPEN 60 (2024) 210e216

[6] Van Dam RM, Snijder MB, Dekker JM, Stehouwer CD, Bouter LM, Heine RJ, [22] Ministry of agriculture, forestry and fisheries, Japan. White paper on fisheries
et al. Potentially modifiable determinants of vitamin D status in an older (summary) (FY 2021 trends in fisheries, FY 2022 fisheries policy). 2021.
population in The Netherlands: the Hoorn Study. Am J Clin Nutr 2007;85(3): [23] Kishi M, Ideno Y, Nagai K, Lee JS, Suzuki S, Hayashi K. Use of dietary supple-
755e61. ments among Japanese female nursing professionals. J Nutr Sci Vitaminol
[7] Scragg R, Camargo CA. Frequency of leisure-time physical activity and serum 2022;68(3):213e20. 30.
25-hydroxyvitamin D levels in the US population: results from the third na- [24] Mousavi SE, Amini H, Heydarpour P, Amini Chermahini F, Godderis L. Air
tional health and nutrition examination survey. Am J Epidemiol 2008;168(6): pollution, environmental chemicals, and smoking may trigger vitamin D
577e86. deficiency: evidence and potential mechanisms. Environ Int 2019;122:67e90.
[8] Cui A, Zhang T, Xiao P, Fan Z, Wang H, Zhuang Y. Global and regional preva- [25] St Helen G, Jacob Iii P, Nardone N, Benowitz NL. IQOS: Examination of Philip
lence of vitamin D deficiency in population-based studies from 2000 to 2022: Morris International’s claim of reduced exposure. Tobac Control 2018
a pooled analysis of 7.9 million participants. Front Nutr 2023;10:1070808. Nov;27(Suppl 1):s30e6.
[9] Nanri A, Foo LH, Nakamura K, Hori A, Poudel-Tandukar K, Matsushita Y, et al. [26] Iwaniec UT, Haynatzki GR, Fung YK, Akhter MP, Haven MC, Cullen DM. Effects
Serum 25-hydroxyvitamin D concentrations and season-specific correlates in of nicotine on bone and calciotropic hormones in aged ovariectomized rats.
Japanese adults. J Epidemiol 2011;21(5):346e53. J Musculoskelet Neuronal Interact 2002;2(5):469e78.
[10] Nakamura K, Kitamura K, Takachi R, Saito T, Kobayashi R, Oshiki R, et al. [27] Fung YK, Iwaniec U, Cullen DM, Akhter MP, Haven MC, Timmins P. Long-term
Impact of demographic, environmental, and lifestyle factors on vitamin D effects of nicotine on bone and calciotropic hormones in adult female rats.
sufficiency in 9084 Japanese adults. Bone 2015;74:10e7. Pharmacol Toxicol 1999;85(6):181e7.
[11] Mizoue T, Kochi T, Akter S, Eguchi M, Kurotani K, Tsuruoka H, et al. Low serum 
[28] Santos A, Amaral TF, Guerra RS, Sousa AS, Alvares L, Moreira P, et al. Vitamin D
25-hydroxyvitamin D concentrations are associated with increased likelihood status and associated factors among Portuguese older adults: results from the
of having depressive symptoms among Japanese workers. J Nutr 2015;145(3): Nutrition UP 65 cross-sectional study. BMJ Open 2017;7(6):e016123.
541e6. [29] Liu X, Baylin A, Levy PD. Vitamin D deficiency and insufficiency among US
[12] Miyamoto H, Kawakami D, Hanafusa N, Nakanishi T, Miyasaka M, Furutani Y, adults: prevalence, predictors and clinical implications. Br J Nutr 2018;119(8):
et al. Determination of a serum 25-hydroxyvitamin D reference ranges in 928e36.
Japanese adults using fully automated liquid chromatography-tandem mass [30] Lin LY, Smeeth L, Langan S, Warren-Gash C. Distribution of vitamin D status in
spectrometry. J Nutr 2023;153(4):1253e64. the UK: a cross-sectional analysis of UK Biobank. BMJ Open 2021;11(1):
[13] Amrein K, Scherkl M, Hoffmann M, Neuwersch-Sommeregger S, e038503.
Ko€stenberger M, Tmava Berisha A, et al. Vitamin D deficiency 2.0: an update [31] Power ML, Schulkin J. Sex differences in fat storage, fat metabolism, and the
on the current status worldwide. Eur J Clin Nutr 2020;74(11):1498e513. health risks from obesity: possible evolutionary origins. Br J Nutr 2008;99(5):
[14] Agüero-Domenech N, Bernabeu E, García-Valentín A, Sarrio  n A, Jover S, 931e40.
Baranda J, et al. Influence of strict lockdown on vitamin D deficiency in [32] Carrelli A, Bucovsky M, Horst R, Cremers S, Zhang C, Bessler M, et al. Vitamin D
pregnant women: a word of caution. Nutrients 2023;15(8):1972. storage in adipose tissue of obese and normal weight women. J Bone Miner
[15] Karonova TL, Chernikova AT, Golovatyuk KA, Bykova ES, Grant WB, Res 2017;32(2):237e42.
Kalinina OV, et al. Vitamin D intake may reduce SARS-CoV-2 infection [33] Bhattoa HP, Bettembuk P, Ganacharya S, Balogh A. Prevalence and seasonal
morbidity in health care workers. Nutrients 2022;14(3):505. variation of hypovitaminosis D and its relationship to bone metabolism in com-
[16] Funaki T, Sanpei M, Morisaki N, Mizoue T, Yamaguchi K. Serious vitamin D munity dwelling postmenopausal Hungarian women. Osteoporos Int 2004;15(6).
deficiency in healthcare workers during the COVID-19 pandemic. BMJ Nutr [34] Brouwer-Brolsma EM, Vaes AMM, Van Der Zwaluw NL, Van Wijngaarden JP,
Prev Health 2022;5(1):134e6. Swart KMA, Ham AC, et al. Relative importance of summer sun exposure,
[17] Ministry of Health, Labour and Welfare. Response to COVID 19 (Novel Coro- vitamin D intake, and genes to vitamin D status in Dutch older adults: the B-
navirus) after the classification change. https://www.mhlw.go.jp/stf/covid-19/ PROOF study. J Steroid Biochem Mol Biol 2016;164:168e76.
kenkou-iryousoudan_00006.html (accessed Aug 29, 2023). [35] Díaz-Lo pez A, Paz-Graniel I, Alonso-Sanz R, Marque s Baldero C, Mateos Gil C,
[18] Mizoue T, Yamamoto S, Oshiro Y, Inamura N, Nemoto T, Horii K, et al. Cu- Arija Val V. Vitamin D deficiency in primary health care users at risk in Spain.
mulative and undiagnosed SARS-CoV-2 infection among the staff of a medical Nutr Hosp 2021;38(5):1058e67.
research centre in Tokyo after the emergence of variants. Epidemiol Infect [36] Faniyi AA, Lugg ST, Faustini SE, Webster C, Duffy JE, Hewison M, et al. Vitamin
2023;151:e48. D status and seroconversion for COVID-19 in UK healthcare workers. Eur
[19] Yamamoto S, Oshiro Y, Inamura N, Nemoto T, Horii K, Okudera K, et al. Respir J 2021;57(4):2004234.
Durability and determinants of anti-SARS-CoV-2 spike antibodies following [37] Duncan A, Talwar D, McMillan DC, Stefanowicz F, O'Reilly DSJ. Quantitative
the second and third doses of mRNA COVID-19 vaccine. Clin Microbiol Infect data on the magnitude of the systemic inflammatory response and its effect
Off Publ Eur Soc Clin Microbiol Infect Dis 2023;S1198e743X(23). 00244- on micronutrient status based on plasma measurements. Am J Clin Nutr
00246. 2012;95(1):64e71.
[20] Tardelli VS, Lago MPPD, Silveira DXD, Fidalgo TM. Vitamin D and alcohol: a [38] Tanaka M, Imano H, Kubota Y, Yamagishi K, Umesawa M, Muraki I, et al.
review of the current literature. Psychiatr Res 2017;248:83e6. Serum high-sensitivity C-reactive protein levels and the risk of atrial fibril-
[21] Nakamura K, Nashimoto M, Okuda Y, Ota T, Yamamoto M. Fish as a major lation in Japanese population: the circulatory risk in communities study.
source of vitamin D in the Japanese diet. Nutrition 2002;18(5):415e6. J Atherosclerosis Thromb 2021;28(2):194e202.

216