African Journal of AIDS Research

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/raar20

An analysis of the relationship between HIV risk

self-perception with sexual behaviour and HIV
status in South African older adults

Makandwe Nyirenda, Nonzwakazi Mnqonywa, Bomkazi Tutshana,

Jayganthie Naidoo, Paul Kowal & Joel Negin

To cite this article: Makandwe Nyirenda, Nonzwakazi Mnqonywa, Bomkazi Tutshana, Jayganthie
Naidoo, Paul Kowal & Joel Negin (2022) An analysis of the relationship between HIV risk self-
perception with sexual behaviour and HIV status in South African older adults, African Journal of
AIDS Research, 21:3, 277-286, DOI: 10.2989/16085906.2022.2090395

To link to this article: https://doi.org/10.2989/16085906.2022.2090395

© 2022 The Author(s). Co-published by NISC

Pty (Ltd) and Informa UK Limited, trading as
Taylor & Francis Group

Published online: 14 Sep 2022.

Submit your article to this journal

Article views: 136

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=raar20
African Journal of AIDS Research 2022, 21(3): 277–286 Copyright © The Authors

Open Access article distributed in terms of the Creative Commons Attribution License
[CC BY 4.0] (https://creativecommons.org/licenses/by/4.0)
AJAR
ISSN 1608-5906 EISSN 1727-9445
https://doi.org/10.2989/16085906.2022.2090395

Research Article

An analysis of the relationship between HIV risk self-perception with

sexual behaviour and HIV status in South African older adults

Makandwe Nyirenda1,2* , Nonzwakazi Mnqonywa3, Bomkazi Tutshana3, Jayganthie Naidoo3, Paul Kowal4
& Joel Negin5

1South African Medical Research Council, Burden of Disease Research Unit, Cape Town, South Africa.
2University of KwaZulu-Natal, College of Health Sciences, School of Nursing and Public Health, Durban, South Africa
3South African Medical Research Council, HIV Prevention Research Unit, Durban, South Africa

4International Health Transitions, Canberra, Australia

5The University of Sydney, School of Public Health, Sydney, Australia

*Correspondence: [email protected]

Objective: To examine how older adults perceive their own risk of acquiring HIV; and how this perception correlates
with their sexual behaviour and HIV status.
Methods: We used cross-sectional survey data for 435 adults aged 50 years and older from South Africa. All
participants completed a questionnaire on their basic socio-demographic and economic factors, self-reported
health, sexual behaviour, HIV knowledge and attitudes, and self-perceived risk of HIV acquisition. In addition,
anthropometrical measurements (weight, height, blood pressure, cholesterol) and HIV testing were conducted.
Multinomial logistic regressions were used to determine the association between self-perceived HIV risk
(categorised as “not at risk”, at “low risk”, at “high risk” and “didn’t know”) and being sexually active and testing
HIV-positive, controlling for socio-demographic, behavioural and health-related factors.
Results: Of the 435 respondents, 9.4% perceived themselves as at high risk of HIV infection, 18.9% as at low
risk and 53.6% believed they were not at risk of HIV. Most respondent who perceived themselves as at low risk
or not-at-risk at all of HIV were not sexually active. Older adults that were sexually active were more likely to
consider themselves as at high risk of acquiring HIV (relative risk ratio [RRR] 2.05; 95% confidence interval (CI)
1.05−4.00; p = 0.036), as well as to test HIV positive (RRR 10.5; 95% CI 3.8−29.1; p < 0.001). Self-perceived HIV risk
was significantly associated with age, sex, population group, and a greater awareness about HIV and how it is
transmitted.
Conclusions: Older persons who perceived themselves as at high risk of HIV were closely associated with sexual
activity and testing HIV positive. Therefore, there is an urgent need for older persons, particularly those who
remain sexually active, to screen and test for HIV routinely. Furthermore, there should be policy and programme
interventions, such as the development of a simple risk-assessment tool for older adults to determine their risk
for HIV. Older persons have been neglected in sexual health and HIV programmes. There is, therefore, a need to
encourage older persons to take up appropriate HIV risk reduction and prevention behaviours.

Keywords: self-perceived risk, sexual behaviour, older adults, South Africa

Introduction adults continue to be sexually active and may be infected

with HIV is often met with disbelief, stigma and discrimination
A combination of population ageing and increasing access (Autenrieth et al., 2018; Bendavid et al., 2012). Although
to antiretroviral therapy (ART) (Soomro et al., 2019; sexual activity is expected to decline with age, the literature
UNAIDS, 2016) as well as sexual activity in old age (Negin shows that older adults are much more sexually active than
et al., 2016; Nyirenda et al., 2018; Rosenberg et al., 2017; is commonly assumed, and continue to be sexually active,
Wang et al., 2015) has resulted in an increasing number of even at very advanced ages (Freeman & Anglewicz, 2012;
older persons aged 50 years and over living with HIV (United Negin et al., 2016; Rosenberg et al., 2017). Coupled with
Nations, 2013). While recent research has debunked many this reality about sexual activity, findings from across Africa
of the stereotypes and myths about lack of sexual activity in show that some populations of older adults may not have
older adults aged 50 years and older (Freeman & Anglewicz, correct information about HIV (Lekalakala-Mokgele, 2014;
2012; Ginsberg et al., 2005; Negin et al., 2016; Odimegwu Negin, Nemser, et al., 2012). But even where they may
& Mutanda, 2017; Wang et al., 2015), the reality that older have correct information about the risk of becoming HIV

African Journal of AIDS Research is co-published by NISC (Pty) Ltd and Informa UK Limited (trading as Taylor & Francis Group)
278 Nyirenda, Mnqonywa, Tutshana, Naidoo, Kowal & Negin

infected, they may not see it as relevant to them, leading study to investigate sexual health, HIV and comorbidity
to risky sexual behaviours among older persons, such as with non-communicable infections among older persons
having multiple partners and not using condoms. This places (SHIOP) conducted in semi-rural (Botha’s Hill) and urban
older adults at increased risk of getting infected with HIV (Chatsworth) communities of Durban, KwaZulu-Natal,
(Gómez-Olivé et al., 2020). South Africa. Further details on the SHIOP study have been
Unfortunately, older adults tend to be classified as low risk published elsewhere (Abbai et al., 2018; Nyirenda et al.,
(Aboderin, 2014; Aboderin & Beard, 2015; Rosenberg et al., 2018). Briefly, SHIOP was conducted between February and
2017), and HIV as a disease of young people (Dellar et al., May 2016, and included men and women aged 50 years and
2015; Naicker et al., 2015). HIV-related sexual behaviours older who were able to communicate in English or IsiZulu,
have, therefore, been studied quite extensively in younger were not cognitively impaired or terminally ill (by interviewer
persons. Relatively, few studies have addressed risky sexual assessment) and were willing and able to provide written
practices in adults aged 50 years and older. Furthermore, informed consent. Participants defined as illiterate, that is,
compared to younger HIV-infected adults, managing HIV in who were not able to read or write in either English or IsiZulu
older adults is complicated by comorbidity with other chronic marked the informed consent form using their thumbprint
conditions (Collins & Armstrong, 2020; Knight, Schatz et al., in the presence of an impartial witness who could not be
2018; Nachega et al., 2012; Negin, Martiniuk, et al., 2012) a member of the study team or employee of the research
whose symptoms may be conflated with HIV disease or organisation.
disregarded as a “normal” part of ageing (Negin et al., 2014). In this cross-sectional study, a convenience approach
These issues have contributed to a lack of HIV prevention was used to select participants. Trained recruiters went from
education and intervention programmes directed towards house to house to identify potential participants. Those that
older adults (Aboderin, 2014; Soomro et al., 2019). met the inclusion criteria were then invited to the research
As a result, little is known about sexual and HIV risk site located within the study area. Individuals who were not
perceptions in older adults (Heidari, 2016). Emerging data able to travel on their own or had mobility challenges were
suggests there could be considerable underestimation transported to and from the research site. Target sample
of risk by high-risk older adults (Syme et al., 2017). It has size was n = 217 participants in each community, with
been argued that older adults may internalise this perception final enrolments of n = 220 from Botha’s Hill and n = 215
that they are at low risk of HIV (Taylor et al., 2016) despite from Chatsworth. All study participants were provided with
practising some very risky sexual behaviours in a high monetary reimbursement of ZAR 150 (≈USD 10 at the
HIV setting. Data from rural South Africa have shown that time) as per local institutional review board guidelines for
the probability of older adults acquiring HIV is not zero time, inconvenience, and other expenses they may have
(Gómez-Olivé et al., 2020). incurred due to study participation. All participants provided
Adoption of HIV prevention interventions is driven by written informed consent and participant privacy was
a person’s accurate perceptions of their risks of acquiring observed during the entire study. Procedures followed in
HIV. The AIDS risk reduction model (Catania et al., 1990) this study were in accordance with the ethical standards of
and the UNAIDS’s 90-90-90 targets for ending AIDS by the Helsinki Declaration of 1975 as revised in 2013 (World
2020 (UNAIDS, 2014a) were modelled on the assumption Medical Association, 2013) and of the South African Medical
that uptake of HIV testing and other safer sexual behaviours Research Council (SAMRC) Ethics Committee which
like condom use and not having multiple partners were approved of this study (EC030-9/2015).
dependent on people’s awareness of HIV and understanding
their own risk of HIV infection (UNAIDS, 2014b). Developing Study procedures
HIV prevention and care programmes that either target All study participants were administered a questionnaire on
or include older adults (Knight, Mukumbang et al., 2018; their basic demographic characteristics, socio-economic
Lekalakala-Mokgele, 2014) requires an understanding about status, physical and mental health, sexual behaviour, and
how older adults perceive their own risk of HIV infection and knowledge and attitudes on HIV including perceptions of risk
how this relates to their sexual behaviours and HIV status. of HIV acquisition. Where, sexual behaviour is defined as any
Currently, there is substantial uncertainty and numerous activity engaged in to satisfy the sexual needs of an individual
misperceptions regarding the levels of HIV and the risks of (Chawla & Sarkar, 2019). In this context, we considered
getting HIV infection among older adults in South Africa (van risky sexual behaviour as any behaviour that could expose
Empel et al., 2021). Increasing the number of people with an the individual to a greater chance of becoming HIV infected
accurate assessment of their HIV risk is likely to contribute such as having unprotected sex, having multiple partners,
to increased uptake of HIV counselling and testing services. and not knowing a partner’s HIV status. Blood pressure,
This study therefore aimed to examine how older adults from weight and height for body mass index computation were also
two communities severely affected by HIV in KwaZulu-Natal, measured. A whole blood sample was collected from each
South Africa, perceived their own risk of acquiring HIV; and participant and was used for HIV testing using two parallel
how this self-perceived risk correlated with their sexual rapid HIV tests: Determine HIV-1/2 (Abbott Laboratories,
behaviour and HIV status. Japan) and UnigoldTM Recombigen® HIV in accordance
with manufacturer instructions. Results of the HIV testing
Methods were available within thirty minutes after completing the
Study population questionnaire and were only disclosed to the participant by a
Information of self-perceived risk, sexual behaviour and trained HIV counsellor after pre- and post-test counselling as
HIV status was extracted from data collected as part of a per HIV standard operating procedures.
African Journal of AIDS Research 2022, 21(3): 277–286 279

Statistical analysis (LR), 53.6% “not at risk” (NR), and 18.2% “don’t know” (DK)
The main outcome variable of interest in this study was (Table 1). Among participants aged 50–59 years, 43.1%
self-perceived risk of HIV acquisition. This was determined perceived themselves as NR, 18.3% as LR, 15.3% as
from the question: “How would you rate your risk of getting HR, and 23.3% as DK. In the older age groups, over 60%
HIV?” This self-assessed risk of HIV was categorised as perceived themselves as NR with only 5.6% among 60–69
“not at risk”, “low risk”, “high risk”, or “don’t know”. Although and 1.4% in 70+ ages perceiving themselves as HR. There
participants were asked later in the questionnaire whether were no significant differences by sex in the perceived risk
they knew their HIV status, there was no further probe of HIV (p = 0.540), although a slightly higher proportion of
on whether they were HIV-negative or positive to avoid older men (11.7%) than older women (8.5%) rated their risk
involuntary disclosure to the interviewer. Therefore, only the of HIV as high. Over two-thirds of non-African participants
HIV result obtained after testing as part of this study was (69.2%) said they were NR compared to less than 50% of
used as respondent’s HIV sero-status in the analysis. HIV African participants (p = 0.001). African participants were
testing was conducted after completing the questionnaire. significantly more likely to rate their risk as high (11.6%)
Thus, all participants including those who self-reported that compared to non-Africans (2.8%) (p = 0.001). There were
they already knew their HIV status were asked for their significant differences by source of income (p = 0.033), with
perceived risk to acquire HIV. The potential bias from this 25.0% of participants with no income perceiving themselves
approach is acknowledged in the limitations section of this as HR compared to 6.7% of participants receiving a
article. government grant. There was similarly a significant
HIV status (negative or positive) was used as a predictor difference by place of residency (p = 002). Older adults from
variable in this analysis. Explanatory variables included in the urban area of Chatsworth were more likely to report
this analysis were sex (male/female); age in years, which NR of HIV (62.3%) than those from the semi-rural Botha’s
was recategorised into broad age groups (50–59, 60–69 and Hill area (45.0%). There were no statistically significant
70+); population group (black African or non-African (which differences in perceived HIV risk by religion (p = 0.856),
combined Asian and one white participant)); religion (none, marital status (p = 0.727), education (p = 0.590) and
Christian, other); employed (yes/no), highest education level employment status (p = 0.430).
attained (none, primary, secondary or higher); marital status
(never been married, currently married, and previously HIV knowledge, attitudes and self-perceived risk of HIV
married (separated, divorced, widowed)); condom use (yes/ Table 2 shows that overall awareness of HIV was very
no); knowledge and attitudes of HIV; and sexual behaviour high, with over 90% of respondents knowing that HIV can
(number of sexual partners in last 12 months, number of be transmitted through sex with an HIV-infected person.
lifetime sexual partners, consulting a health worker for Similarly, other means of contracting HIV such as through
sexual needs or activity, treatment for STIs, and use of blood transfusion (65%), contaminated needles (88%) and
traditional healers for STIs). through mother-to-child transmission (85%) were known
Chi-square and t-tests were used to determine statistical by most older adults (data not shown). Furthermore, nearly
differences in distributions of the categorical and continuous two-thirds (72.6%, n = 316) agreed or strongly agreed
variables by self-perceived HIV risk. Multinomial logistic (72.6%) that even older adults ≥ 50 years were at risk of HIV
regressions were used to determine the association infection, although only approximately one-in-ten (11.1%)
between self-perceived HIV risk and being sexually active of participants saying older adults were also at risk of HIV
and HIV infected, controlling for socio-demographic (age, believed they themselves were at HR of acquiring HIV.
sex, population group, religion, marital status, education, Among participants who disagreed that older people were
employment), sexual behaviour (being sexually active, also at risk of HIV, most of them believed they were not at
condom use), and HIV knowledge and attitude factors (ever risk of getting HIV (71.4%).
tested for HIV, knowledge of partner’s HIV status, belief that When asked what older people could do to prevent
older people are also at risk of HIV). Data were analysed themselves from getting HIV, well over half (58.4%)
using STATA version 14.2 (StataCorp, 2014). mentioned that they could use condoms. However, over
one-in-three (38.4%) of the respondents believed there
Results was nothing they could do to prevent themselves from
Sample characteristics getting HIV. Over 77% of individuals who said there was
Participants in this study were aged 50 to 90 years with “nothing they could do”, perceived themselves as not at risk
a median age of 61 years (interquartile range 12). The of HIV, while only 12.6% of those who said they could use
majority of the 435 participants were women (70.6%); black condoms believed they were at HR of HIV. Encouragingly,
African (75.4%); Christian (78.4%); had attained primary 64.8% of the older people in this study had tested for HIV,
level of education (50%); were not employed (89.9%); and of whom 95.4% knew their own HIV status, but as stated
were recipients of government grants (68.7%). Regarding earlier there was no further prompt for participants to
marital status, a quarter of the older adults had never been disclose what their HIV status was. Of the 38.2% (n = 166)
married; 35.4% were currently married and two-in-five were who did not know their HIV status or preferred not to say,
previously married (separated, divorced or widowed). By the majority (62.7%) believed they were at NR of acquiring
place of residence, distribution was nearly even: 50.6% from HIV; less than 2% believed they were at HR of HIV; and
semi-rural Botha’s Hill and 49.4% from urban Chatsworth. 19.3% did not know their level of HIV risk. Approximately
Overall, the perception of HIV risk among the 435 study 53% (n = 116) knew their partner’s HIV status, the majority
participants was 9.4% “high risk” (HR), 18.9% “low risk” of whom perceived themselves as not at risk or at LR of
280 Nyirenda, Mnqonywa, Tutshana, Naidoo, Kowal & Negin

Table 1. Socio-demographic factors and self-perceived HIV risk among older persons in South Africa (N = 435)

Total Not at risk Low risk High risk Don’t know

p-valueƚ
n (%) n (%) n (%) n (%) n (%)
Overall sample 435 (100.0) 233 (53.6) 82 (18.9) 41 (9.4) 79 (18.2)
Age group <0.001
50–59 202 (46.4) 87 (43.1) 37 (18.3) 31 (15.3) 47 (23.3)
60–69 161 (37.0) 99 (61.5) 32 (19.9) 9 (5.6) 21 (13.0)
70+ 72 (16.6) 47 (65.3) 13 (18.1) 1 (1.4) 11 (15.3)
Sex 0.540
Female 307 (70.6) 164 (53.4) 57 (18.6) 26 (8.5) 60 (19.5)
Male 128 (29.4) 69 (53.9) 25 (19.5) 15 (11.7) 19 (14.8)
Population group 0.001
Non-African¤ 107 (24.6) 74 (69.2) 18 (16.8) 3 (2.8) 12 (11.2)
Black African 328 (75.4) 159 (48.5) 64 (19.5) 38 (11.6) 67 (20.4)
Religion 0.856
Christian 341 (78.4) 183 (53.7) 65 (19.1) 29 (8.5) 64 (18.8)
Hindu 68 (15.6) 38 (55.9) 12 (17.6) 8 (11.8) 10 (14.7)
None 26 (6.0) 12 (46.2) 5 (19.2) 4 (15.4) 5 (19.2)
Marital status 0.727
Never married 108 (24.8) 53 (49.1) 18 (16.7) 12 (11.1) 25 (23.1)
Currently married 154 (35.4) 83 (53.9) 32 (20.8) 14 (9.1) 25 (16.2)
Previously married₴ 173 (39.8) 97 (56.1) 32 (18.5) 15 (8.7) 29 (16.8)
Education attained 0.590
Never been to school* 70 (16.1) 37 (52.9) 10 (14.3) 7 (10.0) 16 (22.9)
Primary 218 (50.1) 115 (52.8) 44 (20.2) 17 (7.8) 42 (19.3)
Secondary or higher 147 (33.8) 81 (55.1) 28 (19.0) 17 (11.6) 21 (14.3)
Employmentᵫ 0.430
Unemployed 391 (89.9) 210 (53.7) 70 (17.9) 38 (9.7) 73 (18.7)
Employed 44 (10.1) 23 (52.3) 12 (27.3) 3 (6.8) 6 (13.6)
Income 0.033
Government grant 299 (68.7) 172 (57.5) 57 (19.1) 20 (6.7) 50 (16.7)
Other 124 (28.5) 55 (44.4) 24 (19.4) 18 (14.5) 27 (21.8)
None 12 (2.8) 6 (50) 1 (8.3) 3 (25.0) 2 (16.7)
Place of residency 0.002
Semi-rural 220 (50.6) 99 (45.0) 53 (24.1) 23 (10.5) 45 (20.5)
Urban 215 (49.4) 134 (62.3) 29 (13.5) 18 (8.4) 34 (15.8)
¤ South Africans of Asian origin, including one white (Caucasian) participant
*Includes participants who responded “don’t know” to education question (n = 4)
ᵫ Includes self-employed participants (n = 8)

₴ Composed of participants who were currently separated, divorced, or widowed

ƚ Pearson chi-square test of strength of the association between self-perceived risk and the factors

HIV (73.2%; n = 85). Among those who did not know their (73.4%). Only 7.3% of those not sexually active perceived
partner’s HIV status or preferred not to say (47%; n = 103), themselves as at HR of HIV compared to 13.0% among the
only a very small proportion (9.7%; n = 10) perceived sexually active. Of concern is the 16.1% of sexually active
themselves as at HR of HIV. older persons who did not know their HIV risk level.
Looking further into the sexual behaviour of participants,
Self-perceived risk and sexual behaviour only 25.1% had ever used a condom. Most respondents who
Overall, 37% (n = 161) of participants in this study said they said they had never used a condom perceived themselves
had been sexually active in the last 12 months (Table 3). as NR (57.7%) compared to 41.3% who thought they were at
Sexual activity was defined as having penetrative vaginal NR among condom users. Approximately similar proportions
intercourse between a man and a woman. Of the sexually of one-in-five of older adults who had ever used a condom
active participants, 92.5% had only one partner, while 7.5% perceived themselves as at LR (17.4%) and at HR (21.1%)
had two or more sexual partners in the last 12 months. of getting HIV. Among those to had ever used a condom,
Approximately 50% (n = 79) of older adults who were 70.6% (n = 77) reported to have used a condom with a most
sexually active perceived themselves as NR of HIV; a further recent partner. Among those using a condom with their most
21.7% said they were only at LR. Only 13.0% of the sexually recent partner, around a third said they were NR (36.4.0%),
active older adults thought they were at HR of HIV infection, followed by 27.3% who said they were at high HIV risk and
while 16.1% of the sexually active did not know their level 22.1% did not know.
of HIV risk. Most older adults who had not been sexually Respondents were asked about their sexual-related
active in the last 12 months rated themselves as NR or LR health seeking behaviour, that is, whether they had
African Journal of AIDS Research 2022, 21(3): 277–286 281

Table 2. HIV knowledge, attitudes and self-perceived risk of HIV in South African older adults (N = 435)

Total* Not at risk Low risk High risk Don’t know

Variable p-value
n n (%) n (%) n (%) n (%)
Overall sample 435 233 (53.6) 82 (18.9) 41 (9.4) 79 (18.2)
Knows HIV can be transmitted through sex 0.002
with infected person
Yes 406 222 (54.7) 78 (19.2) 40 (9.9) 66 (16.3)
No 29 11 (37.9) 4 (13.8) 1 (3.4) 13 (44.8)
Older adults are at risk of HIV 0.006
Agree/strongly agree 316 160 (50.6) 69 (21.8) 35 (11.1) 52 (16.5)
Neither agree nor disagree 56 28 (50.0) 8 (14.3) 4 (7.1) 16 (28.6)
Disagree/strongly disagree 63 45 (71.4) 5 (7.9) 2 (3.2) 11 (17.5)
What can do to protect self from HIV <0.001
Use condoms 254 101 (39.8) 64 (25.2) 32 (12.6) 57 (22.4)
Do nothing 167 129 (77.2) 17 (10.2) 9 (5.4) 12 (7.2)
Other, specify 14 3 (21.4) 1 (7.1) 0 (0.0) 10 (71.4)
Ever tested for HIV 0.001
Yes 282 139 (49.3) 56 (19.9) 38 (13.5) 49 (17.4)
No 153 94 (61.4) 26 (17.0) 3 (2.0) 30 (19.6)
Know your HIV status <0.001
Yes 269 129 (48.0) 55 (20.4) 38 (14.1) 47 (17.5)
No/prefers not to say 166 104 (62.7) 27 (16.3) 3 (1.8) 32 (19.3)
Know partner’s HIV status§ 0.008
Yes 116 52 (44.8) 33 (28.4) 17 (14.7) 14 (12.1)
No/Prefers not to say 103 54 (52.4) 14 (13.6) 10 (9.7) 25 (24.3)
*Percentages are calculated column-wise; for the perceived risk categories, percentages are calculated across the rows
§ Excluded (n = 216) individuals as this question was not applicable to persons who said they had not been sexually active in the last 12 months

Table 3. Sexual behaviour and self-perceived HIV risk among older persons in South Africa

Total Not-at-risk Low-risk High-risk Don’t know

Variable p-value
n n (%) n (%) n (%) n (%)
Sexually active in the last 12 months 0.102
No 274 154 (56.2) 47 (17.2) 20 (7.3) 53 (19.3)
Yes 161 79 (49.1) 35 (21.7) 21 (13.0) 26 (16.1)
Partners last 12 months* 0.346
1 149 75 (50.3) 33 (22.1) 19 (12.8) 22 (14.8)
2+ 12 4 (33.3) 2 (16.7) 2 (16.7) 4 (33.3)
Type of partner last 12 months* 0.590
Spouse 100 51 (51.0) 22 (22.0) 11 (11.0) 16 (16.0)
Regular partner 52 22 (42.3) 13 (25.0) 8 (15.4) 9 (17.3)
Casual partner 9 6 (66.7) 0 (0.0) 2 (22.2) 1 (11.1)
Ever used condoms <0.001
No 326 188 (57.7) 63 (19.3) 19 (5.8) 56 (17.2)
Yes 109 45 (41.3) 19 (17.4) 22 (20.2) 23 (21.1)
Condom used most recent partner** <0.001
No 115 64 (55.7) 27 (23.5) 5 (4.3) 19 (16.5)
Yes 46 15 (32.6) 8 (17.4) 16 (34.8) 7 (15.2)
Consult health worker for sexual needs of activity 0.001
No 379 214 (56.5) 71 (18.7) 31 (8.2) 63 (16.6)
Prefers not to say 14 9 (64.3) 2 (14.3) 0 (0.0) 3 (21.4)
Yes 42 10 (23.8) 9 (21.4) 10 (23.8) 13 (31.0)
*Question restricted to only those who had ever used a condom
**Question restricted to only participants who had ever used a condom and had been sexually active in last 12 months

consulted a health care worker about their sexual needs or HIV infection and only 8.2% perceived themselves at HR.
activity. Among participants who were sexually active in the Among those who had consulted a health worker for sexual
last 12 months, about 26.1% had consulted a health worker needs or activity, similar proportions perceived themselves
regarding their sexual needs or activity. An overwhelming as NR as at HR (23.8%). However, many respondents who
majority (73.9%) of sexually active older people had never had consulted a health worker said they did not know their
consulted a health worker. Most (56.5%) of those never risk (31.0%).
to consult with a health worker believed they were NR of
282 Nyirenda, Mnqonywa, Tutshana, Naidoo, Kowal & Negin

Association of self-perceived risk with sexual activity Discussion

and HIV status
In multinomial logistic regression analyses we found that While HIV-related sexual behaviours have been studied quite
there was an association between sexual activity and extensively in younger persons, few studies have addressed
self-perceived HIV risk as displayed in Figure 1. Overall, sexual health in older adults aged 50 years and older,
older persons who were sexually active were significantly including the growing population of older adults living with HIV
more likely to perceive themselves as at HR of HIV infection (Autenrieth et al., 2018; Houle et al., 2020; Pilowsky & Wu,
(RRR 2.05; 95% CI 1.05−4.00; p = 0.036). When segregated 2015). Using data from a cross-sectional study among adults
by sex, the relationship was only significant among women. aged 50 years and older in South Africa, this article makes a
The odds of perceiving self as at HR of HIV were four times valuable contribution to the limited data on how older adults
greater in women who were sexually active compared to rate their vulnerability to getting HIV infection; and how this
among women who were not sexually active. subjective self-assessment of HIV risk correlated with sexual
In this study, 16.1% of the sample were HIV-positive activity and testing HIV-positive. In this study we found that
(16.6% women; 14.8% men). As shown in Table 4, we nearly three-quarters of the older adults believed they were
found that self-perceived HIV risk was strongly associated not at risk or at very low risk of getting HIV.
with HIV status in older adults in this study. With reference Only about one-in-ten, and approximately 26% among
to older people who perceived themselves as LR, those sexually active respondents, perceived themselves as
who perceived themselves as HR were significantly more at high risk of HIV among these respondents dwelling in
likely to test HIV-positive in both the unadjusted (RRR 15.5; communities severely impacted by HIV. Self-perceived
p < 0.001) and adjusted (RRR 10.5; p < 0.001) models. We risk was not significantly associated with sex, religion,
also found older people saying DK were more likely to test marital status, education or employment status, similar
HIV-positive (RRR 2.6; p = 0.023) in the unadjusted models to previous findings (van Empel et al., 2021), but was
but was marginally not statistically significant in adjusted significantly associated with age, population group, income
analyses (RRR 2.4; p = 0.054). On the other hand, older and place of residency. That is, increasing age was strongly
people perceiving themselves as NR were less likely to be associated with lower proportions rating themselves as
HIV infected (RRR 0.4; p = 0.082) in adjusted models. When at high risk of HIV, consistent with data from the United
we excluded participants who already knew their HIV status Nations (UNAIDS, 2014b). Furthermore, African relative to
prior to testing in the study, those perceiving themselves as non-African participants were more likely to believe they
at HR of HIV were still more likely to test HIV-positive (RRR were at increased risk of HIV. By income status, those with
17.9; p = 0.021). no income sources were more likely to perceive themselves
as at increased risk of HIV, as were participants from the
semi-rural areas.
Our findings appear to contradict a recent analysis of the
Point estimate perceived risk of HIV acquisition and prevalence using data
Overall 95% CI from a population-based household survey nested within
a longitudinal surveillance system in rural South Africa,
which found that a large majority of the older adults with a
mean age of 61 years (± 12.6) overestimated their risk of
Males acquiring HIV (van Empel et al., 2021). It must be noted
though that the study by van Empel and others (2021) asked
respondents to rate the risk of HIV acquisition in a single
Females heterosexual sex act involving sero-discondant couples,
which may be very different from perceived risk in the
general population where the HIV status of the partner is
less likely to be known but assumed to be negative (Houle
0 1 2 3 4 5 6 7 8 9 10 et al., 2020). The low self-perceived HIV risk in this study
RELATIVE RISK RATIO was not a result of ignorance about HIV, as about nine-in-ten
Figure 1. Likelihood to perceive self as at as at high risk of HIV of the study participants had heard about HIV and had
among sexually active older persons in South Africa excellent knowledge of how HIV is transmitted. In addition,

Table 4. Unadjusted and adjusted risk of HIV infection by self-perceived HIV risk in South African older persons

Unadjusted Adjusted*
Relative risk ratio (95% Cl) p-value Relative risk ratio (95% Cl) p-value
Low risk 1.00 (base outcome) 1.00 (base outcome)
Not at risk 0.4 (0.1–0.9) 0.024 0.4 (0.2–1.1) 0.082
High risk 15.5 (6.1–39.4) <0.001 10.5 (3.8–29.1) <0.001
Don’t know 2.6 (1.1–6.0) 0.023 2.4 (1.0–6.0) 0.054
*Adjusted for age, sex, population group, employment status, sexual activity, ever tested for HIV, knows own HIV status, and knows partner’s
HIV status
African Journal of AIDS Research 2022, 21(3): 277–286 283

we found that over three-quarters were affirmative that even among persons who are not currently sexually active and
older adults were at risk of getting HIV; but it appears study may not have been for some years. Likewise, a higher
participants may have associated this risk with “others”, not perceived risk of HIV is to be expected among older persons
themselves. Our findings of most participants believing they who are currently sexually active as our findings show
were at low to no risk of HIV appear to be in line with global a near-universal knowledge that HIV can be transmitted
data that suggests older adults tend to underestimate their during sex with an HIV-infected partner. While these findings
risk of getting infected with HIV despite a trend of a growing are in the expected direction, the literature suggests there
proportion of older adults living with HIV (UNAIDS, 2014b). could be underestimation of HIV risk among older persons.
Although sexual activity does decline with increasing A review article of mainly American studies concluded that
age, the rate of sexual activity of 37% overall and over 64% older adults tend to underestimate their vulnerability to HIV
among men is prevalent enough to pose a considerable infection and report a low prevalence of condom use even
risk of acquiring HIV. Studies from across sub-Saharan when they were in non-marital or short-term relationships
Africa have shown sexual activity in older adults range (Pilowsky & Wu, 2015). That is, not everyone who perceives
from 30% to 75% (Chepngeno-Langat & Hosegood, 2012; themselves as at low or no risk for HIV is truly at low or no
Freeman & Anglewicz, 2012; Negin et al., 2016; Nicolosi risk due to the possibility of risk underestimation among
et al., 2004; Odimegwu & Mutanda, 2017; Rosenberg et older people. According to the optimistic bias theory (Sharot,
al., 2017; Todd et al., 2009; Trompeter et al., 2012). In this 2011), persons may underestimate their risk of getting a
study, we showed that how an individual perceived their disease like HIV on the belief that it is unlikely to happen
risk of HIV infection was strongly related to their sexual to them (Chowdhury et al., 2014; O’Sullivan, 2015). Thus,
behaviour patterns and their HIV status. That is, we found older persons, based on the optimistic bias theory, may
that individuals who were sexually active were more likely underestimate their risk of getting HIV as they compare
to perceive themselves as being at high risk of HIV infection themselves to population groups that are widely regarded
compared to not at risk. Furthermore, our results showed as most vulnerable to HIV infection, viz., adolescent girls,
that sexually active individuals who stated that they were at sex workers, men who have sex with men and injecting
high risk of HIV were indeed more likely to test HIV-positive. drug users (Dellar et al., 2015; HSRC, 2018). Older adults
Very few of the sexually active older adults in this study may also downplay their vulnerability to HIV infection due
had used a condom during their last sexual encounter, and to discrimination and stigmatising behaviour towards older
condoms were rarely mentioned when asked what could people living with HIV (Knight, Schatz, et al., 2018; UNAIDS,
be done to protect themselves from HIV infection. This is 2014b). Thus, although we are limited by our data to
consistent with findings from another study in South Africa determine whether there was underestimation of HIV risk in
that found knowledge of condom use to be high in older this study, it is probable given these theoretical bases that
adults, but usage very low (Lekalakala-Mokgele, 2014). this may have contributed to the large proportion of older
The fact that a large proportion of older persons were in adults in this study reporting a low HIV risk.
stable marital relations with a single partner may explain A study among young adults aged 15–24 years showed
the low condom use we observed. As reported by Taylor that many of the young people who assessed themselves
and others (2016), not being at risk of falling pregnant may as at no or little risk were actually at moderate or high risk
also contribute to low condom use among post-menopausal of contracting HIV (Prata et al., 2006). More importantly,
women (Taylor et al., 2016). that study showed that for both young men and young
In this study, nearly one-in-five respondents did not know women, safer sexual practices measured by condom use
their level of risk to HIV infection. Not being able to perceive were nearly twice as high in those who correctly assessed
level of HIV risk is deeply concerning as it could perpetuate their risk compared to those who did not (Prata et al., 2006).
risky sexual behaviours and place individuals and their Therefore, it is imperative to develop programmes that can
partners at greater risk of HIV infection. This can be seen get older adults, including those who believe they are at low
from our findings that respondents who did not know their risk of HIV, to test more regularly during their routine visits to
level of HIV risk were also associated with greater likelihood health facilities or other testing centers within the community.
of being sexually active and testing HIV-positive. Another Currently, older adults are being overlooked in sexual
worrying finding was that majority of older adults, irrespective health and HIV prevention programmes due to their lack
of their perceived HIV risk category, did not know their of self-understanding of their levels of risk to HIV infection.
partner’s HIV status. Not knowing one’s own HIV status and Findings from this study enhance the understanding of risk
that of one’s sexual partner(s) significantly increases an perceptions among older persons, and how this perceived
individual’s HIV risk as the person would be unlikely to adopt risk is related to sexual behaviour and testing HIV-positive
safer sexual practices or to take up interventions to prevent which is key to developing effective tailor-made interventions
HIV transmission and acquisition (UNAIDS, 2014b). to prevent HIV infection in older people.
Factors that contribute to increasing HIV transmission in A group of older persons for whom some attention is
older adults typically fall into three categories: (1) inadequate particularly required is that of sexually active older persons.
HIV prevention education, (2) insufficient communication Given declining rates of sexual activity and HIV incidence
with health providers regarding HIV transmission, and (3) with increasing age, being able to identify accurately sexually
a poor HIV risk awareness or perception (Savasta, 2004; active older adults who are at high risk of acquiring HIV is
UNAIDS, 2014b). Our analysis shows that majority of key to developing targeted HIV prevention interventions. This
older persons believed they were at low or no risk of HIV. is particularly important in resource-limited settings where
This self-perception of risk is to be expected, particularly population-wide programmes for older people may not be
284 Nyirenda, Mnqonywa, Tutshana, Naidoo, Kowal & Negin

cost effective or feasible (Gómez-Olivé et al., 2020; UNAIDS, targeted at older adults to help identify those at high risk
2021). Older adults need to be made aware that even though of HIV, especially those that remain sexually active, is
they may be in stable marital relationships with one sexual recommended. This would enable older persons to adopt
partner, there is still a risk of them getting HIV (Gómez-Olivé appropriate HIV risk reduction and prevention options.
et al., 2020). The near-universal knowledge about HIV and Such a tool could also be useful for older persons who do
modes of transmission needs to be matched with a true not know or may be underestimating their level of risk of
self-assessment of HIV risk if safer sexual practices are to be getting HIV (Chowdhury et al., 2014; O’Sullivan, 2015; Syme
adopted. The uptake of HIV prevention interventions, whether et al., 2017; UNAIDS, 2014b). Older adults have been left
condom use, pre-exposure prophylaxis (PrEP), medical male behind in HIV prevention programmes and policies because
circumcision, or universal test and treat, is highly dependent they tend to be classified by others and themselves as at
on personal judgement of the risks of acquiring HIV (Pilowsky low HIV risk (Aboderin, 2014; Mutevedzi & Newell, 2011;
& Wu, 2015; UNAIDS, 2014b). Hence, targeted policy and UNAIDS, 2014b, 2021). There exists now more than ever
programme interventions for sexually active older adults are a need for sexual education, couched within older persons’
urgently needed to help older persons determine their true own language and context, to equip them better to manage
level of HIV acquisition risk. their sexual health and to understand their vulnerability to
There are some limitations to be noted in this article. This HIV infection (Gómez-Olivé et al., 2020; Gott et al., 2004).
analysis used data from a cross-sectional study that selected
the sample using a convenient approach due to resource ORCID iDs
constraints, making the sample unrepresentative. Hence,
findings of this study are not generalisable to either the Makandwe Nyirenda – https://orcid.org/0000-0002-1839-877X
South African or the sub-Saharan African population of older Paul Kowal – https://orcid.org/0000-0002-6314-8753
adults. The cross-sectional nature of the data also means Joel Negin – https://orcid.org/0000-0002-2016-311X
that no directional causality can be established between
perceived HIV risk and the exposure factors. We cannot, for References
example, determine whether an individual’s sense of high
HIV risk is because of practising risky sexual behaviour, or Abbai, N. S., Nyirenda, M., Reddy, T., & Ramjee, G. (2018). Good
whether they are engaging in risky behaviour because of a correlation between the Afinion AS100 analyser and the ABX Pentra
low HIV risk perception. Furthermore, although interviewers 400 analyser for the measurement of glycosylated haemoglobin and
were trained to ask the questions accurately and elicit further lipid levels in older adults in Durban, South Africa. South African
Medical Journal, 108(1), 50–55. https://doi.org/10.7196/SAMJ.2017.
explanation where required, there may still have been some
v108i1.12548
ambiguity on respondents’ part in the understanding and Aboderin, I. (2014). Sexual and reproductive health and rights of older
evaluation of HIV risk. Another potential source of bias in men and women: Addressing a policy blind spot. Reproductive
the findings of this study is prior knowledge of HIV status Health Matters, 22(44), 185–190. https://doi.org/10.1016/
influencing how individuals perceived their vulnerability to S0968-8080(14)44814-6
HIV infection. To control for this, we conducted additional Aboderin, I., & Beard, J. R. (2015). Older people’s health in
analyses where we excluded all individuals with prior sub-Saharan Africa. Lancet, 385(9968), e9–e11. https://doi.
knowledge of their HIV status. Findings of these additional org/10.1016/S0140-6736(14)61602-0
analyses showed little difference with analyses of the entire Autenrieth, C. S., Beck, E. J., Stelzle, D., Mallouris, C., Mahy, M., &
sample: the majority perceived themselves as at no or little Ghys, P. (2018). Global and regional trends of people living with
HIV aged 50 and over: Estimates and projections for 2000-2020.
risk of HIV; and those who perceived themselves as at
PLoS One, 13(11), e0207005. https://doi.org/10.1371/journal.
high risk were more likely to be sexually active and to test pone.0207005
HIV-positive. This bias is thus likely to have had little impact Bendavid, E., Ford, N., & Mills, E. J. (2012). HIV and Africa’s elderly:
and changes little our interpretation of the findings. The problems and possibilities. AIDS (London, England), 26, S85–
S91. https://doi.org/10.1097/QAD.0b013e3283558513
Conclusion Catania, J. A., Kegeles, S. M., & Coates, T. J. (1990). Towards an
understanding of risk behavior: An AIDS risk reduction model
This article thus makes an important contribution to the (ARRM). Health Education Quarterly, 17(1), 53–72. https://doi.
knowledge on perception of HIV risk and how that is org/10.1177/109019819001700107
associated with sexual activity and HIV status. The fact Chawla, N., & Sarkar, S. (2019). Defining “High-risk Sexual Behavior”
in the Context of Substance Use. Journal of Psychosexual Health,
that individuals who perceived themselves as at high risk
1(1), 26–31. https://doi.org/10.1177/2631831818822015
were more likely to be HIV-positive means there is need Chepngeno-Langat, G., & Hosegood, V. (2012). Older people and
for interventions to assist older adults have an accurate sexuality: Double jeopardy of ageism and sexism in youth-dominated
understanding of their vulnerability to HIV infection and for societies. Agenda (Durban, South Africa), 26(4), 93–99. https://doi.
some targeted interventions for those who rate themselves org/10.1080/10130950.2012.757864
as at high risk of HIV. A persistent absence of targeted Chowdhury, R., Sharot, T., Wolfe, T., Düzel, E., & Dolan, R.
HIV prevention programmes for older adults increases J. (2014). Optimistic update bias increases in older age.
the vulnerability of this population group to HIV infection Psychological Medicine, 44(9), 2003–2012. https://doi.org/10.1017/
(Chepngeno-Langat & Hosegood, 2012; Heidari, 2016; S0033291713002602
Minichiello et al., 2012).
Further research on this neglected topic that could
inform the development of a simple risk-assessment tool
African Journal of AIDS Research 2022, 21(3): 277–286 285

Collins, L. F., & Armstrong, W. S. (2020). What it means to age Negin, J., Geddes, L., Brennan-Ing, M., Kuteesa, M., Karpiak, S., &
with HIV Infection: Years gained are not comorbidity free. Seeley, J. (2016). Sexual Behavior of Older Adults Living with HIV
JAMA Network Open, 3(6), e208023. https://doi.org/10.1001/ in Uganda. Archives of Sexual Behavior, 45(2), 441–449. https://doi.
jamanetworkopen.2020.8023 org/10.1007/s10508-015-0582-5
Dellar, R. C., Dlamini, S., & Abdool Karim, Q. (2015). Adolescent Negin, J., Martiniuk, A., Cumming, R. G., Naidoo, N., Phaswana-
girls and young women: Key populations for HIV epidemic control. Mafuya, N., Madurai, L., Williams, S. & Kowal, P. (2012). Prevalence
Journal of the International AIDS Society, 18(2 Suppl 1), 19408. of HIV and chronic comorbidities among older adults. AIDS
https://doi.org/10.7448/IAS.18.2.19408. (London, England), 26(Suppl 1), S55–S63. https://doi.org/10.1097/
Freeman, E., & Anglewicz, P. (2012). HIV prevalence and sexual QAD.0b013e3283558459
behaviour at older ages in rural Malawi. International Journal of STD Negin, J., Nemser, B., Cumming, R., Lelerai, E., Ben Amor, Y., &
& AIDS, 23(7), 490–496. https://doi.org/10.1258/ijsa.2011.011340 Pronyk, P. (2012). HIV Attitudes, Awareness and Testing Among
Ginsberg, T. B., Pomerantz, S. C., & Kramer-Feeley, V. (2005). Older Adults in Africa. AIDS and Behavior, 16(1), 63–68. https://doi.
Sexuality in older adults: Behaviours and preferences. Age and org/10.1007/s10461-011-9994-y
Ageing, 34(5), 475–480. https://doi.org/10.1093/ageing/afi143 Negin, J., Rozea, A., & Martiniuk, A. L. (2014). HIV behavioural
Gómez-Olivé, F. X., Houle, B., Rosenberg, M., Kabudula, C., interventions targeted towards older adults: A systematic
Mojola, S., Rohr, J. K., Clark, S., Angotti, N., Schatz, E., Kahn, K., review. BMC Public Health, 14, Article 507. https://doi.
Bärnighausen, T., & Menken, J. (2020). Brief Report: HIV Incidence org/10.1186/1471-2458-14-507
Among Older Adults in a Rural South African Setting: 2010-2015. Nicolosi, A., Laumann, E. O., Glasser, D. B., Moreira, E. D., Jr., Paik, A.,
Journal of Acquired Immune Deficiency Syndromes (1999), 85(1), & Gingell, C. (2004). Sexual behavior and sexual dysfunctions after
18–22. doi:10.1097/qai.0000000000002404 age 40: The global study of sexual attitudes and behaviors. Urology,
Gott, M., Galena, E., Hinchliff, S., & Elford, H. (2004). “Opening a can 64(5), 991–997. https://doi.org/10.1016/j.urology.2004.06.055
of worms”: GP and practice nurse barriers to talking about sexual Nyirenda, M., Abbai, N. S., Naidoo, J., & Ramjee, G. (2018). Sexual
health in primary care. Family Practice, 21(5), 528–536. https://doi. activity and healthcare-seeking behaviour: A cross-sectional study
org/10.1093/fampra/cmh509 of older adults in an HIV-endemic South African setting. Southern
Heidari, S. (2016). Sexuality and older people: A neglected issue. African Journal of Demography, 18(1), 7–57. https://www.jstor.org/
Reproductive Health Matters, 24(48), 1–5. https://doi.org/10.1016/j. stable/90022347
rhm.2016.11.011 O’Sullivan, O. P. (2015). The neural basis of always looking on the
Houle, B., Yu, S.-T., Angotti, N., Schatz, E., Kabudula, C. W., bright side. Dialogues in Philosophy, Mental and Neuro Sciences,
Gómez-Olivé, F. X., Clark, S. J., Menken, J., & Mojola, S. A. (2020). 8(1), 11–15. https://www.crossingdialogues.com/Ms-A14-09.htm
Clusters of HIV risk and protective sexual behaviors in Agincourt, Odimegwu, C. O., & Mutanda, N. (2017). Covariates of high-risk
rural South Africa: Findings from the Ha Nakekela Population-Based sexual behaviour of men aged 50 years and above in sub-Saharan
Study of Ages 15 and Older. Archives of Sexual Behavior, 49(6), Africa. Journal of Social Aspects of HIV/AIDS Research Alliance,
2057–2068. https://doi.org/10.1007/s10508-020-01663-5 14(1), 162–170. https://doi.org/10.1080/17290376.2017.1392340
HSRC. (2018). The Fifth South African National HIV Prevalence, Pilowsky, D. J., & Wu, L. T. (2015). Sexual risk behaviors and HIV
Incidence, Behaviour and Communication Survey, 2017 (SABSSM risk among Americans aged 50 years or older: A review. Substance
V). http://www.hsrc.ac.za/uploads/pageContent/9234/SABSSMV_ Abuse and Rehabilitation, 6, 51–60. https://doi.org/10.2147/SAR.
Impact_Assessment_Summary_ZA_ADS_cleared_PDFA4.pdf S78808
Knight, L., Mukumbang, F. C., & Schatz, E. (2018). Behavioral and Prata, N., Morris, L., Mazive, E., Vahidnia, F., & Stehr, M. (2006).
cognitive interventions to improve treatment adherence and access Relationship between HIV risk perception and condom use:
to HIV care among older adults in sub-Saharan Africa: An updated Evidence from a population-based survey in Mozambique.
systematic review. Systematic Reviews, 7(1), Article 114. https://doi. International Family Planning Perspectives, 32(4), 192–200. https://
org/10.1186/s13643-018-0759-9 doi.org/10.1363/3219206
Knight, L., Schatz, E., & Mukumbang, F. C. (2018). “I attend at Vanguard Rosenberg, M. S., Gomez-Olive, X. F., Rohr, J. K., Houle, B. C.,
and I attend here as well”: Barriers to accessing healthcare services Kabudula, C. W., Wagner, R. G., . . . Bärnighausen, T. W. (2017).
among older South Africans with HIV and non-communicable Sexual behaviors and HIV status: a population-based study
diseases. International Journal for Equity in Health, 17(1), Article among older adults in rural South Africa. Journal of Acquired
147. https://doi.org/10.1186/s12939-018-0863-4 Immune Deficiency Syndromes (1999), 74(1), e9–e17. https://doi.
Lekalakala-Mokgele, E. (2014). Understanding of the risk of HIV org/10.1097/QAI.0000000000001173
infection among the elderly in Ga-Rankuwa, South Africa. Journal Savasta, A. M. (2004). HIV: Associated transmission risks in older
of Social Aspects of HIV/AIDS Research Alliance, 11(1), 67–75. adults—an integrative review of the literature. The Journal of the
https://doi.org/10.1080/17290376.2014.931816 Association of Nurses in AIDS Care, 15(1), 50–59. https://doi.
Minichiello, V., Rahman, S., Hawkes, G., & Pitts, M. (2012). STI org/10.1177/1055329003252051
epidemiology in the global older population: Emerging challenges. Sharot, T. (2011). The optimism bias. Current Biology, 21(23), R941–
Perspectives in Public Health, 132(4), 178–181. https://doi. R945. https://doi.org/10.1016/j.cub.2011.10.030
org/10.1177/1757913912445688 Soomro, N., Fitzgerald, G., Seeley, J., Schatz, E., Nachega, J. B., &
Mutevedzi, P. C., & Newell, M. L. (2011). A missing piece in the puzzle: Negin, J. (2019). Comparison of antiretroviral therapy adherence
HIV in mature adults in sub-Saharan Africa. Future Virology, 6(6), among HIV-infected older adults with younger adults in Africa:
755–767. https://doi.org/10.2217/fvl.11.43 Systematic review and meta-analysis. AIDS and Behavior, 23(2),
Nachega, J. B., Hsu, A. J., Uthman, O. A., Spinewine, A., & Pham, P. A. 445–458. https://doi.org/10.1007/s10461-018-2196-0
(2012). Antiretroviral therapy adherence and drug-drug interactions StataCorp. (2014). Stata: Release 14. Statistical Software (Version
in the aging HIV population. AIDS (London, England), 26(Suppl 1), 13.1). StataCorp LP.
S39–S53. https://doi.org/10.1097/QAD.0b013e32835584ea Syme, M. L., Cohn, T. J., & Barnack-Tavlaris, J. (2017). A Comparison
Naicker, N., Kharsany, A. B., Werner, L., van Loggerenberg, F., of Actual and Perceived Sexual Risk Among Older Adults. Journal
Mlisana, K., Garrett, N., & Abdool Karim, S. S. (2015). Risk factors of Sex Research, 54(2), 149–160. https://doi.org/10.1080/0022449
for HIV acquisition in high risk women in a generalised epidemic 9.2015.1124379
setting. AIDS and Behavior, 19(7), 1305–1316. https://doi.
org/10.1007/s10461-015-1002-5
286 Nyirenda, Mnqonywa, Tutshana, Naidoo, Kowal & Negin

Taylor, T. N., Munoz-Plaza, C. E., Goparaju, L., Martinez, O., Holman, UNAIDS. (2021). 2021 UNAIDS global AIDS update: confronting
S., Minkoff, H. L., Karpiak, S. E., Gandh, M., Cohen, M. H., Golub, inequalities – lessons for pandemic responses from 40
E. T., Levine, A. M., Adedime, A. A., Gonsalves, R., Bryan, T., years of AIDS. https://www.unaids.org/en/resources/
Connors, N., Schechter, G., & Wilson, T. E. (2016). “The pleasure is documents/2021/2021-global-aids-update
better as I’ve gotten older”: Sexual health, sexuality, and sexual risk United Nations. (2013). HIV and aging: A special
behaviors among older women living with HIV. Archives of Sexual supplement to the UNAIDS report on the global AIDS
Behavior. Advance online publication. https://doi.org/10.1007/ epidemic 2013. https://www.unaids.org/en/resources/
s10508-016-0751-1 documents/2013/20131101_JC2563_hiv-and-aging
Todd, J., Cremin, I., McGrath, N., Bwanika, J. B., Wringe, A., Marston, van Empel, E., de Vlieg, R. A., Montana, L., Gómez-Olivé, F. X.,
M., Kasamba, I., Mushati, P., Lutalo, T., Hosegood, V., & Zaba, Kahn, K., Tollman, S., Berkman, L., Bärnighausen, T. W., &
B. (2009). Reported number of sexual partners: Comparison of Manne-Goehler, J. (2021). Older adults vastly overestimate both
data from four African longitudinal studies. Sexually Transmitted HIV acquisition risk and hiv prevalence in rural South Africa.
Infections, 85(Suppl 1), i72–i80. https://doi.org/10.1136/ Archives of Sexual Behavior, 50(7), 3257–3276. https://doi.
sti.2008.033985 org/10.1007/s10508-021-01982-1
Trompeter, S. E., Bettencourt, R., & Barrett-Connor, E. (2012). sexual Wang, V., Depp, C. A., Ceglowski, J., Thompson, W. K., Rock, D.,
activity and satisfaction in healthy community-dwelling older women. & Jeste, D. V. (2015). Sexual health and function in later life: A
The American Journal of Medicine, 125(1), 37–43.e31. https://doi. population-based study of 606 older adults with a partner. The
org/10.1016/j.amjmed.2011.07.036 American Journal of Geriatric Psychiatry, 23(3), 227–233. https://
UNAIDS. (2014a). 90-90-90: An ambitious treatment target to help doi.org/10.1016/j.jagp.2014.03.006
end the AIDS epidemic. Retrieved from http://www.unaids.org/sites/ World Medical Association. (2013). World Medical Association
default/files/media_asset/90-90-90_en_0.pdf Declaration of Helsinki: Ethical Principles for Medical Research
UNAIDS. (2014b). The Gap Report: Beginning of the end of the AIDS Involving Human Subjects. Journal of the American Medical
epidemic. http://www.unaids.org/sites/default/files/media_asset/ Association, 310(20), 2191–2194. https://doi.org/10.1001/
UNAIDS_Gap_report_en.pdf jama.2013.281053
UNAIDS. (2016). Get on the Fast-Track: the life-cycle approach
to HIV. https://www.unaids.org/en/resources/campaigns/
get-on-the-fast-track

Manuscript submitted August 2021/ revised February 2022/ accepted June 2022