Food insecurity , HIV status and prior testing at South African primary healthcare clinics

South African Journal of Science http://www.sajs.co.za Volume 114 | Number 9/10 September/October 2018 © 2018. The Author(s). Published under a Creative Commons Attribution Licence. Food insecurity, HIV status and prior testing at South African primary healthcare clinics AUTHORS: Makandwe Nyirenda1 Renee Street1,2 Tarylee Reddy3 Susie Hoffman4,5 Suraya Dawad1,6 Kelly Blanchard7 Theresa M. Exner5 Elizabeth A. Kelvin8,9 Joanne E. Mantell5 Gita Ramjee1

HIV and food insecurity are two prominent causes of morbidity and mortality in sub-Saharan Africa.Food insecurity has been associated with risky sexual practices and poor access to healthcare services.We describe the association between household food insecurity and previous HIV testing and HIV status.We used logistic regression to analyse the association between food insecurity and prior HIV counselling and testing (HCT) and testing HIV positive.A total of 2742 adults who presented for HCT at three primary healthcare clinics in KwaZulu-Natal, South Africa, participated in the study.The prevalence of household food insecurity was 35%.The prevalence of food insecurity was highest in adults who had incomplete high schooling (43%), were unemployed (39%), and whose primary source of income was government grants (50%).Individuals who were food insecure had significantly higher odds of testing HIV positive (adjusted odds ratio 1.41, 95% CI 1.16-1.71),adjusted for demographic and socio-economic variables.There was no association between food insecurity and prior HCT.The findings of this study highlight the important role food insecurity may play in HIV risk.Interventions to turn food-insecure into food-secure households are needed to reduce their household members' vulnerability to HIV acquisition.The absence of such interventions is likely to severely impact ambitious global targets of ending AIDS by 2030 through the 90-90-90 targets and test-and-treat-all initiatives.

Significance:
• One in three adults presenting for HIV counselling and testing came from households with some degree of food insufficiency.
• Experience of food insecurity was very high in young people who did not complete high school and were currently not studying.
• Findings support the need for socio-economic and structural interventions to transform food-insecure into food-secure households.
• Failure or lack of such interventions will contribute to the failure to achieve global targets like the UNAIDS 90-90-90 programme.

Introduction
To end AIDS as a global pandemic by 2030, UNAIDS has set the world ambitious targets of 90-90-90. 1That is, by 2020, 90% of all persons living with HIV will be tested and know their status; of those tested positive, 90% will be on antiretroviral treatment; and finally, 90% of those on antiretroviral treatment will achieve viral suppression.HIV counselling and testing (HCT) is thus the initial step in the cascade of HIV prevention, care and treatment towards ending HIV as a public health pandemic. 1,2A number of factors have been identified as contributory to the low uptake of HIV testing services including hunger, poor nutrition and food insecurity. 3In South Africa, the recent National Health and Nutrition Examination Survey revealed that 28% of the population are at risk of hunger and 26% are household food insecure (ever experienced hunger). 41][12] In South Africa, one of the main reasons cited for not accepting free antiretroviral drugs was fear of taking medication on an empty stomach as a consequence of food insufficiency. 13,14In Tanzania, supplementary food cost was cited as an access challenge in sustaining long-term antiretroviral treatment. 15od insecurity also has been suggested to be a major contributor to risky sexual behaviour and HIV spread in neighbouring Botswana and Swaziland 16 and elsewhere.Those who report food insecurity have been found to be less likely to use condoms than those who report food security. 17,18ch less research has focused on the effect of food insecurity on health-seeking behaviour including HIV testing.
To date, there is no information on the role of household food insecurity in HIV testing history and/or testing HIV positive among South Africans.
To address this gap, the aim of this study was to determine the association of household food insecurity with prior HIV testing and HIV serostatus among individuals presenting for HCT at public primary healthcare clinics in KwaZulu-Natal, South Africa.We hypothesised that household food insecurity would be associated with lower likelihood to have previously tested for HIV and higher likelihood to be HIV positive when tested.

Study population
Data for these cross-sectional analyses were drawn from interviews conducted between November 2010 and May 2012 as part of the Pathways to Care study, a prospective cohort study of newly diagnosed HIV-positive

Confounders
Informed by available empirical literature, 17,20,21 potential confounders controlled for in the analysis of HIV status included age categorised into age ranges (<24, 25-28, 29-31, 32-36, ≥37 years), gender, clinic site, educational attainment (completed high school or not), relationship status (not married or in a relationship; married or in a relationship, not living together; married or in a relationship, and living together), employment status (employed part-time/full-time/self-employed or unemployed), primary source of income (none, family, employment, spouse/ partner, government grant, or other), caring for dependent adults and/or children (none, children only, adults only, both), ability to borrow money (up to ZAR100) if needed for medical expenses (yes/no), and time taken to travel to the clinic (<30 min, 30-60 min, >60 min).

Statistical analysis
Chi-square tests were used to evaluate the association between each demographic and socio-economic factor and household food insecurity.The relationship between household food insecurity and previous HIV testing and HIV status was examined using univariable and multivariable logistic regression models.A variable was deemed a potential confounder if its inclusion in the model containing food security resulted in a 10% or higher change in the model coefficient for food security.

Results
The majority of the participants in this analysis were women (52.8%), were aged less than 24 years (36.5%), had not completed high school (61.5%), were not currently studying (87.4%) and were married or in a relationship but not living together (62.6%) (Table 1).Further characteristics of the sample were that most were unemployed (65.1%) but were currently caring for a dependent adult or child (58.5%).Nearly two-thirds of participants came from households with at least one government grant recipient (63.1%).Nearly all participants (96.4%) travelled less than 1 hour to get to the primary healthcare clinic.
The overall prevalence of food insecurity in this sample was 35.3% (Table 1).There were no statistically significant gender differences in reported household food insecurity (p=0.989).Prevalence of food insecurity increased with age and was highest among adults older than 37 years (45.5%, p<0.001).Household food insecurity was also more prevalent in adults who did not complete high school (42.4%); were currently not in school (37.1%) compared with those who had completed high school (24.0%); and individuals who were married or in a co-habiting relationship (27.3%) compared with those who were married or in a relationship but not co-habiting (22.4%).

HOW TO CITE:
Nyirenda Table 2 shows the unadjusted odds ratio (OR) of ever having been tested associated with food insecurity and the OR adjusting for confounding variables (gender, age, relationship status, employment status, primary source of income, financially supporting other adults or children, and travel time to clinic).We found that household food insecurity was not significantly associated with history of previous HIV testing (OR=0.95,95% confidence interval (CI)=0.80-1.12,p=0.520).
Table 2 also presents the unadjusted and adjusted ORs for the association between household food insecurity and testing HIV positive.In both the unadjusted (OR=1.70,95%CI=1.43-2.01)and adjusted models (OR=1.46,95%CI=1.22-1.75),individuals who had ever experienced household food insecurity had higher odds of testing HIV positive when they presented for HCT at primary healthcare facilities.

Discussion
As UNAIDS has set ambitious global targets of 90-90-90 to be reached by 2020 in order to end AIDS as a global pandemic by 2030, achieving or exceeding the first 90% (that is, 90% of all adults knowing their HIV status) is key to ending the epidemic given the expected fallout along the cascade steps from testing to treatment and viral suppression. 22,23We examined the links between household food insecurity and previous HIV testing and testing HIV positive to ascertain if food insecurity is a barrier to HIV testing or a risk factor for HIV infection.We found that 35.3% of study participants came from households with some degree of food insufficiency.Although there was no association between household food insecurity and previous HIV testing, individuals who were food insecure had significantly higher odds of testing HIV positive [OR 1.47 (1.22-1.77)].
Our study revealed that 29.1% of those reporting household food insecurity were young adults (<24 years old).The incidence of HIV in South Africa is highest among young women below age 25. [24][25][26] The majority (61.9%) of participants in our study had not completed high school.Experience of food insecurity was higher among participants who had not completed high school compared to those who had completed high school (42.4% versus 24.0%, respectively).Weiser et al. 16 found that educational level did not significantly influence the association between food insecurity and risky sexual behaviour.However, De Swardt et al. 27 found a positive relationship between educational level and strategies to survive food crises among South African households.Higher educational level was shown to be associated with greater ability to survive a food crisis. 27Elsewhere it has further been shown that completing secondary education helps to protect against HIV acquisition, more so among young girls. 18,28V and food insecurity are leading causes of morbidity and mortality in sub-Saharan Africa.Eradication of both HIV and food insecurity is central to the Sustainable Development Goals. 29 The first three of these goals are to: (1) end poverty of every form; (2) end hunger, ensure food security and better nutrition through sustainable farming; and (3) ensure health and well-being for all.Despite South Africa willingly consenting to the SDGs, the proportions of at-risk and food-insecure households have remained largely unchanged over the years. 4,30Considering the association between food insecurity and high-risk sexual behaviour that has been demonstrated by others among women in southern Africa 16 , it is a likely contributory factor to the continued high HIV risk among young women and may explain the association we found in this study 17,20,21 .
There are some limitations to this study that should be mentioned.This analysis relied on self-reports of food sufficiency in the household by an individual member of the household.In addition, study participants were interviewed about food security cross-sectionally.Thus, we were unable to determine whether food insecurity 'caused' someone to become HIV positive or whether being HIV positive created a condition of food insecurity.A more robust design would have been to conduct a longitudinal cohort study using food diaries.This approach was, however, not feasible and beyond the scope of the Pathways to Care study.The study was also limited by the use of a single question to assess food insecurity.Finally, we did not have data that would have enabled us to explore sexual risk behaviour as a possible mediator of the association that we found between HIV positive status and food insecurity.
Despite these limitations, this analysis makes an important contribution to the literature.A strength of the analysis is that study participants were newly diagnosed HIV-positive individuals; therefore, we could show the likely role of food insecurity in seeking HIV counselling and testing.Findings of this study, together with other available literature, support the need for socio-economic and structural interventions that will transform households experiencing food insecurity into food-secure households so as to reduce household members' vulnerability to HIV acquisition.This recommendation is consistent with the Sustainable Development Goals.Failure or lack of such prudent interventions is likely to contribute to the failure to achieve the UNAIDS 90-90-90 goals.

Conclusion
Household food insecurity is related to similar risk factors for HIV, such as low education, not attending school, not being married and low socioeconomic status. 25,31,32Although we did not find an association between food insecurity and prior HCT, we found that testing HIV positive was associated with household food insecurity among adults attending HCT at public primary healthcare facilities.These findings highlight the important role that food insecurity plays in HIV risk, and are supported by findings from other sub-Saharan African countries. 16,17Thus, interventions such as targeted food support in poor and food-insecure populations are key to mitigating vulnerability to HIV infection.In South Africa, such an intervention could be implemented with similar means testing, as the child support or old age grants. 33,34If left unaddressed, household food insecurity is likely to adversely affect the global targets of ending AIDS by 2030 through the 90-90-90 and test-and-treat initiatives.

Table 1 :
Description of the study sample in association with food insecurity among adults presenting for HIV counselling and testing at three primary healthcare clinics in KwaZulu-Natal, South Africa, 2010-2012 http://www.sajs.co.za Volume 114 | Number 9/10 September/October 2018

in the household became ill and you needed R100 for treatment, how difficult would it be to get the money?
http://www.sajs.co.za

Table 2 :
Association between household food security and history of previous HIV testing, and testing HIV positive Adjusted for education, employment status, currently studying, primary source of income, supports other adults or children, anyone in household receiving grants, difficulty in obtaining ZAR100 for treatment; n=2717.† † Adjusted for age group, education, relationship status, employment status, currently studying, primary source of income; n=2678.