Share this article on:

A Pilot Study of Food Supplementation to Improve Adherence to Antiretroviral Therapy Among Food-Insecure Adults in Lusaka, Zambia

Cantrell, Ronald A PhD*†; Sinkala, Moses MBChB, MPH; Megazinni, Karen DrPH, PA-C§; Lawson-Marriott, Sibi MAAS; Washington, Sierra MD; Chi, Benjamin H MD, MSc*†; Tambatamba-Chapula, Bushimbwa MBChB, MA#; Levy, Jens PhD*; Stringer, Elizabeth M MD, MSc*†; Mulenga, Lloyd MBChB*; Stringer, Jeffrey S A MD*†

JAIDS Journal of Acquired Immune Deficiency Syndromes: October 2008 - Volume 49 - Issue 2 - p 190-195
doi: 10.1097/QAI.0b013e31818455d2
Epidemiology and Social Science

Background: The provision of food supplementation to food-insecure patients initiating antiretroviral therapy (ART) may improve adherence to medications.

Methods: A home-based adherence support program at 8 government clinics assessed patients for food insecurity. Four clinics provided food supplementation, and 4 acted as controls. The analysis compared adherence (assessed by medication possession ratio), CD4, and weight gain outcomes among food-insecure patients enrolled at the food clinics with those enrolled at the control clinics.

Results: Between May 1, 2004, and March 31, 2005, 636 food- insecure adults were enrolled. Food supplementation was associated with better adherence to therapy. Two hundred fifty-eight of 366 (70%) patients in the food group achieved a medication possession ratio of 95% or greater versus 79 of 166 (48%) among controls (relative risk = 1.5; 95% confidence interval: 1.2 to 1.8). This finding was unchanged after adjustment for sex, age, baseline CD4 count, baseline World Health Organization stage, and baseline hemoglobin. We did not observe a significant effect of food supplementation on weight gain or CD4 cell response.

Conclusions: This analysis suggests that providing food to food-insecure patients initiating ART is feasible and may improve adherence to medication. A large randomized study of the clinical benefits of food supplementation to ART patients is urgently needed to inform international policy.

From the *Centre for Infectious Disease Research in Zambia, Lusaka, Zambia; †School of Medicine and Public Health, University of Alabama, Birmingham, AL; ‡Catholic Medical Missions Board, Lusaka, Zambia; §Westat Inc, Rockville, MD; World Food Programme, Accra, Ghana; ¶School of Medicine, University of California, San Francisco, CA; and #Lusaka Urban District Health Management Team, Lusaka, Zambia.

Received for publication April 2, 2008; accepted June 23, 2008.

Supported by the University of Alabama at Birmingham and the United Nations World Food Programme. Additional support was provided by a Doris Duke Charitable Foundation Award for Operations Research for AIDS Care and Treatment in Africa (2005047). Investigator salary or trainee support was provided by the National Institutes of Health (K01-TW05708; K01-TW06670; D43-TW001035).

Presented in part at the International AIDS Society's XVI International AIDS Conference as abstract MOAB0401, August 13-18, 2006, Toronto, Ontario, Canada.

Correspondence to: Ronald A. Cantrell, MPH, Centre for Infectious Disease Research in Zambia, Plot 1275 Lubutu Road, PO Box 34681, Lusaka, Zambia (e-mail:

Back to Top | Article Outline


HIV disease causes progressive disability and in its more advanced stages can interfere with an individual's ability to obtain food.1,2 In Zambia, one third of the population is “food-insecure”-a condition where people lack consistent access to enough food of sufficient variety to meet dietary needs3-and nearly one fifth is infected with HIV.4 Malnutrition and HIV may interact in a vicious cycle, where malnutrition accelerates immunosuppression and HIV infection worsens malnutrition.5 Because energy demands are increased by 10% -30% in HIV-infected adults,6-9 meeting these caloric needs can be especially challenging in the face of poverty. Poor nutrition may also affect a patient's adherence to antiretroviral therapy (ART) by depriving him or her of the energy to travel to the pharmacy to collect antiretroviral drugs10 or by potentiating drug toxicity.11,12 Evidence has shown that micronutrient supplementation can delay HIV disease progression among adults not yet requiring HIV treatment.13,14

Observations from developing countries link malnutrition-characterized by low body mass index and severe anemia-to an increased risk of mortality among patients initiating ART.15-17 The benefits of providing food supplementation to food-insecure individuals on ART, however, have only been speculated10,18,19 but not tested. The goal of this pilot project was to evaluate the effect of food supplementation on adherence to ART among food-insecure adults initiating ART in Lusaka, Zambia.

Back to Top | Article Outline


In May 2004, the Zambian Ministry of Health implemented HIV care and treatment services in Lusaka.15 Alongside ART service expansion, a number of supporting services were provided, including a home-based adherence support initiative. Linked to the home-based adherence support program, some ART sites received a food donation from the United Nations World Food Programme. The food supplementation pilot program was launched at 4 sites in September 2004 and expanded to 4 additional sites 14 months later. The staggered rollout of this activity offered an opportunity to compare adherence and clinical response among food-insecure patients starting ART with and without available food supplementation.

HIV-infected patients enroll into care at primary centers in Lusaka as part of the Zambian Ministry of Health's national program for HIV care and treatment. The protocol for clinical care in this program has been described elsewhere.15,20 In brief, all patients with a confirmed HIV diagnosis undergo physical examination including weight measurement, World Health Organization (WHO) staging, CD4+ T-lymphocyte (CD4) counting with a Beckman Coulter Epics XL and FlowCARE pan-leukocyte gating (PLG) CD4 reagents (Fullerton, CA), and hemoglobin measurement at their initial visit. ART eligibility is determined according to WHO guidelines21,22 Those starting ART return to the clinic pharmacy each month to collect their drugs. Each pharmacy dispensation includes a 3-day buffer of extra tablets. Clinical visits are performed at weeks 2, 4, 8, and 12, then 3 monthly. A weight measurement, CD4 cell count, and WHO staging are performed every 6 months, more frequently if clinically indicated. Clinical and laboratory data are entered into an electronic medical record system, which also tracks scheduled and attended clinical and pharmacy appointments.23

Patients initiating ART are offered home-based adherence support. This program employs community health workers (CHWs) who visit patients in their homes to monitor and encourage ART adherence. The CHWs begin with daily visits and gradually reduce to monthly visits over a 15-week period. If patients are deemed poorly adherent by the CHW, they are visited more frequently and may be referred to their respective clinics for further assessment and counseling. At the initial home visit, the CHW also performs a household food security assessment, using a standardized instrument that gathers information on income, household size, and food consumption patterns (Table 1).



Food-insecure patients receive monthly rations provided by World Food Programme. The initial commitment is for 6 months of food supplementation, with the option to receive an additional 6 months if the household still meets criteria upon reassessment. At the beginning of the pilot program, patients who were not the primary income earners in their households received an individual ration of micronutrient-fortified corn-soya blend and vegetable oil, whereas primary income earners received the same individual ration along with an additional household ration of micronutrient-fortified corn-soya blend, oil, maize meal, and beans sufficient for 6 household members (Table 1). In July 2005, a programmatic change was made to base the ration size solely upon the number of people living in the household, instead of primary income earner status. Households with fewer than 3 people received an individual ration, whereas households with 3 or more people received a ration sufficient for 6 adults.

This analysis takes advantage of a staggered rollout of the pilot food supplementation program to make its comparisons. We classified patients in the 4 clinics assigned to begin ration distribution in September 2004 as the “food” group and patients in the 4 clinics assigned to begin ration distribution at a later time as the “control” group. (NB: eligible patients in the control group were offered food at a later time, once the pilot program became available at their respective clinics, but that intervention occurred after the outcomes we describe in this report had been measured.)

We limited the data to adult (older than 15 years) patients on ART who were food-insecure and enrolled in the voluntary home-based adherence support program. Adherence to ART was measured during the first 12 months of therapy by timeliness of pharmacy visits using a variation of the medication possession ratio (MPR).24,25 We divided the number of days late for pharmacy refills by total days on therapy in the first year and then subtracted that percentage from 100% to calculate the MPR. We did not count a person as late until after 3 days to account for the extra pills they received. The end product estimates the amount of time that a given patient has medication on hand during the first 12 months of therapy,26,27 an adherence measure that has correlated with good ART outcomes in both North American and African settings.28 For purposes of analysis, we dichotomized the outcome into “adherent” (≥ 95% MPR) and “nonadherent” (<95% MPR).29

Because patients accessed food supplementation based upon the clinic they attended, we adjusted for potential correlation of patient characteristics within clinics using hierarchical logistic regression models (SAS GLIMMIX Procedure) for categorical variables30 and hierarchical linear regression models (SAS MIXED Procedure) for continuous variables,31 with random intercepts for clinic. To calculate relative risk (RR) estimates, we converted the odds ratio estimates from hierarchical logistic regression.32 Analysis of variance was used to calculate adjusted mean changes in weight and CD4 count. All analyses were performed with SAS version 9.1.3 (SAS Institute, Cary, NC). This study was approved by the Institutional Review Boards of the University of Zambia (Lusaka, Zambia) and the University of Alabama at Birmingham (Birmingham, AL).

Back to Top | Article Outline


Between May 1, 2004, and March 31, 2005, 8 participating clinics enrolled 1335 adults into the home-based adherence support program, of whom 636 (48%) met criteria for household food insecurity. These included 442 of 881 (50%) patients at the 4 participating food clinics and 194 of 454 (43%) patients at the 4 control clinics. At 12 months post- ART initiation, 532 patients (84%) remained alive and in the program. The distribution of alive, dead, and withdrawn or lost at 12 months was not significantly different across groups. In the food group, 366 (83%) were alive, 28 (6%) were dead, and 48 (11%) had withdrawn or were lost compared with 166 (85%), 15 (8%), and 13 (7%), respectively, in the control group (P = 0.23).

Back to Top | Article Outline

Baseline Characteristics

The 636 food-insecure adults were a median of 35 years of age (interquartile range [IQR]: 31, 42), and 414 (65%) were females. The mean (SD) weight of 411 female patients and 218 male patients with a baseline measure was 52.9 (9.9) and 56.1 (7.9) kg, respectively. The mean CD4 cell count (SD) among 614 patients with a baseline measure was 131 cells (107) per cubic millimeter, with 134 (22%) having fewer than 50 CD4 cells per cubic millimeter. Of 605 patients with baseline hemoglobin measured, the mean (SD) concentration was 10.7 (2.1) g/dL, with 59 (10%) having a concentration <8.0 g/dL.

Patients in the food group received a median of 9 monthly rations (IQR: 7, 10). Compared with controls, patients in the food group were somewhat more likely to be severely immunosuppressed but less likely to be WHO stage III or IV. Otherwise, the 2 groups did not differ according to other measured baseline characteristics: age, sex, hemoglobin concentration, weight, or presence of active tuberculosis (Table 2).



Back to Top | Article Outline

Adherence to ART

Adherence to ART was higher among patients in the food group compared with controls. Two hundred fifty-eight of 366 (70%) patients in the food group achieved an MPR of 95% or greater versus 79 of 166 (48%) among controls [RR = 1.5; 95% confidence interval (CI): 1.2 to 1.8]. This finding remained consistent after adjustment for sex, age, baseline CD4 cell count, baseline WHO stage, and baseline hemoglobin concentration (adjusted RR = 1.5; 95% CI: 1.2 to 1.8; Fig. 1 and Table 3).





Back to Top | Article Outline

Clinical Response

We did not observe a significant difference in weight gain or CD4 cell response between individuals in each group. The mean increase in weight adjusted for age, sex, baseline CD4 count, WHO stage, and hemoglobin among the 329 patients in the food group with data available was 5.4 kg (95% CI: 4.7 to 6.1) at 6 months versus 5.1 kg (95% CI: 4.0 to 6.2) among 131 patients in the control group (P = 0.68). At 12 months, the mean increase in weight adjusted for the same factors among the 302 patients in the food group with data available was 6.3 kg (95% CI: 5.5 to 7.2) versus 5.4 kg (95% CI: 4.0 to 6.8) among 113 patients in the control group (P = 0.34). This pattern persisted when stratifying by sex, with men trending toward more adjusted weight gain at 12 months, 5.5 kg (95% CI: 4.3 to 6.7) among 105 men in the food group with data available versus 3.2 kg (95% CI: 1.1 to 5.3) among 35 men in the control group (P = 0.12) (Table 4).



The mean increase in CD4 cell count adjusted for age, sex, baseline CD4 count, WHO stage, and hemoglobin among the 279 patients in the food group with data available was 154 cells per cubic millimeter (95% CI: 136 to 172) at 6 months versus 171 cells per cubic millimeter (95% CI: 144 to 198) among 125 patients in the control group (P = 0.50). At 12 months, the mean adjusted increase in CD4 cell count among the 241 patients in the food group with data available was 182 cells per cubic millimeter (95% CI: 160 to 204) versus 180 cells per cubic millimeter (95% CI: 147 to 214) among 105 patients in the control group (P = 0.96).

Back to Top | Article Outline


This analysis suggests that providing food supplements to food-insecure patients initiating ART improves adherence. Patients in the food group received a median of 9 monthly rations and were 50% more likely than controls to be adherent, defined here as possessing their medication on at least 95% of days during the first year of therapy. Although we did not observe a statistically significant effect of food supplementation on the clinical outcomes of weight and CD4 cell response, there is a trend toward modest benefit, particularly among men.

This pilot study demonstrated the feasibility of food supplementation in the context of a home-based adherence support program and assessed the prevalence of food insecurity among patients starting ART. That 48% of participants met criteria for food insecurity is concerning and, if unaddressed, could pose a serious threat to the long-term success of ART programs in impoverished settings such as ours. We have previously reported from the Lusaka cohort that low body mass index among adults15 and low weight-for-age Z scores among children20 are associated with poor ART survival. This information, coupled with the known relationship between adherence and good long-term ART outcomes, argues strongly for nutrition as an essential component of integrated HIV care and treatment.

A weakness of this study is its small patient numbers and limited follow-up time. This is especially evident in the observed clinical effects. Our available sample size and clustered design have only 19% power to detect a 1-kg increase in weight between groups, and the smallest weight gain we would have 80% power to detect would be 2.9 kg (assuming alpha = 0.05 and intraclass correlation coefficient = 0.01).33,34 Additionally, the provision of ART without food supplementation is known in this same population to increase weight and CD4 cell count,12 so isolating the incremental effect attributable to food is difficult in this small nonrandomized analysis. Unmeasured confounding is another limitation as we did not record information on access to and consumption of other sources of food. Finally, it is possible that we lack enough follow-up time to observe the full effect of the food intervention, particularly regarding adherence to therapy, where there is known to be considerable delay between poor adherence and failure of the antiretroviral regimens. A definitive study would need to be much larger, have longer patient follow-up, and include virologic suppression as a primary outcome.

These results indicate us that food insecurity is an important reality for many of our ART patients and that it extends not only to the individual but also into their entire household. Over the first 10 months of the pilot program (approximately half the study period), we provided an individual ration to participants who were not the primary income earners in their household and a larger household ration to participants who were primary income earners. As the study proceeded, it became evident to the team that this distinction was not particularly helpful and that there was considerable sharing of allocated rations among all people living with the primary beneficiary. For this reason, a modification was made to provide a more substantial ration to all (although in neither case was more than 68% recommended daily energy allowance provided; Table 1).

The issue of food sharing, we believe, can explain some of our study findings. Because food is such a valuable commodity, our participants were highly motivated to come to the clinic on time to collect their ration and their medicine (thus, the higher observed adherence). However, if sharing was as common as we believe, it is possible that each patient did not get enough protein energy to produce additional weight gain relative to those who did not get food. It should be noted that these results were achieved with a locally available, fortified corn-soya blend. The use of a more calorically dense product, such as ready-to- use therapeutic food or “Plumpy Nut,” might have produced different results. Although more expensive than corn-soya blend, ready- to- use therapeutic food can be packaged in small containers, labeled with the patient's name, and even dispensed by the clinic pharmacy. Such “food by prescription” might mitigate to some extent the degree of household sharing and ensure that the ART patient receives an adequate supplement.

Although this study does not definitely demonstrate a direct clinical benefit of food supplementation on ART outcomes, it does suggest that providing it will substantially improve adherence. It also shows that food supplementation with commodities that are cheap and easily available locally is feasible as part of a comprehensive adherence support program and can be evaluated in a rigorous way. We believe that food- insecure people should have food, irrespective of whether it improves measurable ART outcomes. However, there has historically been reluctance among donors to allow funding earmarked for HIV/AIDS treatment to go toward procurement of food. A well-designed, randomized study that demonstrated a clear clinical benefit of food supplementation would provide a powerful evidence base from which to advocate for food assistance to become allowable within the context of large medical aid programs, such as those funded by the US President's Emergency Plan for AIDS Relief or the Global Fund for AIDS, Tuberculosis, and Malaria. Such a study seems an urgent international priority.

Back to Top | Article Outline


The authors wish to acknowledge and thank Debra Pankratz [Centre for Infectious Disease Research in Zambia (CIDRZ)], Carolyn Bolton, (CIDRZ), Mary Morris, (CIDRZ), Yvonne Mulenga (Project Concern International), Dan Krebs (University of Chicago School of Medicine), Tom Ventimiglia (Family Health International, Tanzania), and David Stevenson (World Food Programme, Rome) for their contributions to the program described herein.

Back to Top | Article Outline


1. Piot P, Pinstrup-Andersen P. AIDS: The New Challenge to Food Security. 2001-2002. IFPRI Annual Report Essay. Washington, DC: IFPRI; 2002.
2. Bukusuba J, Kikafunda JK, Whitehead RG. Food security status in households of people living with HIV/AIDS (PLWHA) in a Ugandan urban setting. Br J Nutr. 2007;98:211-217.
3. Food and Agriculture Organization of the United Nations. The State of Food Insecurity in the World, 2002. Available at: Accessed November 11, 2007.
4. UNAIDS. Zambia Country Profile. Available at: Accessed December 5, 2007.
5. Semba RD, Tang AM. Micronutrients and the pathogenesis of human immunodeficiency virus infection. Br J Nutr. 1999;81:181-189.
6. Hommes MJ, Romijn JA, Godfried MH, et al. Increased resting energy expenditure in human immunodeficiency virus-infected men. Metabolism. 1990;39:1186-1190.
7. Grunfeld C, Pang M, Shimizu L, et al. Resting energy expenditure, caloric intake, and short-term weight change in human immunodeficiency virus infection and the acquired immunodeficiency syndrome. Am J Clin Nutr. 1992;55:455-460.
8. Melchior JC, Salmon D, Rigaud D, et al. Resting energy expenditure is increased in stable, malnourished HIV-infected patients. Am J Clin Nutr. 1991;53:437-441.
9. Shevitz AH, Knox TA, Spiegelman D, et al. Elevated resting energy expenditure among HIV-seropositive persons receiving highly active antiretroviral therapy. AIDS. 1999;13:1351-1357.
10. Marston B, De Cock KM. Multivitamins, nutrition, and antiretroviral therapy for HIV disease in Africa. N Engl J Med. 2004;351:78-80.
11. Casey KM. Malnutrition associated with HIV/AIDS. Part one: definition and scope, epidemiology, and pathophysiology. J Assoc Nurses AIDS Care. 1997;8:24-32.
12. Salomom J, Detruchis P, Melchior JC. Nutrition and HIV infection. Br J Nutr. 2002;87:111-119.
13. Fawzi W, Msamanga GI, Spiegelman D, et al. A randomized trial of multivitamin supplements and HIV disease progression and mortality. N Engl J Med. 2004;351:23-32.
14. Fawzi W, Msamanga G, Spiegelman D, et al. Studies of vitamins and minerals and HIV transmission and disease progression. J Nutr. 2005;135:938-944.
15. Stringer JS, Zulu I, Levy J, et al. Rapid scale-up of antiretroviral therapy at primary care sites in Zambia: feasibility and early outcomes. JAMA. 2006;296:782-793.
16. Zachariah R, Fitzgerald M, Massaquoi M, et al. Risk factors for high early mortality in patients on antiretroviral treatment in a rural district of Malawi. AIDS. 2006;20:2355-2360.
17. Calmy A, Pinoges L, Szumilin E, et al. Generic fixed-dose combination antiretroviral treatment in resource-poor settings: multicentric observational cohort. AIDS. 2006;20:1163-1169.
18. Mahlungulu S, Grobler LA, Visser ME, et al. Nutritional interventions for reducing morbidity and mortality in people with HIV. Cochrane Database Syst Rev. 2007;CD004536.
19. Piwoz E. Nutrition and HIV/AIDS: evidence, gaps, and priority actions. 2004. The Support for Analysis and Research in Africa (SARA) project. Available at: Accessed December 5, 2007.
20. Bolton-Moore C, Mubiana-Mbewe M, Cantrell RA, et al. Clinical outcomes and CD4 cell response in children receiving antiretroviral therapy at primary health care facilities in Zambia. JAMA. 2007;298:1888-1899.
21. World Health Organization. Scaling Up Antiretroviral Therapy in Resource-Limited Settings: Treatment Guidelines for a Public Health Approach. 2003 Revision. Geneva, Switzerland: WHO Press; 2003.
22. World Health Organization. Antiretroviral Therapy for HIV Infection in Adults and Adolescents in Resource-Limited Settings, Towards Universal Access: Recommendations for a Public Health Approach (2006 version). Geneva, Switzerland: WHO Press; 2006.
23. Fusco H, Hubschman T, Mweeta V, et al. Electronic patient tracking supports rapid expansion of HIV care and treatment in resource-constrained settings. Presented at 3rd IAS Conference on HIV Pathogenesis and Treatment; 2005. Rio de Janiero, Brazil.
24. Sikka R, Xia F, Aubert RE. Estimating medication persistency using administrative claims data. Am J Manag Care. 2005;11:449-457.
25. Dezii CM. Persistence with drug therapy: a practical approach using administrative claims data. Manag Care. 2001;10:42-45.
26. Nachega JB, Hislop M, Dowdy DW, et al. Adherence to highly active antiretroviral therapy assessed by pharmacy claims predicts survival in HIV-infected South African adults. J Acquir Immune Defic Syndr. 2006;43:78-84.
27. Orrell C, Bangsberg DR, Badri M, et al. Adherence is not a barrier to successful antiretroviral therapy in South Africa. AIDS. 2003;17:1369-1375.
28. Mills EJ, Nachega JB, Buchan I, et al. Adherence to antiretroviral therapy in sub-Saharan Africa and North America: a meta-analysis. JAMA. 2006;296:679-690.
29. Weidle PJ, Wamai N, Solberg P, et al. Adherence to antiretroviral therapy in a home-based AIDS care programme in rural Uganda. Lancet. 2006;368:1587-1594.
30. Brown H, Prescott R. Applied Mixed Models in Medicine. New York, NY: John Wiley & Sons; 2006.
31. Singer J. Using SAS PROC MIXED to fit multilevel models, hierarchical models, and individual growth models. J Educ Behav Stat. 1998;23:323-355.
32. Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280:1690-1691.
33. Donner A, Klar N. Statistical considerations in the design and analysis of community intervention trials. J Clin Epidemiol. 1996;49:435-439.
34. Donner A, Klar N. Design and Analysis of Cluster Randomization Trials in Health Research. London, united Kingdom: Arnold; 2000: 1-178.

HIV; adherence; food supplementation; food insecurity; Zambia

© 2008 Lippincott Williams & Wilkins, Inc.