Food insecurity and HIV viral suppression are linked through nutritional, mental health, and behavioral pathways. We examined the effects of a multisectoral agricultural intervention on HIV viral suppression, nutritional status, and mental health among people living with HIV (PLHIV) taking antiretroviral therapy (ART) in Kenya.
For the Shamba Maisha cluster RCT (NCT02815579) we randomly allocated 16 health facilities (1:1; match-paired) in Kenya to intervention or control arms. The intervention included a bank loan to purchase farming commodities including a human-powered irrigation pump, fertilizer and seeds plus training in sustainable agriculture and financial literacy. Participants were ?18 years old, on ART >6 months, moderately-to-severely food insecure, with access to farmland and water. Every 6 months participants were followed for 24 months. The primary outcome was change from baseline to endline in viral suppression (<200 copies/mL), compared between arms using difference-in-differences analyses. Secondary outcomes included changes in food insecurity, ART adherence, clinic attendance, depression, self-confidence, and social support.
Between June 2016 and December 2017, we enrolled 366 and 354 participants in the intervention and control arms, respectively. Fifty five percent of participants were women. Retention at 24 months was 94.0%. HIV viral suppression increased in both arms from baseline to endline: intervention 85.8% to 95.1% and control 82.4% to 94.3% (p=0.86). The proportion of participants who missed a scheduled HIV clinic visit as well as adherence to ART were not different by arm. Levels of food insecurity decreased more in the intervention than control arm. The proportions of those with depression declined more in the intervention arm. Self-confidence and social support both improved more in the intervention arm (see Table).
A multisectoral agricultural intervention reduced food insecurity and depressive symptoms, and improved self-confidence and social support among PLHIV. Because viral suppression approached the UNAIDS goal of ?95% among all participants, who resided in settings with widespread test and treat policies, no additional effects of the intervention on HIV clinical indicators occurred. Interventions that improve livelihoods may help address the structural drivers of poor health and co-morbidities affecting PLHIV in resource limited settings.