UCSF

Introduction

Widowed women make up 18-40% of the 12 million women living with HIV in eastern and southern Africa. Widowhood has also been associated with greater HIV morbidity and mortality. We compared the effectiveness of a multisectoral climate adaptive agricultural livelihood intervention (called Shamba Maisha) on food insecurity, and HIV related health outcomes among widowed and married women living with HIV in western Kenya.

Methods

We implemented Shamba Maisha (NCT02815579) using a cluster-randomized control trial design. The intervention arm received an US$175 in-kind loan to purchase a micro-irrigation pump, seeds, and fertilizer, and received eight training sessions on sustainable agriculture and financial management. Study outcomes were measured every 6 months over a 24-month follow-up period and trends in outcomes assessed using multilevel mixed-effects models.

Results

The trial enrolled 232 (61.5%) married and 145 (38.5%) widowed women. Widowed women (mean age 42.8 ± 8.4 years) were older than married women (35.8 ± 9.0 years) (p < 0.01). Almost all widowed women (97.2%) self-identified as household heads compared to 10.8% of married women. Comparing widowed vs married women, reduction in food insecurity (-3.13, 95%CI -4.42, -1.84 vs. -3.08, 95%CI -4.15, -2.02), depressive symptoms (-0.21, 95%CI -0.36, -0.07 vs. -0.19, 95%CI -0.29, -0.08), internalized stigma (-0.33, 95%CI -0.55, -0.11 vs. -0.38, 95%CI -0.57, -0.19), and anticipated stigma (-0.46 95%CI -0.65, -0.28 vs. -0.35, 95%CI -0.50, -0.21) was similar for both groups. In contrast, improvements in social support (-2.22, 95%CI -3.85, -0.59 vs. -4.00, 95%CI -5.16, -2.84; p = 0.08) and reduction in enacted stigma (0.01, 95%CI -0.06, 0.08 vs. -0.14, 95%CI -0.20, -0.09; p < 0.01) were weaker for widowed than married women.

Conclusions

Our study is among the first comparing the effect of a livelihood intervention on HIV health outcomes among widowed and married women. Widowed women experienced similar benefits as married women on individual-level outcomes, but weaker benefit on outcomes dependent on their external environment like enacted stigma and social support. Future trials and programs targeting widowed women should bolster stigma reduction and social support.