Widowhood stigma as a determinant of HIV risk behaviours among widows in gem sub-county, Siaya county, Kenya

Abstract

The United Nations named widowhood stigma as one of the triple burdens of widowhood alongside sexual and economic vulnerabilities. However, widowhood stigma is hardly measured and its health effects less understood. The overall aim of this study was to assess widowhood stigma as a determinant of HIV risk behaviors among widows in Gem sub-County, Siaya County, Kenya. The specific objectives included 1) to estimate the prevalence of HIV risk behaviors since widowhood, 2) to explore widows’ experiences with widowhood stigma, 3) to develop a measure for widowhood stigma, and 4) to determine the association between widowhood stigma and HIV risk behaviors. Gem sub-County was chosen as the study site because of reported concerns with widowhood stigma, a 27% HIV prevalence among widows higher than 16% in Siaya County, and a unique population-based sampling frame. Using a mixed methods exploratory sequential design, 56 purposively sampled widows first participated in in-depth interviews (n=21), three focus group discussions (n=28), and narrative life histories (n=7). Their data were analysed thematically and used to generate themes and items for measuring widowhood stigma that were used in the subsequent survey. Before the cross-sectional survey with an age-stratified random sample of 480 widows, the widowhood stigma items were first reviewed by 10 local experts for relevance and comprehensiveness and then tested and revised through cognitive interviews with six widows for comprehension, clarity and ease of response. Survey data were analyzed in three steps: 1) exploratory factor analysis, reliability and validity tests to develop the widowhood stigma scale, 2) descriptive statistics to estimate the prevalence of widowhood stigma and HIV risk behaviors, and 3) multivariable logistic regression to determine the association between widowhood stigma and HIV risk behaviours. From the qualitative data, widows experienced structural, interpersonal and intrapersonal widowhood stigma. Interpersonal widowhood stigma included negative labels and stereotypes, social isolation and loss of social status. The widowhood stigma scale had 19 items across 3 sub-scales (internalized stigma, loss of social status and social isolation) measured on a Likert scale of 1-4 with higher mean scores showing more widowhood stigma. The scale’s overall Cronbach’s alpha of 0.93 showed high reliability and its strong positive correlation (rs=0.51) with depression and moderate negative correlation (rs=-0.41) with functional social support showed preliminary construct validity. On prevalence, 1 in 2 widows experienced widowhood stigma and 1 in 2 widows engaged in any HIV risk behaviors, most commonly condomless sex (55.6%), widow inheritance (partnering of widows with the deceased husband’s relative) (60.2%) and sexual cleansing (sex to purify widow from the shadow of death) (60.0%). After adjusting for confounders such as age, education, income source and HIV status, for every unit increase in widowhood stigma scores, the odds of condomless sex (AOR: 1.72, 95%CI: 1.24, 2.41), widow inheritance (AOR: 1.58, 95%CI: 1.14, 2.18), sexual cleansing (AOR: 1.71, 95%CI: 1.24, 2.37), sex with older partners (AOR: 1.41, 95%CI: 1.06, 1.88), multiple sequential widow inheritors (AOR: 1.70, 95%CI: 1.21, 2.40), and widows receiving gift for sex (AOR: 1.52, 95%CI: 1.04, 2.22) increased. Widowhood stigma was highly prevalent and increased HIV risk behaviors since widowhood. Interventions to address widowhood stigma as driver of HIV vulnerability in widowhood is needed. While needing validation in new populations, the widowhood stigma scale should catalyze research on widows’ wellbeing and contribute to the design and evaluation of widowhood stigma reduction interventions and policies in Gem sub-County and other settings of high HIV prevalence.