Title

The impact of severe drought and HIV in Southern Africa: Results from the LePHIA survey 2016-2017

Presenter

Andrea Low

Authors

A. Low¹, K. Frederix², S. McCracken³, E. Gummerson⁴, E. Radin⁴, S. Davia⁵, N. Ahmed⁴, S. Findley⁶, A. Schwitters⁵

Institutions

¹Columbia University, ICAP, New York, United States, ²ICAP, Maseru, Lesotho, ³CDC, Atlanta, United States, ⁴ICAP at Columbia University, New York, United States, ⁵CDC-Lesotho, Maseru, Lesotho, ⁶Columbia University, New York, United States

Background: An analysis of the impact of drought in Africa on HIV by Burke, et al (2014) demonstrated an 11% increase in prevalence in HIV-endemic rural areas attributable to local rainfall shocks. The Lesotho Population-Based HIV Impact Assessment (LePHIA) was conducted November 2016-May 2017, after one of the most severe droughts in Southern African history, allowing for reevaluation of this relationship in a setting of expanded antiretroviral (ART) coverage.
Methods: LePHIA selected a nationally representative sample of adults and children. Individuals completed an interview and HIV testing. Deviations in rainfall for May 2014-June 2016 were measured using precipitation data from Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS), with severe drought (SD) set at < 15% of the average rainfall from 1981-2016. The association between SD and HIV-related outcomes was assessed using logistic regression, incorporating survey weights and stratified by age, gender and geography (urban vs rural), with a particular focus on young persons.
Results: All of Lesotho suffered from drought, with regions receiving 1% to 31% of their historical rainfall. Only 5.7% (507/8825) of households were in regions without SD, corresponding to 5.9% of children (271/4,291) and 5.8% of adults (742/12,887). In rural areas, 39.5% of households in SD areas reported food shortage in the past four weeks vs. 27.8% in non-drought areas (p=0.01). SD was associated with higher HIV prevalence in 15-19 year old females in rural areas (odds ratio [OR]=2.77, 95%CI:1.19-6.47,p=0.02), and lower awareness of HIV status in urban females aged 15-24 (OR=0.23, 95%CI:0.08-0.70, p=0.01). In rural 15-24 year old females, SD was associated with higher reported transactional sex (OR=3.26, 95%CI:1.78-5.98, p< 0.001), early sexual debut (OR=3.11, 95%CI:1.43-6.74, p=0.004), lower attendance of secondary school (OR=0.44, 95%CI:0.0.25-0.78, p=0.005), and lower condom use (at last sex, OR=0.55, 95%CI:0.31-0.99, p=0.05). In a multivariable model, the impact of drought on HIV prevalence in 15-19 year old females persisted after adjusting for household wealth.
Conclusions: Severe drought was associated with higher HIV prevalence in adolescent girls, lower educational attainment and higher risky sexual behavior. Policy-makers should adopt potential mechanisms to mitigate the impact of income shock from natural disasters on vulnerable populations.

Download the e-Poster