Stephen McCracken
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Stephen McCracken has a total of 130 co-authors. Here are the top co-authors by number of shared publications:
Elizabeth Radin
Affiliation ICAP at Columbia University, New York, New York, United States of America
3 shared publications- Andrea Low
ICAP at Columbia University, Columbia University Mailman School of Public Health, New York, USA
2 shared publications - Bharat S. Parekh
Division of Global HIV & TB, CGH, CDC
2 shared publications Elizabeth Gummerson
Affiliation ICAP at Columbia University, New York, New York, United States of America
2 shared publications- George Williams Rutherford
University of California, USA
2 shared publications Hetal Patel
Affiliation Division of Global HIV and TB, US Centers for Disease Control and Prevention (CDC), Atlanta, GA, United States of America
2 shared publications
Latest Publications
- Open Access
- 28 January 2026
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 6 (1) , e0005804
Piloting the feasibility of a population-based joint TB-HIV survey in KwaZulu-Natal Province, South Africa, 2019
South Africa bears a high burden of infectious diseases, including intersecting epidemics of human immunodeficiency virus (HIV) and tuberculosis (TB). Among people living with HIV, TB is the predominant cause of death. Since 2002, South Africa has conducted six national population-based HIV surveys, and its first national TB prevalence survey between 2017 and 2019. Given the epidemiologic overlap of these conditions and dwindling resources, a joint national TB and HIV survey could be advantageous. We piloted a joint survey design in August–September 2019 to assess the feasibility of simultaneously collecting HIV and TB data. The pilot survey utilized the same sampling frame as the 2017–2019 national TB prevalence survey, based on small area layers as building blocks for clusters. Two clusters in KwaZulu-Natal (one urban and one rural) were selected. People of all ages were eligible to participate. Household questionnaires were administered to consenting household heads, followed by invitations to the cluster survey hub, where age-appropriate individual questionnaires were administered. Whole blood samples were tested for HIV, viral load, HIV drug resistance and HIV recent infection status. TB metrics included symptom and chest x-ray screening with sputum testing for those screening positive. Those ≥18 years received other health measurements (weight, height) and screening tests (random blood glucose, cholesterol). The survey successfully combined the collection of both HIV and TB relevant data. Overall, Household-level uptake was 78.6% (363/462), while individual-level uptake at the hub was 48.1% (616/1,280), with lower participation in the urban cluster. Uptake of additional health measurements exceeded 87%. The pilot study demonstrated that combining TB and HIV surveys is possible, but fewer people participated compared to the individual national HIV and TB surveys. Further operational research could explore how to optimize survey design, accommodate differing data requirements, and improve participation in future joint surveys.
- 28 January 2026
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 6 (1) , e0005804
Open AccessPiloting the feasibility of a population-based joint TB-HIV survey in KwaZulu-Natal Province, South Africa, 2019
South Africa bears a high burden of infectious diseases, including intersecting epidemics of human immunodeficiency virus (HIV) and tuberculosis (TB). Among people living with HIV, TB is the predominant cause of death. Since 2002, South Africa has conducted six national population-based HIV surveys, and its first national TB prevalence survey between 2017 and 2019. Given the epidemiologic overlap of these conditions and dwindling resources, a joint national TB and HIV survey could be advantageous. We piloted a joint survey design in August–September 2019 to assess the feasibility of simultaneously collecting HIV and TB data. The pilot survey utilized the same sampling frame as the 2017–2019 national TB prevalence survey, based on small area layers as building blocks for clusters. Two clusters in KwaZulu-Natal (one urban and one rural) were selected. People of all ages were eligible to participate. Household questionnaires were administered to consenting household heads, followed by invitations to the cluster survey hub, where age-appropriate individual questionnaires were administered. Whole blood samples were tested for HIV, viral load, HIV drug resistance and HIV recent infection status. TB metrics included symptom and chest x-ray screening with sputum testing for those screening positive. Those ≥18 years received other health measurements (weight, height) and screening tests (random blood glucose, cholesterol). The survey successfully combined the collection of both HIV and TB relevant data. Overall, Household-level uptake was 78.6% (363/462), while individual-level uptake at the hub was 48.1% (616/1,280), with lower participation in the urban cluster. Uptake of additional health measurements exceeded 87%. The pilot study demonstrated that combining TB and HIV surveys is possible, but fewer people participated compared to the individual national HIV and TB surveys. Further operational research could explore how to optimize survey design, accommodate differing data requirements, and improve participation in future joint surveys.
- Open Access
- 26 June 2023
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 18 (6) , e0275560
The epidemiology of HIV population viral load in twelve sub-Saharan African countries
Background: We examined the epidemiology and transmission potential of HIV population viral load (VL) in 12 sub-Saharan African countries. Methods: We analyzed data from Population-based HIV Impact Assessments (PHIAs), large national household-based surveys conducted between 2015 and 2019 in Cameroon, Cote d’Ivoire, Eswatini, Kenya, Lesotho, Malawi, Namibia, Rwanda, Tanzania, Uganda, Zambia, and Zimbabwe. Blood-based biomarkers included HIV serology, recency of HIV infection, and VL. We estimated the number of people living with HIV (PLHIV) with suppressed viral load (10 VL among unsuppressed males was higher than that among females. Correlations between VL measures and incident HIV varied, were weaker for VL metrics among females compared to males and were strongest for the number of unsuppressed PLHIV per 100 HIV-negative adults (R2 = 0.92). Conclusions: Despite higher proportions of viral suppression, female unsuppressed PLHIV outnumbered males in all countries examined. Unsuppressed male PLHIV have consistently higher VL and a higher risk of transmitting HIV than females. Just 5% of PLHIV account for almost two-thirds of countries’ total VL. Population-level VL metrics help monitor the epidemic and highlight key programmatic gaps in these African countries.
- 26 June 2023
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 18 (6) , e0275560
Open AccessThe epidemiology of HIV population viral load in twelve sub-Saharan African countries
Background: We examined the epidemiology and transmission potential of HIV population viral load (VL) in 12 sub-Saharan African countries. Methods: We analyzed data from Population-based HIV Impact Assessments (PHIAs), large national household-based surveys conducted between 2015 and 2019 in Cameroon, Cote d’Ivoire, Eswatini, Kenya, Lesotho, Malawi, Namibia, Rwanda, Tanzania, Uganda, Zambia, and Zimbabwe. Blood-based biomarkers included HIV serology, recency of HIV infection, and VL. We estimated the number of people living with HIV (PLHIV) with suppressed viral load (10 VL among unsuppressed males was higher than that among females. Correlations between VL measures and incident HIV varied, were weaker for VL metrics among females compared to males and were strongest for the number of unsuppressed PLHIV per 100 HIV-negative adults (R2 = 0.92). Conclusions: Despite higher proportions of viral suppression, female unsuppressed PLHIV outnumbered males in all countries examined. Unsuppressed male PLHIV have consistently higher VL and a higher risk of transmitting HIV than females. Just 5% of PLHIV account for almost two-thirds of countries’ total VL. Population-level VL metrics help monitor the epidemic and highlight key programmatic gaps in these African countries.
- Open Access
- 28 July 2022
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 17 (7) , e0271431
High HIV prevalence and associated factors in Lesotho: Results from a population-based survey
Despite extensive global efforts, sub-Saharan Africa remains disproportionately affected by the HIV epidemic. This generalized epidemic can be seen in Lesotho which in 2014 the HIV prevalence rate of those aged 15–49 years was 24.6%, with and incidence of 1.9 new infections per 100-person-year exposures. To better understand the impact of Lesotho’s national HIV response and significant predictors associated with HIV infection, the Lesotho Population-based HIV Impact Assessment was conducted. This survey provided a nationally representative sample of individuals aged 15–59 years old in which participants were tested for HIV and given an individual questionnaire that included socio-demographic and behavioral risk questions. The association of factors between survey questions and HIV incident was assessed using logistic regression. Multivariate logistic regression models for men and women were constructed for each outcome using variables known to be or plausibly associated with recent or chronic infection. Overall annualized incidence among people aged 15–49 was 1.19% (95% CI 0.73–1.65) per year. The overall prevalence of HIV was 25.6% with women having significantly higher prevalence. Multiple variables, including decreased wealth status, lower education levels, marital status, condom use at first sex, and circumcision (men only) were identified as being significantly associated with HIV infection for both men and women. In combination with improving the awareness of HIV status, an increased focus is needed on AGYW and men 35–49 years old to prevent new infections. HIV education and prevention programs should focus heavily on younger age groups prior to and soon after sexual debut to prevent HIV transmission. The findings of the survey showed significant room for improvement in increasing awareness of HIV status and reinforcing the need for continued HIV prevention and treatment efforts in Lesotho to prevent new infections.
- 28 July 2022
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 17 (7) , e0271431
Open AccessHigh HIV prevalence and associated factors in Lesotho: Results from a population-based survey
Despite extensive global efforts, sub-Saharan Africa remains disproportionately affected by the HIV epidemic. This generalized epidemic can be seen in Lesotho which in 2014 the HIV prevalence rate of those aged 15–49 years was 24.6%, with and incidence of 1.9 new infections per 100-person-year exposures. To better understand the impact of Lesotho’s national HIV response and significant predictors associated with HIV infection, the Lesotho Population-based HIV Impact Assessment was conducted. This survey provided a nationally representative sample of individuals aged 15–59 years old in which participants were tested for HIV and given an individual questionnaire that included socio-demographic and behavioral risk questions. The association of factors between survey questions and HIV incident was assessed using logistic regression. Multivariate logistic regression models for men and women were constructed for each outcome using variables known to be or plausibly associated with recent or chronic infection. Overall annualized incidence among people aged 15–49 was 1.19% (95% CI 0.73–1.65) per year. The overall prevalence of HIV was 25.6% with women having significantly higher prevalence. Multiple variables, including decreased wealth status, lower education levels, marital status, condom use at first sex, and circumcision (men only) were identified as being significantly associated with HIV infection for both men and women. In combination with improving the awareness of HIV status, an increased focus is needed on AGYW and men 35–49 years old to prevent new infections. HIV education and prevention programs should focus heavily on younger age groups prior to and soon after sexual debut to prevent HIV transmission. The findings of the survey showed significant room for improvement in increasing awareness of HIV status and reinforcing the need for continued HIV prevention and treatment efforts in Lesotho to prevent new infections.
- Open Access
- 17 June 2022
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 2 (6) , e0000363
Seroprevalence of SARS-CoV-2 in four states of Nigeria in October 2020: A population-based household survey
The observed epidemiology of SARS-CoV-2 in sub-Saharan Africa has varied greatly from that in Europe and the United States, with much lower reported incidence. Population-based studies are needed to estimate true cumulative incidence of SARS-CoV-2 to inform public health interventions. This study estimated SARS-CoV-2 seroprevalence in four selected states in Nigeria in October 2020. We implemented a two-stage cluster sample household survey in four Nigerian states (Enugu, Gombe, Lagos, and Nasarawa) to estimate age-stratified prevalence of SARS-CoV-2 antibodies. All individuals in sampled households were eligible for interview, blood draw, and nasal/oropharyngeal swab collection. We additionally tested participants for current/recent malaria infection. Seroprevalence estimates were calculated accounting for the complex survey design. Across all four states, 10,629 (96·5%) of 11,015 interviewed individuals provided blood samples. The seroprevalence of SARS-CoV-2 antibodies was 25·2% (95% CI 21·8–28·6) in Enugu State, 9·3% (95% CI 7·0–11·5) in Gombe State, 23·3% (95% CI 20·5–26·4) in Lagos State, and 18·0% (95% CI 14·4–21·6) in Nasarawa State. Prevalence of current/recent malaria infection ranged from 2·8% in Lagos to 45·8% in Gombe and was not significantly related to SARS-CoV-2 seroprevalence. The prevalence of active SARS-CoV-2 infection in the four states during the survey period was 0·2% (95% CI 0·1–0·4). Approximately eight months after the first reported COVID-19 case in Nigeria, seroprevalence indicated infection levels 194 times higher than the 24,198 officially reported COVID-19 cases across the four states; however, most of the population remained susceptible to COVID-19 in October 2020.
- 17 June 2022
- Published by Public Library of Science (PLoS) in PLOS Global Public Health
- Vol. 2 (6) , e0000363
Open AccessSeroprevalence of SARS-CoV-2 in four states of Nigeria in October 2020: A population-based household survey
The observed epidemiology of SARS-CoV-2 in sub-Saharan Africa has varied greatly from that in Europe and the United States, with much lower reported incidence. Population-based studies are needed to estimate true cumulative incidence of SARS-CoV-2 to inform public health interventions. This study estimated SARS-CoV-2 seroprevalence in four selected states in Nigeria in October 2020. We implemented a two-stage cluster sample household survey in four Nigerian states (Enugu, Gombe, Lagos, and Nasarawa) to estimate age-stratified prevalence of SARS-CoV-2 antibodies. All individuals in sampled households were eligible for interview, blood draw, and nasal/oropharyngeal swab collection. We additionally tested participants for current/recent malaria infection. Seroprevalence estimates were calculated accounting for the complex survey design. Across all four states, 10,629 (96·5%) of 11,015 interviewed individuals provided blood samples. The seroprevalence of SARS-CoV-2 antibodies was 25·2% (95% CI 21·8–28·6) in Enugu State, 9·3% (95% CI 7·0–11·5) in Gombe State, 23·3% (95% CI 20·5–26·4) in Lagos State, and 18·0% (95% CI 14·4–21·6) in Nasarawa State. Prevalence of current/recent malaria infection ranged from 2·8% in Lagos to 45·8% in Gombe and was not significantly related to SARS-CoV-2 seroprevalence. The prevalence of active SARS-CoV-2 infection in the four states during the survey period was 0·2% (95% CI 0·1–0·4). Approximately eight months after the first reported COVID-19 case in Nigeria, seroprevalence indicated infection levels 194 times higher than the 24,198 officially reported COVID-19 cases across the four states; however, most of the population remained susceptible to COVID-19 in October 2020.
- Open Access
- 2 September 2021
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 16 (9) , e0256865
Migration in Namibia and its association with HIV acquisition and treatment outcomes
Background: In the 21st century, understanding how population migration impacts human health is critical. Namibia has high migration rates and HIV prevalence, but little is known about how these intersect. We examined the association between migration and HIV-related outcomes using data from the 2017 Namibia Population-based HIV Impact Assessment (NAMPHIA). Methods and findings: The NAMPHIA survey selected a nationally representative sample of adults in 2017. All adults aged 15–64 years were invited to complete an interview and home-based HIV test. Recent infection (1000 copies/mL) and antiretroviral analyte data. Awareness of HIV status and antiretroviral use were based on self-report and/or detectable antiretrovirals in blood. Viremia was defined as having a viral load ≥1000 copies/mL, including all participants in the denominator regardless of serostatus. We generated community viremia values as a weighted proportion at the EA level, excluding those classified as recently infected. Significant migrants were those who had lived outside their current region or away from home >one month in the past three years. Recent cross-community in-migrants were those who had moved to the community two years ago to determine the association of migration and timing with recent infection or viral load suppression (VLS). All proportions are weighted. Of eligible adults, we had HIV results and migration data on 9,625 (83.9%) of 11,474 women and 7,291 (73.0%) of 9,990 men. Most respondents (62.5%) reported significant migration. Of cross-community in-migrants, 15.3% were recent. HIV prevalence was 12.6% and did not differ by migration status. Population VLS was 77.4%. Recent cross-community in-migration was associated with recent HIV infection (aOR: 4.01, 95% CI 0.99–16.22) after adjusting for community viremia. Significant migration (aOR 0.73, 95% CI: 0.55–0.97) and recent cross-community in-migration (aOR 0.57, 95% CI: 0.35–0.92) were associated with lower VLS, primarily due to lack of awareness of HIV infection. The study was limited by lack of precise data on trajectory of migration. Conclusions: Despite a high population-level VLS, Namibia still has migrant populations that are not accessing effective treatment for HIV. Targeting migrants with effective prevention and testing programs in communities with viremia could enable further epidemic control.
- 2 September 2021
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 16 (9) , e0256865
Open AccessMigration in Namibia and its association with HIV acquisition and treatment outcomes
Background: In the 21st century, understanding how population migration impacts human health is critical. Namibia has high migration rates and HIV prevalence, but little is known about how these intersect. We examined the association between migration and HIV-related outcomes using data from the 2017 Namibia Population-based HIV Impact Assessment (NAMPHIA). Methods and findings: The NAMPHIA survey selected a nationally representative sample of adults in 2017. All adults aged 15–64 years were invited to complete an interview and home-based HIV test. Recent infection (1000 copies/mL) and antiretroviral analyte data. Awareness of HIV status and antiretroviral use were based on self-report and/or detectable antiretrovirals in blood. Viremia was defined as having a viral load ≥1000 copies/mL, including all participants in the denominator regardless of serostatus. We generated community viremia values as a weighted proportion at the EA level, excluding those classified as recently infected. Significant migrants were those who had lived outside their current region or away from home >one month in the past three years. Recent cross-community in-migrants were those who had moved to the community two years ago to determine the association of migration and timing with recent infection or viral load suppression (VLS). All proportions are weighted. Of eligible adults, we had HIV results and migration data on 9,625 (83.9%) of 11,474 women and 7,291 (73.0%) of 9,990 men. Most respondents (62.5%) reported significant migration. Of cross-community in-migrants, 15.3% were recent. HIV prevalence was 12.6% and did not differ by migration status. Population VLS was 77.4%. Recent cross-community in-migration was associated with recent HIV infection (aOR: 4.01, 95% CI 0.99–16.22) after adjusting for community viremia. Significant migration (aOR 0.73, 95% CI: 0.55–0.97) and recent cross-community in-migration (aOR 0.57, 95% CI: 0.35–0.92) were associated with lower VLS, primarily due to lack of awareness of HIV infection. The study was limited by lack of precise data on trajectory of migration. Conclusions: Despite a high population-level VLS, Namibia still has migrant populations that are not accessing effective treatment for HIV. Targeting migrants with effective prevention and testing programs in communities with viremia could enable further epidemic control.
- Open Access
- 11 October 2020
- Published by Oxford University Press (OUP) in Journal of Survey Statistics and Methodology
- Vol. 9 (5) , 1035-1049
Dealing with Inaccurate Measures of Size in Two-Stage Probability Proportional to Size Sample Designs: Applications in African Household Surveys
The units at the early stages of multi-stage area samples are generally sampled with probabilities proportional to their estimated sizes (PPES). With such a design, an overall equal probability (EP) sample design would yield a constant number of final stage units from each final stage cluster if the measures of size used in the PPES selection at each sampling stage were directly proportional to the number of final stage units. However, there are often sizable relative differences between the measures of size used in the PPES selections and the number of final stage units. Two common approaches for dealing with these differences are: (1) to retain a self-weighting sample design, allowing the sample sizes to vary across the sampled primary sampling units (PSUs) and (2) to retain the fixed sample size in each PSU and to compensate for the unequal selection probabilities by weighting adjustments in the analyses. This article examines these alternative designs in the context of two-stage sampling in which PSUs are sampled with PPES at the first stage, and an equal probability sample of final stage units is selected from each sampled PSU at the second stage. Two-stage sample designs of this type are used for household surveys in many countries. The discussion is illustrated with data from the Population-based HIV Impact Assessment surveys that were conducted using this design in several African countries.
- 11 October 2020
- Published by Oxford University Press (OUP) in Journal of Survey Statistics and Methodology
- Vol. 9 (5) , 1035-1049
Open AccessDealing with Inaccurate Measures of Size in Two-Stage Probability Proportional to Size Sample Designs: Applications in African Household Surveys
The units at the early stages of multi-stage area samples are generally sampled with probabilities proportional to their estimated sizes (PPES). With such a design, an overall equal probability (EP) sample design would yield a constant number of final stage units from each final stage cluster if the measures of size used in the PPES selection at each sampling stage were directly proportional to the number of final stage units. However, there are often sizable relative differences between the measures of size used in the PPES selections and the number of final stage units. Two common approaches for dealing with these differences are: (1) to retain a self-weighting sample design, allowing the sample sizes to vary across the sampled primary sampling units (PSUs) and (2) to retain the fixed sample size in each PSU and to compensate for the unequal selection probabilities by weighting adjustments in the analyses. This article examines these alternative designs in the context of two-stage sampling in which PSUs are sampled with PPES at the first stage, and an equal probability sample of final stage units is selected from each sampled PSU at the second stage. Two-stage sample designs of this type are used for household surveys in many countries. The discussion is illustrated with data from the Population-based HIV Impact Assessment surveys that were conducted using this design in several African countries.
- Open Access
- 14 January 2019
- Published by Public Library of Science (PLoS) in PLOS Medicine
- Vol. 16 (1) , e1002727
Association between severe drought and HIV prevention and care behaviors in Lesotho: A population-based survey 2016–2017
A previous analysis of the impact of drought in Africa on HIV demonstrated an 11% greater prevalence in HIV-endemic rural areas attributable to local rainfall shocks. The Lesotho Population-Based HIV Impact Assessment (LePHIA) was conducted after the severe drought of 2014–2016, allowing for reevaluation of this relationship in a setting of expanded antiretroviral coverage. LePHIA selected a nationally representative sample between November 2016 and May 2017. All adults aged 15–59 years in randomly selected households were invited to complete an interview and HIV testing, with one woman per household eligible to answer questions on their experience of sexual violence. Deviations in rainfall for May 2014–June 2016 were estimated using precipitation data from Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS), with drought defined as p = 0.004), and higher HIV prevalence (OR 2.77, 95% CI 1.19–6.47, p = 0.02). It was also associated with lower educational attainment in rural females ages 15–24 years (OR 0.44, 95% CI 0.25–0.78, p = 0.005). Multivariable analysis adjusting for household wealth and sexual behavior showed that experiencing drought increased the odds of HIV infection among females 15–24 years old (adjusted OR [aOR] 1.80, 95% CI 0.96–3.39, p = 0.07), although this was not statistically significant. Migration was associated with 2-fold higher odds of HIV infection in young people (aOR 2.06, 95% CI 1.25–3.40, p = 0.006). The study was limited by the extensiveness of the drought and the small number of participants in the comparison group. Drought in Lesotho was associated with higher HIV prevalence in girls 15–19 years old in rural areas and with lower educational attainment and riskier sexual behavior in rural females 15–24 years old. Policy-makers may consider adopting potential mechanisms to mitigate the impact of income shock from natural disasters on populations vulnerable to HIV transmission.
- 14 January 2019
- Published by Public Library of Science (PLoS) in PLOS Medicine
- Vol. 16 (1) , e1002727
Open AccessAssociation between severe drought and HIV prevention and care behaviors in Lesotho: A population-based survey 2016–2017
A previous analysis of the impact of drought in Africa on HIV demonstrated an 11% greater prevalence in HIV-endemic rural areas attributable to local rainfall shocks. The Lesotho Population-Based HIV Impact Assessment (LePHIA) was conducted after the severe drought of 2014–2016, allowing for reevaluation of this relationship in a setting of expanded antiretroviral coverage. LePHIA selected a nationally representative sample between November 2016 and May 2017. All adults aged 15–59 years in randomly selected households were invited to complete an interview and HIV testing, with one woman per household eligible to answer questions on their experience of sexual violence. Deviations in rainfall for May 2014–June 2016 were estimated using precipitation data from Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS), with drought defined as p = 0.004), and higher HIV prevalence (OR 2.77, 95% CI 1.19–6.47, p = 0.02). It was also associated with lower educational attainment in rural females ages 15–24 years (OR 0.44, 95% CI 0.25–0.78, p = 0.005). Multivariable analysis adjusting for household wealth and sexual behavior showed that experiencing drought increased the odds of HIV infection among females 15–24 years old (adjusted OR [aOR] 1.80, 95% CI 0.96–3.39, p = 0.07), although this was not statistically significant. Migration was associated with 2-fold higher odds of HIV infection in young people (aOR 2.06, 95% CI 1.25–3.40, p = 0.006). The study was limited by the extensiveness of the drought and the small number of participants in the comparison group. Drought in Lesotho was associated with higher HIV prevalence in girls 15–19 years old in rural areas and with lower educational attainment and riskier sexual behavior in rural females 15–24 years old. Policy-makers may consider adopting potential mechanisms to mitigate the impact of income shock from natural disasters on populations vulnerable to HIV transmission.