A simplified point-of-service model for hepatitis C in people who inject drugs in South Africa.

BACKGROUND: Globally, 9% of people who inject drugs (PWID), a key hepatitis C-infected population, reside in sub-Saharan Africa. In South Africa, hepatitis C seroprevalence in PWID is high. It is almost 84% in Pretoria and hepatitis C genotypes 1 and 3 predominate. Access to hepatitis C care for PWID is inadequate given low referral rates, socio-structural barriers, homelessness and limited access to harm reduction. Traditional care models do not address the needs of this population. We piloted a simplified complete point-of-service care model, a first of its kind in the country and sub-continental region. METHODS: Community-based recruitment from Pretoria's PWID population occurred over 11 months. Participants were screened with point-of-care rapid diagnostic tests for HBsAg (Alere Determine™), hepatitis C and HIV antibodies (OraQuick®). Qualitative HCV viremia was confirmed on site with Genedrive® (Sysmex), similarly at week 4, end of treatment and to confirm sustained virological response. Viremic hepatitis C participants were initiated on 12 weeks of daily sofosbuvir and daclatasvir. Harm reduction and adherence support, through directly observed therapy, peer support, a stipend and transport, was provided. RESULTS: A total of 163 participants were screened for hepatitis C antibody, and 66% were positive with 80 (87%) viremic. An additional 36 confirmed hepatitis C viremic participants were referred. Of those eligible to initiate treatment, 87 (93%) were commenced on sofosbuvir and daclatasvir, with 98% (n = 85) male, 35% (n = 30) HIV co-infected, 1% (n = 1) HBV co-infected and 5% (n = 4) HIV/HBV/HCV triple infected. Some 67% (n = 58) accessed harm reduction packs, 57% (n = 50) opioid substitution therapy and 18% (n = 16) stopped injecting. A per protocol sustained virological response of 90% (n = 51) was achieved with 14% (n = 7) confirmed reinfections following a sustained virological response. HCV RNA qualitative testing performance was acceptable with all sustained virological responses validated against a laboratory assay. Mild adverse effects were reported in 6% (n = 5). Thirty-eight percent (n = 33) of participants were lost to follow-up. CONCLUSION: In our setting, a simplified point-of-service hepatitis C care model for PWID yielded an acceptable sustained virological response rate. Retention in care and follow-up remains both challenging and central to success. We have demonstrated the utility of a model of care for our country and region to utilize this more community acceptable and simplified practice.

Dietary sodium intake in urban and rural Malawi, and directions for future interventions.

Background: High dietary sodium intake is a major risk factor for hypertension. Data on population sodium intake are scanty in sub-Saharan Africa, despite a high hypertension prevalence in most countries. Objective: We aimed to determine daily sodium intake in urban and rural communities in Malawi. Design: In an observational cross-sectional survey, data were collected on estimated household-level per capita sodium intake, based on how long participants reported that a defined quantity of plain salt lasts in a household. In a subset of 2078 participants, 24-h urinary sodium was estimated from a morning spot urine sample. Results: Of 29,074 participants, 52.8% of rural and 50.1% of urban individuals lived in households with an estimated per capita plain salt consumption >5 g/d. Of participants with urinary sodium data, 90.8% of rural and 95.9% of urban participants had estimated 24-h urinary sodium >2 g/d; there was no correlation between household per capita salt intake and estimated 24-h urinary sodium excretion. Younger adults were more likely to have high urinary sodium and to eat food prepared outside the home than were those over the age of 60 y. Households with a member with previously diagnosed hypertension had reduced odds (OR: 0.59; 95% CI: 0.51, 0.68) of per capita household plain salt intake >5 g/d, compared with those where hypertension was undiagnosed. Conclusions: Sodium consumption exceeds the recommended amounts for most of the population in rural and urban Malawi. Population-level interventions for sodium intake reduction with a wide focus are needed, targeting both sources outside the home as well as home cooking. This trial was registered at clinicaltrials.gov as NCT03422185.

Glycated haemoglobin A1c (HbA1c) for detection of diabetes mellitus and impaired fasting glucose in Malawi: a diagnostic accuracy study.

OBJECTIVES: To examine the accuracy of glycated haemoglobin A1c (HbA1c) in detecting type 2 diabetes and impaired fasting glucose among adults living in Malawi. DESIGN: A diagnostic validation study of HbA1c. Fasting plasma glucose (FPG) ≥7.0 mmol/L was the reference standard for type 2 diabetes, and FPG between 6.1 and 6.9 mmol/L as impaired fasting glucose. PARTICIPANTS: 3645 adults (of whom 63% were women) recruited from two demographic surveillance study sites in urban and rural Malawi. This analysis excluded those who had a previous diagnosis of diabetes or had history of taking diabetes medication. RESULTS: HbA1c demonstrated excellent validity to detect FPG-defined diabetes, with an area under the receiver operating characteristic (AUROC) curve of 0.92 (95% CI 0.90 to 0.94). At HbA1c ≥6.5% (140 mg/dL), sensitivity was 78.7% and specificity was 94.0%. Subgroup AUROCs ranged from 0.86 for participants with anaemia to 0.94 for participants in urban Malawi. There were clinical and metabolic differences between participants with true diabetes versus false positives when HbA1c was ≥6.5% (140 mg/dL). CONCLUSIONS: The findings from this study provide justification to use HbA1c to detect type 2 diabetes. As HbA1c testing is substantially less burdensome to patients than either FPG testing or oral glucose tolerance testing, it represents a useful option for expanding access to diabetes care in sub-Saharan Africa.

Changes in Fertility at the Population Level in the Era of ART in Rural Malawi.

INTRODUCTION: HIV reduces fertility through biological and social pathways, and antiretroviral treatment (ART) can ameliorate these effects. In northern Malawi, ART has been available since 2007 and lifelong ART is offered to all pregnant or breastfeeding HIV-positive women. METHODS: Using data from the Karonga Health and Demographic Surveillance Site in Malawi from 2005 to 2014, we used total and age-specific fertility rates and Cox regression to assess associations between HIV and ART use and fertility. We also assessed temporal trends in in utero and breastfeeding HIV and ART exposure among live births. RESULTS: From 2005 to 2014, there were 13,583 live births during approximately 78,000 person years of follow-up of women aged 15-49 years. The total fertility rate in HIV-negative women decreased from 6.1 [95% confidence interval (CI): 5.5 to 6.8] in 2005-2006 to 5.1 (4.8-5.5) in 2011-2014. In HIV-positive women, the total fertility rate was more stable, although lower, at 4.4 (3.2-6.1) in 2011-2014. In 2011-2014, compared with HIV-negative women, the adjusted (age, marital status, and education) hazard ratio was 0.7 (95% CI: 0.6 to 0.9) and 0.8 (95% CI: 0.6 to 1.0) for women on ART for at least 9 months and not (yet) on ART, respectively. The crude fertility rate increased with duration on ART up to 3 years before declining. The proportion of HIV-exposed infants decreased, but the proportion of ART-exposed infants increased from 2.4% in 2007-2010 to 3.5% in 2011-2014. CONCLUSIONS: Fertility rates in HIV-positive women are stable in the context of generally decreasing fertility. Despite a decrease in HIV-exposed infants, there has been an increase in ART-exposed infants.

Prevalence and risk factors for anemia severity and type in Malawian men and women: urban and rural differences.

BACKGROUND: The global burden of anemia is large especially in sub-Saharan Africa, where HIV is common and lifestyles are changing rapidly with urbanization. The effects of these changes are unknown. Studies of anemia usually focus on pregnant women or children, among whom the burden is greatest. We describe prevalence and risk factors for anemia among rural and urban men and women of all ages in Malawi. METHODS: We analyzed data from a population-wide cross-sectional survey of adults conducted in two sites, Karonga (rural) and Lilongwe (urban), commencing in May 2013. We used multinomial logistic regression models, stratified by sex to identify risk factors for mild and moderate-to-severe anemia. RESULTS: Anemia prevalence was assessed among 8,926 men (age range 18-100 years) and 14,978 women (age range: 18-103 years). Weighted prevalence levels for all, mild, and moderate-to-severe anemia were 8.2, 6.7 and 1.2% in rural men; 19.4, 12.0 and 7.4% in rural women; 5.9, 5.1 and 0.8% in urban men; and 23.4, 13.6 and 10.1% in urban women. Among women, the odds of anemia were higher among urban residents and those with higher socioeconomic status. Increasing age was associated with higher anemia prevalence in men. Among both men and women, HIV infection was a consistent risk factor for severity of anemia, though its relative effect was stronger on moderate-to-severe anemia. CONCLUSIONS: The drivers of anemia in this population are complex, include both socioeconomic and biological factors and are affecting men and women differently. The associations with urban lifestyle and HIV indicate opportunities for targeted intervention.

Sustained 10-year gain in adult life expectancy following antiretroviral therapy roll-out in rural Malawi: July 2005 to June 2014.

Background: Improved life expectancy in high HIV prevalence populations has been observed since antiretroviral therapy (ART) scale-up. However, it is unclear if the benefits are sustained, and the mortality among HIV-positive individuals not (yet) on ART is not well described. We assessed temporal change in mortality over 9 years in rural Malawi. Methods: Within a demographic surveillance site in northern rural Malawi, we combined demographic, HIV and ART uptake data. We calculated life expectancy using Kaplan-Meier estimates, and compared mortality rates and rate ratios using Poisson regression, by period of ART availability (July 2005-June 2008, July 2008-June 2011 and July 2011-June 2014). Results: Among 32 664 individuals there were 1424 deaths; 1930 individuals were known HIV-positive, of whom 1382 started ART. Overall, life expectancy at age 15 years increased by 10 years within 5 years of ART introduction, and plateaued. Age-standardized adult mortality rates declined from 11.3/1000 to 7.5/1000 person-years between the first and last time period. In July 2011-June 2014 compared with July 2005-June 2008, mortality declined in HIV-positive individuals on ART (rate ratio adjusted (aRR) for age, sex, location and education, 0.3; 95% confidence interval (CI) 0.2-0.5) and in those not (yet) on ART (aRR 0.3; 95%CI 0.1-0.5) but not in HIV-negative individuals (aRR 1.1; 95%CI 0.7-1.9). Conclusions: Total population adult life expectancy increased toward that of HIV-negative individuals by 2011 and remained raised. The reduction in all-cause and HIV-related mortality in HIV-positive individuals not (yet) on ART suggests ART uptake is occurring at an earlier disease stage, particularly in women.

Accessing clinical services and retention in care following screening for hypertension and diabetes among Malawian adults: an urban/rural comparison.

OBJECTIVE: Interventions to impact on the burden of chronic noncommunicable diseases, such as hypertension and diabetes, include screening of asymptomatic adults, but little is known about the subsequent course of clinical care. We report on the uptake of referral for clinical assessment and retention in care, following a large urban/rural population screening program in Malawi. METHODS: Adult residents were screened for raised blood pressure and raised fasting blood glucose at a demographic surveillance site in rural Karonga District and in urban Area 25, Lilongwe with well supported chronic care clinics. Successful uptake was defined as presenting for clinical assessment within 6 weeks of referral, and nonattenders were followed at home. Logistic regression was used to examine association of uptake with demographic and clinical factors. Retention was assessed using survival analysis techniques. RESULTS: A total of 27 305 participants were screened for hypertension and diabetes between May 2013 and September 2015. Of these, 4075 (14.9%) were referred for suspected hypertension (3640), diabetes (172), or both (263). Among those referred, 2480 (60.9%), reported for clinical assessment. Factors associated with uptake of care included being female, rural residency, older age, unemployment, prior medication, and diabetes. Retention, for those enrolled in care following a formal clinical assessment, was associated with the final diagnosis following clinical assessment, rural residency, and older age. CONCLUSION: Screening for hypertension and diabetes identifies large numbers of individuals who need further clinical assessment, but strategies are needed to ensure better linkage and retention into care.

Hypertension and diabetes in Africa: design and implementation of a large population-based study of burden and risk factors in rural and urban Malawi.

BACKGROUND: The emerging burden of cardiovascular disease and diabetes in sub-Saharan Africa threatens the gains made in health by the major international effort to combat infectious diseases. There are few data on distribution of risk factors and outcomes in the region to inform an effective public health response. A comprehensive research programme is being developed aimed at accurately documenting the burden and drivers of NCDs in urban and rural Malawi; to design and test intervention strategies. The programme includes population surveys of all people aged 18 years and above, linking individuals with newly diagnosed hypertension and diabetes to healthcare and supporting clinical services. The successes, challenges and lessons learnt from the programme to date are discussed. RESULTS: Over 20,000 adults have been recruited in rural Karonga and urban Lilongwe. The urban population is significantly younger and wealthier than the rural population. Employed urban individuals, particularly males, give particular recruitment challenges; male participation rates were 80.3 % in the rural population and 43.6 % in urban, whilst female rates were 93.6 and 75.6 %, respectively. The study is generating high quality data on hypertension, diabetes, lipid abnormalities and risk factors. CONCLUSIONS: It is feasible to develop large scale studies that can reliably inform the public health approach to diabetes, cardiovascular disease and other NCDs in Sub-Saharan Africa. It is essential for studies to capture both rural and urban populations to address disparities in risk factors, including age structure. Innovative approaches are needed to address the specific challenge of recruiting employed urban males.

Human Immunodeficiency Virus, Antiretroviral Therapy and Markers of Lymphatic Filariasis Infection: A Cross-sectional Study in Rural Northern Malawi.

BACKGROUND: Lymphatic filariasis (LF) and human immunodeficiency virus (HIV) are major public health problems. Individuals may be co-infected, raising the possibility of important interactions between these two pathogens with consequences for LF elimination through annual mass drug administration (MDA). METHODOLOGY AND PRINCIPAL FINDINGS: We analysed circulating filarial antigenaemia (CFA) by HIV infection status among adults in two sites in northern Malawi, a region endemic for both LF and HIV. Stored blood samples and data from two geographically separate studies were used: one a recruitment phase of a clinical trial of anti-filarial agent dosing regimens, and the other a whole population annual HIV sero-survey. In study one, 1,851 consecutive adult volunteers were screened for HIV and LF infection. CFA prevalence was 25.4% (43/169) in HIV-positive and 23.6% (351/1487) in HIV-negative participants (p=0.57). Geometric mean CFA concentrations were 859 and 1660 antigen units per ml of blood (Ag/ml) respectively, geometric mean ratio (GMR) 0.85, 95%CI 0.49-1.50. In 7,863 adults in study two, CFA prevalence was 20.9% (86/411) in HIV-positive and 24.0% (1789/7452) in HIV-negative participants (p=0.15). Geometric mean CFA concentrations were 630 and 839 Ag/ml respectively (GMR 0.75, 95%CI 0.60-0.94). In the HIV-positive group, antiretroviral therapy (ART) use was associated with a lower CFA prevalence, 12.7% (18/142) vs. 25.3% (67/265), (OR 0.43, 95%CI 0.24-0.76). Prevalence of CFA decreased with duration of ART use, 15.2% 0-1 year (n=59), 13.6% >1-2 years (n=44), 10.0% >2-3 years (n=30) and 0% >3-4 years treatment (n=9), p<0.01 χ2 for linear trend. CONCLUSIONS/SIGNIFICANCE: In this large cross-sectional study of two distinct LF-exposed populations, there is no evidence that HIV infection has an impact on LF epidemiology that will interfere with LF control measures. A significant association of ART use with lower CFA prevalence merits further investigation to understand this apparent beneficial impact of ART.

Earlier menarche is associated with a higher prevalence of Herpes simplex type-2 (HSV-2) in young women in rural Malawi.

Remarkably little is known about associations between age at menarche and sexually transmitted infections, although girls with earlier menarche tend to have earlier sexual debut and school drop-out, so an association might be expected. In a population-based survey of >3000 women aged 15-30 in northern Malawi we show that those with earlier menarche had earlier sexual debut, earlier marriage and were more often Herpes simplex type-2 (HSV-2) positive. Compared to those with menarche aged <14, the age-adjusted odds ratios for HSV-2 were 0.89 (95%CI 0.71-1.1), 0.71 (0.57-0.89) and 0.69 (0.54-0.89) for menarche aged 14, 15 and 16+ respectively. This association persisted after adjusting for socio-economic factors, including schooling, and for sexual behaviour. No such association was seen with HIV infection, which is much less common and less uniformly distributed than HSV-2 in this population. The extra vulnerability of girls with earlier menarche needs to be recognised. DOI: http://dx.doi.org/10.7554/eLife.01604.001.

Underestimation of HIV prevalence in surveys when some people already know their status, and ways to reduce the bias.

OBJECTIVE: To quantify refusal bias due to prior HIV testing, and its effect on HIV prevalence estimates, in general-population surveys. DESIGN: Four annual, cross-sectional, house-to-house HIV serosurveys conducted during 2006-2010 within a demographic surveillance population of 33 000 in northern Malawi. METHODS: The effect of prior knowledge of HIV status on test acceptance in subsequent surveys was analysed. HIV prevalence was then estimated using ten adjustment methods, including age-standardization; multiple imputation of missing data; a conditional probability equations approach incorporating refusal bias; using longitudinal data on previous and subsequent HIV results; including self-reported HIV status; and including linked antiretroviral therapy clinic data. RESULTS: HIV test acceptance was 55-65% in each serosurvey. By 2009/2010 79% of men and 85% of women had tested at least once. Known HIV-positive individuals were more likely to be absent, and refuse interviewing and testing. Using longitudinal data, and adjusting for refusal bias, the best estimate of HIV prevalence was 7% in men and 9% in women in 2008/2009. Estimates using multiple imputations were 4.8 and 6.4%, respectively. Using the conditional probability approach gave good estimates using the refusal risk ratio of HIV-positive to HIV-negative individuals observed in this study, but not when using the only previously published estimate of this ratio, even though this was also from Malawi. CONCLUSION: As the proportion of the population who know their HIV-status increases, survey-based prevalence estimates become increasingly biased. As an adjustment method for cross-sectional data remains elusive, sources of data with high coverage, such as antenatal clinics surveillance, remain important.

Using age-specific mortality of HIV infected persons to predict anti-retroviral treatment need: a comparative analysis of data from five African population-based cohort studies.

OBJECTIVES: To present a simple method for estimating population-level anti-retroviral therapy (ART) need that does not rely on knowledge of past HIV incidence. METHODS: A new approach to estimating ART need is developed based on calculating age-specific proportions of HIV-infected adults expected to die within a fixed number of years in the absence of treatment. Mortality data for HIV-infected adults in the pre-treatment era from five African HIV cohort studies were combined to construct a life table, starting at age 15, smoothed with a Weibull model. Assuming that ART should be made available to anyone expected to die within 3 years, conditional 3-year survival probabilities were computed to represent proportions needing ART. The build-up of ART need in a successful programme continuously recruiting infected adults into treatment as they age to within 3 years of expected death was represented by annually extending the conditional survival range. RESULTS: The Weibull model: survival probability in the infected state from age 15 = exp(-0.0073 × (age - 15)(1.69)) fitted the pooled age-specific mortality data very closely. Initial treatment need for infected persons increased rapidly with age, from 15% at age 20-24 to 32% at age 40-44 and 42% at age 60-64. Overall need in the treatment of naïve population was 24%, doubling within 5 years in a programme continually recruiting patients entering the high-risk period for dying. CONCLUSION: A reasonable projection of treatment need in an ART naive population can be made based on the age and gender profile of HIV-infected people.

Adult mortality and probable cause of death in rural northern Malawi in the era of HIV treatment.

OBJECTIVES: Developing countries are undergoing demographic transition with a shift from high mortality caused by communicable diseases (CD) to lower mortality rates caused by non-communicable diseases (NCD). HIV/AIDS has disrupted this trend in sub-Saharan Africa. However, in recent years, HIV-associated mortality has been reduced with the introduction of widely available antiretroviral therapy (ART). Side effects of ART may lead to increased risk of cardiovascular diseases, raising the prospects of an accelerated transition towards NCD as the primary cause of death. We report population-based data to investigate changes in cause of death owing to NCD during the first 4 years after introduction of HIV treatment. METHODS: We analysed data from a demographic surveillance system in Karonga district, Malawi, from September 2004 to August 2009. ART was introduced in mid-2005. Clinician review of verbal autopsies conducted 2-6 weeks after a death was used to establish a single principal cause of death. RESULTS: Over the entire period, there were 905 deaths, AIDS death rate fell from 505 to 160/100,000 person-years, and there was no evidence of an increase in NCD rates. The proportion of total deaths attributable to AIDS fell from 42% to 17% and from NCD increased from 37% to 49%. DISCUSSION: Our findings show that 4 years after the introduction of ART into HIV care in Karonga district, all-cause mortality has fallen dramatically, with no evidence of an increase in deaths owing to NCD.

Measuring concurrency: an empirical study of different methods in a large population-based survey and evaluation of the UNAIDS guidelines.

BACKGROUND: Recent UNAIDS guidelines recommend measuring concurrency 6 months before the interview date, based on overlapping partnership dates. This has theoretical advantages, but little is known about how well it can be measured in practice. METHODS: The assumptions underlying the UNAIDS measure were tested using data from a sexual behaviour survey conducted in rural northern Malawi. All resident adults aged 15-59 were eligible. Questions included self-reported concurrency and dates for all marital and nonmarital partnerships in the past 12 months. RESULTS: A total of 6796 women and 5253 men were interviewed, 83 and 72% of those eligible, respectively. Since few women reported multiple partners, detailed analysis was restricted to men. Overall 19.2% [95% confidence interval (CI) 18.1-20.2] of men self-reported concurrent relationships in the past year (almost all of those with more than one partner). Using overlapping dates the estimate was 16.7% (15.7-17.7). Excluding partnerships which tied on dates (making overlap uncertain) or restricting the analysis to the three most recent partners gave similar results. The UNAIDS 6-month measure was 12.0% (11.1-12.9), and current concurrency was 11.5% (10.6-12.4). The difference between dates-based and self-reported 12-month measures was much larger for unmarried men: 11.1% (9.7-12.4) self-reported; 7.1% (6.9-8.2) on dates. Polygyny (15% of married men) and the longer duration of relationships stabilized the estimates for married men. Nonmarital partnerships were under-reported, particularly those starting longer ago. CONCLUSIONS: The difficulties of recall of dates for relationships, and under-reporting of partners lead to underestimation of concurrency using date-based measures. Self-reported concurrency is much easier to measure and appears more complete.

Assessing the validity of sexual behaviour reports in a whole population survey in rural Malawi.

BACKGROUND: Sexual behaviour surveys are widely used, but under-reporting of particular risk behaviours is common, especially by women. Surveys in whole populations provide an unusual opportunity to understand the extent and nature of such under-reporting. METHODS: All consenting individuals aged between 15 and 59 within a demographic surveillance site in northern Malawi were interviewed about their sexual behaviour. Validity of responses was assessed by analysis of probing questions; by comparison of results with in-depth interviews and with Herpes simplex type-2 (HSV-2) seropositivity; by comparing reports to same sex and opposite sex interviewers; and by quantifying the partnerships within the local community reported by men and by women, adjusted for response rates. RESULTS: 6,796 women and 5,253 men (83% and 72% of those eligible) consented and took part in sexual behaviour interviews. Probing questions and HSV-2 antibody tests in those who denied sexual activity identified under-reporting for both men and women. Reports varied little by sex or age of the interviewer. The number of marital partnerships reported was comparable for men and women, but men reported about 4 times as many non-marital partnerships. The discrepancy in reporting of non-marital partnerships was most marked for married women (men reported about 7 times as many non-marital partnerships with married women as were reported by married women themselves), but was only apparent in younger married women. CONCLUSIONS: We have shown that the under-reporting of non-marital partnerships by women was strongly age-dependent. The extent of under-reporting of sexual activity by young men was surprisingly high. The results emphasise the importance of triangulation, including biomarkers, and the advantages of considering a whole population.

Age at menarche, schooling, and sexual debut in northern Malawi.

BACKGROUND: Age at sexual debut is a key behavioural indicator used in HIV behavioural surveillance. Early age at menarche may precipitate early sex through perceived readiness for sex, or through school drop-out, but this is rarely studied. We investigated trends and circumstances of sexual debut in relation to schooling and age at menarche. METHODS AND FINDINGS: A cross-sectional sexual behaviour survey was conducted on all individuals age 15-59 within a demographic surveillance site in Karonga District, Malawi. Time trends were assessed using birth cohorts. Survival analysis was used to estimate the median age at menarche, sexual debut and first marriage. The 25(th) centile was used to define "early" sex, and analyses of risk factors for early sex were restricted to those who had reached that age, and were done using logistic regression. Of the 8232 women and 7338 men resident in the area, 88% and 78%, respectively, were seen, and, 94% and 92% of these were interviewed. The median reported age at first sex was 17.5 for women and 18.8 for men. For women, ages at menarche, sexual debut and first marriage did not differ by birth cohort. For men, age at sexual debut and first marriage decreased slightly in later birth cohorts. For both men and women increased schooling was associated with later sexual debut and a longer delay between sexual debut and first marriage, but the associations were stronger for women. Earlier age at menarche was strongly associated with earlier sexual debut and marriage and lower schooling levels. In women early sexual debut (<16 years) was less likely in those with menarche at age 14-15 (odds ratio (OR) 0.31, 95%CI 0.26-0.36), and ≥16 (OR 0.04, 95%CI 0.02-0.05) compared to those with menarche at <14. The proportion of women who completed primary school was 46% in those with menarche at <14, 60% in those with menarche at 14-15 and 70% in those with menarche at ≥16. The association between age at menarche and schooling was partly explained by age at sexual debut. The association between age at menarche and early sex was not altered by adjusting for schooling. CONCLUSIONS: Women with early menarche start sex and marry early, leading to school drop-out. It is important to find ways to support those who reach menarche early to access the same opportunities as other young women.