Background
Child marriage, defined as marriage prior to 18 years of age is expected to become the reality of 100
million adolescent girls in the developing world during the next 10 years. Sub-Saharan Africa and South
Asia regions have one of the highest rates of child marriages in the world. The practice is prevalent in
Pakistan, which disproportionately affects young girls in rural, low income and low education
households. Our study objectives were: To determine the association between child marriage and
maternal healthcare services utilization in Pakistan beyond those attributed to social vulnerabilities
(women’s economic status, education, ethnicity, and place of residence) (Paper I). To determine the
association between child marriage and high fertility and poor fertility health indicators beyond those
attributed to social vulnerabilities (Paper II). To determine the association between child marriage and
morbidity and mortality of children under 5 years of age in Pakistan beyond those attributed to social
vulnerabilities (Paper III). To determine the association between child marriage and controlling
behaviors and spousal violence by husbands against adolescent and young women in Pakistan beyond
those attributed to social vulnerabilities (Paper IV). To describe women’s knowledge and attitude
towards child marriage practice who themselves were married as children to gain insight, especially
cultural factors for designing future interventions (Paper V).
Methods
Nationally representative cross-sectional data from Pakistan Demographic and Health Survey (PDHS),
2006-07 was used for papers I, II and III, and PDHS, 2012-13 for paper IV. In paper I, we limited the
data to ever-married women aged 15–24 years with at least one childbirth (n=1,404) to identify
differences in prenatal care provision (skilled or unskilled medical care provider), antenatal care
(antenatal visits; care at home or hospital), care at delivery (assistance by unskilled medical care
provider), and place of birth by early (younger than 18 years) compared with adult (18 years or older)
age at marriage. In paper II, we limited the data to ever-married women aged 20-24 years (n=1,560) to
identify differences in poor fertility outcomes [high fertility (three or more childbirths); rapid repeat
childbirth (<24 months between births); unwanted pregnancy (any ever); pregnancy termination (any
stillbirth, miscarriage or abortion ever)] by early compared with adult age at marriage. In paper III, we
limited the data to children from the past five years, reported by ever- married women aged 15-24 years
(n=2630 births of n= 2138 mothers) to identify differences in infectious diseases in past 2 weeks
(diarrhea, acute respiratory infection [ARI], ARI with fever), children under 5 years of age mortality and
infant mortality, and low birth weight by early compared with adult age at marriage. In paper IV, we
limited the data to currently married women aged 15-24 years who had participated in the domestic
violence module (n=589) to identify differences in controlling behaviors and spousal violence
experiences between early and adult age at marriage. Associations between child marriage and health
outcomes were assessed by calculating adjusted odds ratio (AOR) using logistic regression models after
controlling for covariates including social equity indicators (education, wealth index, rural residence). In
paper V, women of reproductive age (15-49 years) who were married prior to 18 years, for at least 5
years and had at least one childbirth were recruited from most populous slum areas of Lahore, Pakistan.
Themes for the interview were developed using published literature and everyday observations of the
researchers. Interviews were conducted by trained interviewers in Urdu language and were translated
into English. The interviews were tape-recorded, transcribed, analyzed and categorized into themes.
Results
Paper I: Overall, 66.1% of ever-married women aged 15–24 years in Pakistan with at least one childbirth
were married before the age of 18 years. More than half (61.9%) of women married as children had no
formal education, and the majority (71.0%) resided in rural areas. Child marriage was significantly
associated with decreased likelihood of any prenatal care (AOR=0.73; 95% CI: 0.534-0.993) and
prenatal care by skilled medical care providers (AOR=0.64; 95% CI: 0.476-0.871), and increased
likelihood of delivery assistance by unskilled medical providers (AOR=1.90; 95% CI: 1.435-2.518) and
xiii
delivery at home (AOR=2.17; 95% CI: 1.617-2.915). Paper II: Overall, 50% of ever-married women
aged 20-24 years in Pakistan were married before the age of 18 years. Child marriage was significantly
(p<0.001) associated with low social equity indicators (poverty, rural residence, and no formal
education). Adjusted logistic regression models showed that girl child marriage was significantly
associated with high fertility (AOR=6.62; 95% CI: 3.527-12.429), rapid repeat childbirth (AOR: 2.88;
95% CI: 1.832-4.543), unwanted pregnancy (AOR=2.90; 95% CI: 1.755-4.794), and pregnancy
termination (AOR=1.75; 95% CI: 1.097-2.783). Paper III: Majority (74.5%) of births were from mothers
aged <18 years. Marriage before 18 years increased the likelihood of recent diarrhea among children
born to young mothers (AOR=1.59; 95% CI: 1.18-2.14). Even though maternal child marriage was
associated with children under 5 years of age mortality and infant mortality in unadjusted models,
associations were lost in the adjusted models. We did not find a relation between child marriage and low
birth weight infants, and ARI. Paper IV: About one-third of women aged 15–24 years in Pakistan
reported experiencing controlling behaviors (31.8%) and spousal violence (31.3%) by their husbands.
Compared to adult marriage, child marriage was significantly associated with controlling behaviors
(AOR=1.50; 95% CI: 1.042–2.157), emotional violence (AOR=1.86; 95% CI: 1.254–2.767), and
physical violence (AOR=2.44; 95% CI: 1.582–3.760). Paper V: Nineteen of 20 participants who agreed
to participate were married between 11-17 years. Most respondents were uneducated, poor and were
working as housemaids. The majority participants were unaware of the negative health outcomes of child
marriages. They appeared satisfied by the decision of their parents of marrying them before 18 years,
and even condemned banning child marriages in Pakistan. Strong influence of culture and community
perceptions, varying interpretation of religion, and protecting family honor are some of the reasons that
were narrated by the participants, which seems playing a role in continuation of child marriage practice
in Pakistan.
Conclusions
Interventions such as implementation and enforcement of strict laws prohibiting the practice,
empowering women by increasing education, promoting civil, sexual and reproductive health rights,
creating job opportunities, and enhancing women role in family decision-making can help in tackling
child marriage practice in Pakistan. Significant efforts are needed locally and at governmental level to
abolish the traditional cultural practices, which can directly reduce the high numbers of child marriages
in Pakistan. Efforts to increase the age of marriage and delayed childbearing may have population level
effect in reducing disparities between women married as children and adults, and improving maternal
and child health in Pakistan.