**Background:** Malnutrition during adolescence is associated with poor health outcomes across the life course and intergenerational effects. In South Asia, undernutrition affects >38% of adolescent girls, with limited decline over the past decade and an increasing burden of overweight/obesity. Pakistan has a high burden of malnutrition, costing an estimated US$7.6 billion/year, but nationally representative surveys have traditionally focused on married women 15–49 years, leaving a gap in data on late adolescent girls and unmarried young women. This study aimed to characterise the burden of malnutrition and assess how structural and intermediary social determinants of health (SDoH) contributed to malnutrition among late adolescent and young women in rural Pakistan.
**Methods:** This cross-sectional study used enrolment data from the Matiari emPowerment and Preconception Supplementation (MaPPS) trial, collected from June 2017 to July 2018 in Matiari District, Pakistan. Participants were 25,447 non-pregnant adolescent and young women aged 15–23 years (14,771 aged 15–18.9 years; 10,676 aged 19–23 years), representing 17,044 households. Anthropometric measurements (height, weight, MUAC) were collected in duplicate using standardised methods. WHO-based cut-offs were applied: underweight (BMI <18.5 kg/m²), normal (18.5–24.9 kg/m²), overweight (25–29.9 kg/m²), obese (≥30 kg/m²), and stunting (height <145 cm). Explanatory variables included socioeconomic status, education, occupation, religion, marital status, parity, food security (Household Food Insecurity Access Scale), self-reported health, mental health (DASS-21), self-efficacy (Generalised Self-Efficacy Scale), decision-making autonomy, and food practices. Hierarchical multivariable models were generated separately for late adolescent girls and young women, using a four-level framework adapted from the UNICEF and WHO SDoH frameworks.
**Key Results:** The mean age was 18.8±2.3 years. Nearly half (44.7%) had no formal education; 38.1% were employed outside the home; 22.3% were married; and 13.7% had been pregnant. The prevalence of underweight was 36.9% (95% CI 36.3% to 37.5%), with more late adolescent girls underweight (41.3%) than young women (30.8%) (p<0.001). Overweight/obesity affected 10.2% overall (7.9% overweight, 2.3% obese), with higher prevalence among young women (14.3%) than adolescents (7.2%) (p<0.001). Stunting affected 9.2% (95% CI 8.9% to 9.6%) of participants (adolescents: 10.2%; young women: 7.9%; p<0.001). Among stunted participants, 35.7% were additionally underweight and 7.3% were overweight/obese. Mean HAZ among adolescents was −1.58±0.83 and mean BAZ was −0.67±1.13.
In hierarchical models for BMI categories among late adolescent girls, factors associated with underweight (vs. normal) included: being from a poorer wealth quintile (richest vs. poorest: RRR 0.65, 95% CI 0.56 to 0.74), being unmarried (RRR 1.42, 95% CI 1.28 to 1.59), being food insecure (food secure vs. insecure: RRR 0.91, 95% CI 0.83 to 0.99), reporting poorer health, and having lower self-efficacy. Factors associated with overweight/obese included: being from a higher wealth quintile (richest vs. poorest: RRR 4.65, 95% CI 3.49 to 6.20), being married, being food secure, and being employed in skilled/non-manual labour.
Among young women, underweight was associated with lower wealth quintile (richest vs. poorest: RRR 0.59, 95% CI 0.48 to 0.72), being non-Muslim (Muslim vs. non-Muslim: RRR 0.76, 95% CI 0.61 to 0.93), being unmarried (RRR 1.25, 95% CI 1.11 to 1.42), poorer health, and less decision-making autonomy. Overweight/obesity was associated with higher wealth quintile (richest vs. poorest: RRR 5.24, 95% CI 3.93 to 6.98), being married, and being employed in skilled/non-manual labour.
For stunting, among late adolescent girls, protective factors included any education (secondary or higher vs. none: OR 0.48, 95% CI 0.39 to 0.59), higher wealth quintile (richest vs. poorest: OR 0.42, 95% CI 0.35 to 0.51), food security (OR 0.77, 95% CI 0.66 to 0.90), higher self-efficacy, and greater decision-making autonomy. Among young women, stunting was associated with lower education (secondary or higher vs. none: OR 0.79, 95% CI 0.66 to 0.95), lower wealth quintile, food insecurity (OR 0.73, 95% CI 0.59 to 0.90), and having never been pregnant (OR 1.27, 95% CI 1.10 to 1.47).
**Clinical Implications:** This study reveals a high dual burden of malnutrition among late adolescent and young women in rural Pakistan, with underweight (36.9%) being the most prevalent form but overweight/obesity (10.2%) also emerging. Poverty-related structural factors consistently predicted undernutrition across age groups, while empowerment proxies (self-efficacy, decision-making autonomy) were protective, particularly among adolescents. The findings underscore the need for integrated, multisectoral interventions that address both direct nutritional support and broader structural barriers, including poverty, food insecurity, and limited empowerment. The inclusion of unmarried young women in nutrition programmes is critical, as they represent a previously understudied and vulnerable group. The Government of Pakistan's recent national adolescent nutrition strategy provides a policy framework, but implementation will require coordinated efforts between the public health system and education sector.