**Background:** Natural disasters and armed conflicts are known to cause immediate adverse health effects, but their long-term consequences for children exposed in utero or early life are poorly understood, particularly in low- and middle-income countries. The Developmental Origins of Health and Disease (DOHaD) hypothesis suggests that environmental stressors during pregnancy can program offspring health outcomes decades later. Life history theory predicts that under harsh or dangerous conditions, organisms shift energy allocation toward survival and reproduction at the expense of growth and maintenance. Sri Lanka offers a unique setting to study both a discrete natural disaster (the 2004 Indian Ocean tsunami) and a prolonged civil war (1983-2009), with geographically well-demarcated affected regions.
**Methods:** This cross-sectional study recruited 110 adolescents aged 12-13 years (mean 13.2 years, SD 0.3; 42% girls) whose mothers were pregnant during the tsunami on 26 December 2004. Participants were recruited from four regions: Galle (tsunami-only, n=22), Kilinochi and Manmunai (conflict-only, n=35), Mullativu and Koralai Pattu (conflict-plus-tsunami, n=29), and Akmeemana (control, n=24). Exclusion criteria included mothers who were not pregnant during the tsunami (20 additional adolescents excluded). Data collection included anthropometry (height, weight, circumferences, skinfold thickness measured in duplicate/triplicate following standardized protocols), blood pressure (electronic monitor, lowest of two readings), pubertal stage (self-assessed Tanner stages), and venous blood samples (full blood count, fasting glucose, HbA1c, lipids, insulin, cortisol). Socioeconomic status was assessed using the Multidimensional Poverty Index (MPI). WHO Child Growth Standards were used for height-for-age and BMI-for-age z-scores. Statistical analysis used univariable and multivariable regression models comparing each exposure group against controls, adjusted for age, gender, socioeconomic status (square root of MPI), and maternal height. Multiple testing correction used the Benjamini-Hochberg method (adjusted P-value threshold of 0.01 from initial 0.05 across 72 comparisons).
**Key Results:** The sample was generally short and thin compared to WHO standards (mean height-for-age z-score -0.95, mean BMI-for-age z-score -1.18). Only one child was overweight and none were obese. In adjusted models, BMI-for-age z-scores were significantly higher in all exposed groups compared to controls: conflict group 1.3 (95% CI 0.4 to 2.2), tsunami group 1.0 (95% CI 0.2 to 1.9), and conflict-plus-tsunami group 2.0 (95% CI 1.1 to 2.9). Weight was also higher: conflict 6.85 kg (95% CI 1.45 to 12.25), tsunami 5.95 kg (95% CI 0.93 to 10.97), conflict-plus-tsunami 9.30 kg (95% CI 3.82 to 14.78). Skinfold thicknesses were significantly greater in conflict and conflict-plus-tsunami groups: biceps (conflict 3.5 mm, 95% CI 0.9 to 6.3; conflict-plus-tsunami 6.0 mm, 95% CI 3.2 to 8.7), triceps (conflict 3.1 mm, 95% CI 0.3 to 5.9; conflict-plus-tsunami 5.7 mm, 95% CI 2.9 to 8.5), subscapular (conflict 3.6 mm, 95% CI 0.9 to 6.4; conflict-plus-tsunami 5.2 mm, 95% CI 2.4 to 8.0), and suprailiac (conflict 5.1 mm, 95% CI 1.7 to 8.6; conflict-plus-tsunami 7.3 mm, 95% CI 3.8 to 10.8). No significant differences in skinfold thickness were found in the tsunami-only group. Diastolic blood pressure was higher in conflict (5.6 mmHg, 95% CI 0.5 to 10.7) and conflict-plus-tsunami (6.7 mmHg, 95% CI 1.5 to 11.8) groups. Serum insulin was elevated in conflict (4.5 mIU/L, 95% CI 0.5 to 8.4) and conflict-plus-tsunami (4.4 mIU/L, 95% CI 0.4 to 8.3) groups. No significant differences were found in height, head circumference, or most blood biochemistry after adjustment. Pubertal stage was more advanced in conflict-affected groups (mean 2.4 in conflict, 2.2 in conflict-plus-tsunami vs 1.3 in controls).
**Clinical Implications:** This exploratory study suggests that early-life exposure to conflict and natural disasters may be associated with increased adiposity, higher blood pressure, and elevated insulin levels in adolescence, potentially increasing long-term risk of noncommunicable diseases such as type 2 diabetes and cardiovascular disease. The findings support life history theory predictions that prenatal stress shifts resource allocation toward fat storage. The null findings for height and head circumference suggest these growth parameters may be more canalized or that height differences may emerge later after growth cessation. The study's small sample size (n=110) is a major limitation, preventing subgroup analyses by sex and reducing statistical power. After Benjamini-Hochberg correction (adjusted P-value 0.01), anthropometric and skinfold differences remained significant but blood pressure differences did not. The authors caution against overinterpretation and recommend viewing results as composite measures of life history strategy. These findings can inform post-disaster interventions targeting at-risk adolescents and support advocacy for long-term health monitoring following natural disasters and conflicts.