Background Early-life experiences affect later health outcomes such as obesity and hypertension. However, research tends to investigate single exposure-outcome relationships. Ideally, research should investigate joint causality of multiple determinants of ill health to design effective preventive interventions. We aimed to identify exposures across five pre-hypothesised early-life domains, explore their association to the risk of obesity and hypertension and model hypothetical prevention scenarios.

Methods We used data from 17,415 participants in the 1958 National Child Development Study (NCDS) and 17,196 participants in the 1970 British Cohort Study (BCS70). The outcome was measured obesity (BMI over 30) and measured or self-reported (BCS70 only) hypertension (BP over 140/90 mmHg) at ages 42 (NCDS) and 46 (BCS70). Early-life domains included: ‘antenatal, neonatal and birth’, ‘developmental attributes, ‘education’, ‘socioeconomic’ and ‘parental and family environment’. Confounders included parental separation, parental death and ethnicity. Logistic regression tested the association between domain risk scores and combined obesity-hypertension. Adjusted population attributable fractions (PAFs) explored the reduction in the outcome risk if domain scores were reduced.

Results In unadjusted models, greater risk scores across the five domains were associated with increased risk of obesity-hypertension. For the NCDS, in adjusted models, there remained a significant association between scoring 3+ in the ‘antenatal, neonatal and birth’ domain (RR1.81 95%CI 1.17-2.77) and scoring 4+ in the ‘socioeconomic’ domain (RR1.91 95%CI 1.19-3.06) and obesity-hypertension. For the BCS70, there remained a significant association between scoring 4 or 5+ in the ‘education’ domain (4 RR1.82 95%CI 1.16-2.85; 5+ RR1.65 95%CI 1.09-2.50) and obesity-hypertension. Adjusted PAFs based on NCDS data suggested a 44% (95%CI 15-63%) reduction in obesity-hypertension if those with the highest risk score in the ‘antenatal, neonatal, and birth’ domain had a risk score of 0. A reduction of 48% (95%CI 16-68%) in obesity-hypertension may be achieved if those with the highest risk score in the ‘socioeconomic’ domain had a risk score of 0. BCS70 data suggested for the ‘education’ domain a 39% (95%CI 8-60%) and 36% (95%CI 2-58%) reduction in obesity-hypertension if those with a risk score of 5+ had a risk score of 0 or 1. There could be a further 42% (95%CI 15-60%) and 39% (95%CI 10-58%) reduction in obesity-hypertension if those with a risk score of 4 had a risk score of 0 or 1.

Conclusion Targeted prevention interventions aimed at certain early-life domains could have an impact on obesity-hypertension prevalence, particularly for those with high risk scores in childhood.