Genetic and Environmental Factors and Cardiovascular Disease Risk in Adolescents

Abstract

Importance: Cardiovascular risk factors in youth have been associated with future cardiovascular disease (CVD), but conventional observational studies are vulnerable to genetic and environmental confounding.

Objective: To examine the role of genetic and environmental factors shared by full siblings in the association of adolescent cardiovascular risk factors with future CVD.

Design, setting, and participants: This is a nationwide cohort study with full sibling comparisons. All men who underwent mandatory military conscription examinations in Sweden between 1972 and 1995 were followed up until December 31, 2016. Data analysis was performed from May 1 to November 10, 2022.

Exposures: Body mass index (BMI), cardiorespiratory fitness, blood pressure, handgrip strength, and a combined risk z score in late adolescence.

Main outcomes and measures: The primary outcome was fatal or nonfatal CVD, as recorded in the National Inpatient Register or the Cause of Death Register before 2017.

Results: A total of 1 138 833 men (mean [SD] age, 18.3 [0.8] years), of whom 463 995 were full brothers, were followed up for a median (IQR) of 32.1 (26.7-37.7) years, during which 48 606 experienced a CVD outcome (18 598 among full brothers). All risk factors were associated with CVD, but the effect of controlling for unobserved genetic and environmental factors shared by full siblings varied. In the sibling analysis, hazard ratios for CVD (top vs bottom decile) were 2.10 (95% CI, 1.90-2.32) for BMI, 0.77 (95% CI, 0.68-0.88) for cardiorespiratory fitness, 1.45 (95% CI, 1.32-1.60) for systolic blood pressure, 0.90 (95% CI, 0.82-0.99) for handgrip strength, and 2.19 (95% CI, 1.96-2.46) for the combined z score. The percentage attenuation in these hazard ratios in the sibling vs total cohort analysis ranged from 1.1% for handgrip strength to 40.0% for cardiorespiratory fitness. Consequently, in the sibling analysis, the difference in cumulative CVD incidence at age 60 years (top vs bottom decile) was 7.2% (95% CI, 5.9%-8.6%) for BMI and 1.8% (95% CI, 1.0%-2.5%) for cardiorespiratory fitness. Similarly, in the sibling analysis, hypothetically shifting everyone in the worst deciles of BMI to the middle decile would prevent 14.9% of CVD at age 60 years, whereas the corresponding number for cardiorespiratory fitness was 5.3%.

Conclusions and relevance: In this Swedish national cohort study, cardiovascular risk factors in late adolescence, especially a high BMI, were important targets for CVD prevention, independently of unobserved genetic and environmental factors shared by full siblings. However, the role of adolescent cardiorespiratory fitness in CVD may have been overstated by conventional observational studies.

Figure 1.. Hazard Ratios (HRs) for Cardiovascular Disease by Risk Factors in the Total Cohort and Sibling Cohort

Estimates were obtained using flexible parametric survival models, by which the associations across the restricted cubic splines of exposures were computed with the median value as the referent. Shaded areas denote 95% CIs. The models were adjusted for birth cohort, conscription cohort, and conscription office. For graphical purposes, the x-axis was limited to span from the first to the 99th percentile of the exposure distribution. The combined risk z score represents the unweighted mean across exposure z scores.

Figure 2.. Population-Attributable Fraction for Cardiovascular Disease (CVD) by Risk Factors in the Total Cohort and Sibling Cohort, When Considering a Major or Moderate Intervention

The major intervention entails shifting everyone to the best decile, whereas the moderate intervention involves shifting everyone from the worst to the middle decile. The combined z score represents the unweighted mean across exposure z scores (see the Methods section). Error bars indicate 95% CIs.