Figure 2

Estimated GRS effect sizes within each physical activity (PA) or/and smoking (SM) group. GRS was computed for each individual as the summation of risk alleles across 11 obesity variants, PA was ascertained at the age of 36 years and either at the age of 53 years (for BMI) or 43 years (for WHtR) during interviews with a research nurse, SM was ascertained at the age of 36 years and either at the age of 53 years (for BMI) or 43 years (for WHtR) during interviews with a research nurse. Participants were categorised as being a ‘smoker’ if they currently smoked or a ‘non-smoker’ if they had never smoked or were an ex-smoker. The GRS effect sizes in panels a and b are estimated from the models shown in Supplementary Table 1, the GRS effect sizes in panels c and d are estimated from the models shown in Supplementary Table 2 and the GRS effect sizes in panels e and f are estimated from the models shown in Supplementary Table 3. Results for participants who changed group (e.g., active at the age of 36 years but inactive at the age of 53 years) are not shown because this was not explicitly modelled (i.e., there was no parameter for participants who changed groups). However, it is implicit that the trajectories for individuals who changed group would diverge at the age of 53 years (for BMI) or 43 years (for WHtR) from the trajectory of the old group (e.g., active) to follow the gradient of the trajectory for the new group (e.g., inactive).