This study quantifies the causal effect of birth weight on cardiovascular biomarkers in adulthood using the Cardiovascular Risk in Young Finns Study (YFS). We apply a multivariable Mendelian randomization (MVMR) method that provides a novel approach to improve inference in causal analysis based on a mediation framework. The results show that birth weight is linked to triglyceride levels (β = −0.294; 95% CI [−0.591, 0.003]) but not to low-density lipoprotein (LDL) cholesterol levels (β = 0.007; 95% CI [−0.168, 0.183]). The total effect of birth weight on triglyceride levels is partly offset by a mediation pathway linking birth weight to adult BMI (β = 0.111; 95% CI [−0.013, 0.234]). The negative total effect is consistent with the fetal programming hypothesis. The positive indirect effect via adult BMI highlights the persistence of body weight throughout a person's life and the adverse effects of high BMI on health. The results are consistent with previous findings that both low birth weight and weight gain increase health risks in adulthood.