Fetal programming of early-onset type 2 diabetes: a Swedish nationwide cohort and sibling analysis

Abstract

Incidence of early-onset (< 40 years) type 2 diabetes (T2D) is increasing. While multiple risk factors have been identified, particularly obesity and low socioeconomic status, early-life factors are hypothesised to play a role via fetal programming. We investigated sociodemographic and early-life factors in relation to early-onset T2D using a family-based design that accounts for shared genetic and environmental factors. We included 1,814,062 individuals born in Sweden 1983 to 2002 with follow-up data until 2020, and identified early-onset (age 19-39) T2D cases (n = 3505) through National Diabetes, Patient and Prescribed Drug Registers. Perinatal and sociodemographic factors were retrieved from registers. We used a cohort and sibling design, with multivariable-adjusted Cox proportional hazards regression. Sociodemographic factors associated with early-onset T2D included low parental education, single parenthood, younger parental age and non-Swedish origin. The latter association did not remain after mutual adjustment. Regarding perinatal factors, a higher incidence was noted in relation to lower birth weight (hazard ratio 2.38 [95% confidence interval: 1.98-2.87] and 1.43[1.33-1.54] for < 2500 g and 2500-3500 g, respectively, vs 3500-4500 g), small-for-gestational-age (SGA) (2.24[1.96-2.56]), large-for-gestational-age (LGA) (1.19[1.01-1.39]), and maternal obesity (2.34[2.04-2.69]), diabetes (1.59[1.36-1.85]), smoking (1.59[1.48-1.71]), and infection (1.21[1.03-1.41]) during pregnancy. In the sibling analysis, only low birth weight and SGA remained associated with early-onset T2D. Early-onset T2D is associated with sociodemographic and multiple perinatal factors; only growth restriction likely reflects fetal programming, while other perinatal-related associations might involve confounders. This study highlights the need for early-life, targeted strategies to prevent T2D and reduce health inequities.

Keywords: Birth weight; Early-onset type 2 diabetes; Fetal programming; Maternal diabetes; Perinatal factors; Preterm birth.

Fig. 1

Population selection

Fig. 2

Forest plot summarizing the associations between early life exposures and early-onset type 2 diabetes in the full cohort and in the sibling cohort. Results are mutually adjusted for all variables presented in the forest plot except for the size for gestational age. Size for gestational age was adjusted for all variables except gestational age and birth weight