Relative risk of mortality associated with diabetes as a function of birth weight.

OBJECTIVE

Birth weight is a risk factor for both diabetes and mortality. Diabetes is a risk factor for mortality. Whether the excess mortality observed for diabetes varies with birth weight is unclear.

RESEARCH DESIGN AND

METHODS

Among all 2,508 Rochester, Minnesota, residents who first met research criteria for adult-onset diabetes in 1960-1995, 171 were born locally in-hospital after 1922 (i.e., birth weights available) as singleton, term infants. Each case subject and two age- and sex-matched nondiabetic control subjects (born locally, residing locally when the case subject met the criteria for diabetes) were followed through 31 December 2000 for vital status.

RESULTS

Of the diabetic case subjects, 16% (27 of 171) died vs. 7% (25 of 342) of control subjects (P = 0.004). The difference was less for normal-birth-weight (NBW) (2,948-<3,856 g) individuals (12% [12 of 102] vs. 8% [20 of 246], P = 0.31) than for abnormal-birth-weight individuals (low birth weight [LBW] 20% [8 of 39] vs. 2% [1 of 46], P = 0.01; high birth weight [HBW] 23% [7 of 30] vs. 8% [4 of 50], P = 0.16), as confirmed with age- and sex-adjusted Cox proportional hazards (diabetes-associated hazard ratio 1.4 [95% CI 0.69-2.90] for NBW vs. 4.8 [1.7-13.3] for abnormal birth weight, test for interaction P = 0.056). The observed diabetes deaths were greater than expected, based on mortality for the general population (27 vs. 13.3, P < 0.001), with 70% of excess deaths occurring among LBW (8 vs. 2.2, P < 0.001) and HBW (7 vs. 3.1, P = 0.03) individuals.

CONCLUSIONS

The excess mortality observed for diabetes appears disproportionately concentrated among abnormal-birth-weight individuals, thus identifying a subset of at-risk diabetic individuals and reinforcing the importance of NBW deliveries.