Can inaccuracy of reported parental history of diabetes explain the maternal transmission hypothesis for diabetes?

Background The mode of inheritance of type 2 diabetes mellitus is still under discussion. Several studies have suggested an excess maternal transmission, however, more recent studies could not always confirm these findings. Methods We investigated the frequency of a maternal and paternal history of diabetes among diabetic and non-diabetic subjects and assessed the association between diabetes and a parental history of diabetes among participants of the MONICA Augsburg study. As an extension to previous studies, unknown parental status was taken into account. Results Of the 542 diabetic probands, 25.3% reported a positive maternal history of diabetes and 10.9% reported a positive paternal history of diabetes. Among the 12 209 non-diabetic participants a positive maternal history was also more common than a positive paternal history (12.5% versus 7.1%). Conversely, an unknown paternal status was more common than an unknown maternal status in both groups (diabetic subjects: 27.9% versus 16.8%, non-diabetic subjects: 16.8% versus 8.4%). Adjusted odds ratios (OR) for the association between a parental history of diabetes and diabetes status were similar for a positive maternal (OR = 2.9, 95% CI: 2.3-3.6) and paternal history (OR = 2.8, 95% Cl : 2.1-3.8) and for an unknown maternal (OR 1.3, 95% CI: 1.0-1.8) and paternal history (OR = 1.5, 95% CI : 1.2-1.9). Conclusion Our findings do not support a strong excess maternal transmission of diabetes. Epidemiological biases and failure to account for 'don't know' responses may in part explain the previously observed predominance of a maternal history of diabetes.