All-cause mortality and pharmacological treatment intensity following a high risk screening program for diabetes. A 6.6 year follow-up of the ADDITION study, Denmark

Aim: To study all-cause mortality and pharmacological treatment intensity in relation to baseline glucose metabolism and HbA1c following high risk screening for diabetes in primary care. Methods: Persons aged 40-69 years (N = 163,185) received mailed diabetes risk questionnaires. 20,916 persons without diabetes but with high risk of diabetes were stratified by glucose metabolism (normal glucose tolerance (NGT), dysglycemia (IFG or IGT) or diabetes) and by HbA1c at screening (<6%, 6.0-6.4% or ≥6.5%). Median follow-up was 6.6 years. Excess mortality was calculated by hazard ratio. Results: HR for all-cause mortality increased with increasing levels of HbA1c at screening in people with NGT and dysglycemia. In people with screen detected diabetes the opposite relation was found. In people with diabetes redeemed prescription rates for lipid-, blood pressure- and glucose-lowering drugs increased significantly following screening and prescription rates increased with increasing levels of HbA1c at screening. The same trend in redeemed prescriptions was seen for people with dysglycemia and NGT, but the absolute rates were significantly lower than those among people with screen detected diabetes. Conclusions: This study confirms HbA1c as an independent predictor of all-cause mortality in non-diabetic individuals. A likely explanation for the inverse relation found between all-cause mortality and HbA1c at screening among those with screen detected diabetes would be that intensive treatment near-normalizes mortality. The small group of people with NGT and HbA1c ≥ 6.5%, who had the highest all-cause mortality, may benefit from being labelled and treated as having diabetes although this group may have special characteristics not accounted for in this study.