Tighter glycemic control is associated with ADL physical dependency losses in older patients using sulfonylureas or mitiglinides: Results from the DIMORA study.

BACKGROUND There is growing evidence that tight glycemic control may be more harmful than beneficial in older persons with Type 2 diabetes (T2DM). It remains controversial if tight glycemic control (lower glycated hemoglobin A1c (A1c)) is associated with functional impairments in older frail patients with T2DM. We explored associations between A1c and losses in Activities of Daily Living (ADLs) in diabetic nursing home (NH) patients and tested for differences according to anti-diabetic treatment: diet, anti-diabetic oral drug (AOD), insulin, combined insulin+AOD. METHODS We conducted a cross-sectional study on 1845 older NH patients with T2DM from 150 sites across Italy. Complete evaluations on ADLs, glycemic control, anti-diabetic treatments, comorbidities, and clinical data were recorded. ANOVA was applied to compare clinical characteristics across A1c tertiles. Multivariate regression models evaluated associations between A1c and ADL losses. RESULTS Patients had a mean age [SD]=82 [8] years; BMI=25.5 kg/m(2) [4.7]; Fasting Plasma Glucose (FPG)=7.4 [3.0] mmol/l; Post-prandial glucose (PPG)=10.3 [3.6] mmol/l; A1c=7.0% (54 mmol/mol), ADL losses=3.7 [1.8]. Compared to higher A1c tertiles, patients in the lower tertile had greater ADL losses, were more likely to use AODs, while less likely to use insulin or insulin+AOD. After adjusting for multiple confounders, impairments in ADLs were associated with tighter A1c levels (B=-0.014; p=0.002). Regression models according to anti-diabetic treatment showed that tighter A1c levels continued as independent determinants of ADL losses in patients using AODs (B=-0.023; p=0.001), particularly in those using sulfonylureas (B=-0.043; p<0.001) or mitiglinides (B=-0.044; p=0.050). CONCLUSIONS Tighter glycemic control was associated with ADL physical dependency losses, especially in those using sulfonylureas and mitiglinides.