Reconsidering the Basal Proportion of Insulin Dose: Glycemic and Microvascular Outcomes in Type 1 Diabetes Mellitus

Abstract Optimal glycemic control is required to lower the risk of complications in type 1 diabetes mellitus (T1DM). This can be achieved with multiple daily insulin injections (MDI) or with continuous subcutaneous insulin infusion (CSII). Most diabetes guidelines recommend a proportion of basal insulin (basal proportion of total insulin dose; %B/T) around 50% of the total daily dose (TDD), although there is scarce evidence that suggests that a lower %B/T is associated with lower HbA1c levels. Our objective was to evaluate the association of the %B/T with glycemic and microvascular outcomes. We included 132 T1DM adults of the Diabetes Clinic in a tertiary care center, 117 (88.6%) using MDI and 15 (11.4%) using CSII. Data from the medical records and insulin pumps software during outpatient visits were retrospectively collected. Individuals with end-stage renal disease, solid-organ transplant, pregnancy, and glucocorticoid use were excluded. A positive correlation between %B/T and HbA1c levels was found, r=0.26 (p=0.002). Three groups were analyzed according to the %B/T: ≤40%, 41–59% and ≥60%, observing differences in HbA1c concentrations: 7.1% (6.7–8.0%), 7.8% (7.2–9.1%) and 8.7% (7.6–10.2%), respectively (p=0.003). Regarding microvascular complications, the cases of nephropathy were 0 (0%), 23 (30.7%) and 18 (40%) across those groups (p=0.029) even though there was no difference in T1DM duration across groups. There were also no differences in body mass index, TDD, TDD/weight (units/kg/day), nor in the rates of retinopathy or neuropathy. Multiple regression analysis identified %B/T as an independent predictor of the HbA1c concentration. A difference in the rates of hypoglycemic episodes per month was found among individuals with a %B/T ≤50%: 2 (1–5) versus 6 (2.5–12) episodes per month in those having a higher %BT (p=0.002). There are limitations in our study, including the retrospective nature of the analysis, no data about meal content and a low usage of CGM (thus relying on variable self-monitoring of blood glucose). Therefore, we cannot asseverate that lowering the %B/T would improve glycemic and microvascular outcomes. Nevertheless, our findings indicate that the %B/T correlates with HbA1c levels and are consistent with those previously described. It also suggests a relationship with hypoglycemia and to the best of our knowledge, it is the first time that an association between %B/T and nephropathy has been noted.