(071) Sex Disparities in Medication Changes After Metformin Initiation are Associated With Small, Clinically Non-Relevant, Sex Differences in HbA1c Levels During Follow-up

Background: Sex disparities in medication changes after the initiation of metformin in people with type 2 diabetes have previously been shown. To what extent these sex disparities in medication changes are accompanied by differences in diabetes control, i.e., glycated haemoglobin (HbA1c) levels, remains unknown.

Objectives: To assess whether there are differences between men and women in their HbA1c level after a medication change following initial metformin treatment.

Methods: In a previous cohort study using data from the Groningen Initiative to ANalyze Type 2 diabetes Treatment (GIANTT) database, we identified people with type 2 diabetes who had a medication change during two to five years of follow-up after metformin initiation. Medication change was defined as (1) an addition of any other non-insulin hypoglycaemic agent (OHA), (2) switching from metformin to any OHA, (3) initiation of any OHA after a period of >270 days of metformin discontinuation (OHA initiation), or (4) insulin initiation. The outcome was the HbA1c level at the first measurement between three to nine months after medication change. The determinant was the sex of the patients. Linear regression analyses were conducted on the whole cohort and by each type of medication change, adjusted for age, history of a cardiovascular disease (yes/no), and HbA1c level at the time of medication change.

Results: The HbA1c level of the 4,184 included patients (48% women) was lower at follow-up than at the time of medication change (decrease in median from 7.5% to 6.7%). Both women and men had a reduction at follow-up (decrease in median from 7.4% to 6.8% in women, and from 7.6% to 6.7% in men). In the whole cohort, the adjusted HbA1c level at follow-up was slightly higher for women (β = 0.06, 95% CI: 0.00–0.11, P=0.04). Slightly higher adjusted HbA1c levels in women were also observed in each of the separately analysed specific types of medication change (range β = 0.01 [95% CI: -0.23–0.25] for OHA initiation to β = 0.19 [95% CI: -0.15–0.54] for insulin start). These differences were not statistically significant (P>0.05).

Conclusions: We observed a small, clinically non-relevant, difference between men and women in their HbA1c level at follow-up after a medication change following initial metformin treatment. This suggests that the observed sex disparities in medication changes do not result in meaningful differences in diabetes control.