Do Weight trajectories influence diabetes control? A prospective study in Switzerland (CoLaus study)

Identify anthropometric trajectories among subjects with type 2 diabetes mellitus (T2DM), and associate them with glycaemic control.

Prospective study including 268 community-dwelling participants with T2DM (34% women, mean age 68.7 ± 8.9 years) followed for 10.7 years (range: 8.8-13.6 years). T2DM control was considered for 1) fasting plasma glucose (FPG) < 7.0 mmol/L, or 2) HbA₁c < 7.0% (53 nmol/mol). Changes in weight or waist and weight variability were considered.

One half (FPG) and one third (HbA₁c) of participants presented with uncontrolled T2DM. Half of the participants presented with obesity and 75% with abdominal obesity. During follow-up, half of the participants maintained their weight, 25% gained > 5 kg, and 25% lost < 5 kg; almost half increased their waist by > 5 cm. Using FPG as criterion, participants who lost > 5 cm waist were more likely to be controlled: multivariable-adjusted odds ratio (OR) and 95% confidence interval (CI): 3.10 (1.23-7.78). Participants with controlled T2DM also presented with a higher weight variability: multivariable adjusted mean ± standard error 4.8 ± 0.3 vs. 3.9 ± 0.3 kg, p = 0.028. Using HbA₁c as criterion, participants who lost > 5 kg were less likely to be controlled: OR and (95% CI): 0.35 (0.18-0.66). Similar findings were obtained when restricting the analysis to participants who were diabetic throughout the whole study period.

In a Swiss community-based sample of participants with T2DM, T2DM control rates could be implemented. Neither weight nor waist variability was significantly and consistently associated with T2DM control.