Obesity is associated with prediabetes, a risk factor for
developing T2D. We investigated the effects of liraglutide 3.0mg, as
adjunct to diet and exercise (D&E), on weight loss, prediabetes prevalence
and onset of T2D over 56 weeks. Effects of liraglutide cessation were
investigated in a 12‑week re-randomized period.
Adults (BMI ≥27kg/m2 with ≥1 comorbidity or
≥30kg/m2) received 500kcal/day deficit diet and exercise. Randomization
was 2:1 to once-daily s.c. liraglutide 3.0 mg (n=2487) or placebo (n=1244),
stratified by prediabetes status. At week 56, individuals without prediabetes on
liraglutide were re-randomized 1:1 to liraglutide or placebo (D&E continued).
Clinicaltrials.gov ID: NCT01272219.
Of 3731 individuals (age 45.1 years, 78.5% female, weight
106.2kg, BMI 38.3kg/m2, 61.2% with prediabetes), 71.9% on liraglutide
and 64.4% on placebo completed study. Data are observed means/proportions from the full analysis set with
LOCF. Statistical analysis is ANCOVA (continuous endpoints) or logistic
regression (categorical endpoints).
After 56 weeks, individuals on liraglutide lost 8.0%
(8.4kg) weight versus 2.6% (2.8kg) on placebo (estimated treatment difference [ETD]
5.4% [5.6kg], p<0.0001). Liraglutide improved fasting and post-load glycaemia
versus placebo (ETD FPG -0.38mmol/L, PG [OGTT, AUC] 2.02 h*mmol/L, HbA1c ‑0.23%-points; p<0.0001 for all).
More individuals reverted to normoglycaemia on liraglutide than placebo (69.2%
vs. 32.7%; odds ratio [OR] 4.85, p<0.0001). More individuals taking placebo
progressed to prediabetes (20.7% vs. 7.2%; OR 3.3, p<0.0001) or developed
T2D (n=14, 1.3 events/100 patient years of exposure [PYE]) than liraglutide (n=4,
0.2 events/100 PYE; OR 8.06, p=0.0003).
From week 56–68, individuals re-randomized to placebo
regained more weight (2.9% vs. 0.7%; ETD 2.2%, p<0.0001) and progressed to
prediabetes (8.0–22.4% vs. 9.1–8.6%; p<0.0001)
than individuals on liraglutide. No-one developed T2D.
Consistent with effects on weight and glycaemia, liraglutide 3.0mg
was superior to placebo in reducing prevalence of prediabetes and T2D.
Continued treatment was required to sustain these effects.