Effects of Aerobic and Resistance Exercise on Glycemic Control
Objective: To compare the effectiveness of aerobic training, resistance training, or a combination of both on hemoglobin (Hb)A1c values (mean plasma glucose concentration over the previous 2 to 3 months) in the management of type 2 diabetes.
Design: Randomized, controlled, 4-arm trial of 26-weeks duration, with power to show a moderate difference between groups. Outcomes were assessed blindly.
Setting: Community study that took place from 1999 through 2003 at 8 exercise facilities in Ontario, Canada.
Participants: Participants were recruited through advertising and physician referrals. Inclusion criteria were type 2 diabetes for >6 months, HbA1c values of 6.6% to 9.9%, and age, 39 to 70 years. Exclusion criteria were current insulin therapy; participation in regular exercise or in resistance training during the previous 6 months; recent changes in medications or body weight (≥5%); serum creatinine level of ≥200 μmol/L; proteinuria ≥1 g/d; blood pressure >160/95 mm Hg; and reasons to restrict physical activity. Participants were randomized if they attended >10 of 12 exercise sessions during 4 run-in weeks. Of 2282 persons screened, 258 entered the run-in phase, and 251 were randomized (mean age, 54 years; duration of diabetes, 5.35 y; HbA1c, 7.68%; 64% were men).
Intervention: Participants were allocated to no-cost supervised aerobic training, resistance training, or combined exercise training at a community exercise facility, or to control. Participants exercised 3 times per week with progression in duration and intensity. Aerobic training was on treadmills or bicycle ergometers with time increasing to 45 minutes and heart rate increasing to 75% max. Resistance training comprised 7 exercises on weight machines, progressing to 2 or 3 sets of 7 to 9 repetitions of each exercise. The combined exercise group did both types of exercise, thus their total duration of exercise sessions was approximately 90 min. Median exercise training adherence was ≥80%. Control participants were asked to revert to prestudy exercise levels and were promised an exercise program after the study.
Main outcome measures: The primary outcome was change in HbA1c values after 6 months. Changes in plasma lipid levels, blood pressure, and body composition were also measured. Follow-up was 82%.
Main results: In intention-to-treat adjusted linear mixed-effects analyses, mean HbA1c fell for each exercise group and rose slightly for the control group. The differences in change between the aerobic group and the resistance group compared with the control group were -0.51 (95% CI, -0.87 to -0.14) and -0.38 (CI, -0.72 to -0.22), respectively. The changes were also greater in the combined exercise group versus the aerobic group and the resistance group (-0.46; CI, -0.83 to -0.09 and -0.59; CI, -0.95 to -0.23, respectively). Among participants with a baseline HbA1c ≥7.5 (median value) the decreases were greater than for those with a baseline HbA1c <7.5 (P < 0.001). Changes in plasma lipid levels and blood pressure did not differ by group. Some body composition measurements decreased more in the exercise groups than the control group, but change in abdominal visceral fat did not differ between groups. The aerobic group reported 4 serious adverse effects (2 persons with worsening osteoarthritis, 1 person with angina, and 1 person with spinal stenosis). There were no severe episodes of hypoglycemia, and 10 exercising participants and 1 control participant reported mild hypoglycemia. Overall, adverse events occurred in more exercise participants than control participants (38% vs 14%; P = 0.001), with musculoskeletal injury or discomfort requiring modification of activity occurring in 26% of the exercise participants versus 14% of control participants (P = 0.059).