Skip to main content
. Author manuscript; available in PMC: 2017 Nov 18.
Published in final edited form as: N Engl J Med. 2017 May 18;376(20):1943–1955. doi: 10.1056/NEJMoa1616338

Table 2.

Effect of Specific Exercise Modes, Added to Diet-Induced Weight Loss, on Outcomes.*


Outcome		 Control
(N =40)		 Aerobic
(N = 40)		 Resistance
(N = 40)		 Combination
(N =40)		 P Value†
Group–Time
Interaction		 Aerobic
vs.
Control		 Resistance
vs.
Control		 Aerobic
vs.
Resistance		 Combination
vs.
Aerobic		 Combination
vs.
Resistance
Primary outcome
PPT score
 Baseline   28.6±0.5     29.3±0.3     28.8±0.4     27.9±0.4
 Change at 3 mo     0.4±0.3       2.9±0.4‡       3.2±0.4‡       4.0±0.4‡
 Change at 6 mo     1.0±0.4       3.9±0.4‡       3.9±0.4‡       5.5±0.4‡ <0.001 <0.001 <0.001   0.87   0.002 0.004
Secondary outcomes
Other frailty measures
 VO2peak (ml/kg/min)
  Baseline   17.0±0.5     17.6±0.5     17.0±0.6     17.2±0.6
  Change at 6 mo     0.1±0.3       3.3±0.3‡       1.3±0.3‡       3.1±0.3‡ <0.001 <0.001   0.007 <0.001   0.63 0.001
 FSQ score
  Baseline   29.8±0.5     30.1±0.5     29.3±0.6     29.8±0.6
  Change at 6 mo     0.4±0.3       2.0±0.3‡       2.3±0.3‡       3.6±0.3‡ <0.001   0.002 <0.001   0.46   0.005 0.03
Body composition
 Body weight (kg)
  Baseline   97.9±2.9     96.9±2.3   101.8±2.9     99.0±2.9
  Change at 6 mo   −0.9±0.5     −9.0±0.6‡     −8.5±0.5‡     −8.5±0.5‡ <0.001 <0.001 <0.001   0.76§   0.72§ 0.96§
 Lean mass (kg)
  Baseline   54.9±2.3     55.0±1.9     58.1±2.3     56.5±1.8
  Change at 6 mo     0.0±0.2     −2.7±0.3‡     −1.0±0.3     −1.7±0.3‡ <0.001 <0.001   0.03   0.001   0.047 0.20
 Fat mass (kg)
  Baseline   43.0±1.5     41.9±1.3     44.3±1.5     42.5±1.6
  Change at 6 mo   −0.9±0.4     −6.3±0.5‡     −7.3±0.4‡     −7.0±0.5‡ <0.001 <0.001 <0.001   0.18§   0.36§ 0.67§
 Thigh muscle (cm3)
  Baseline  1302±63    1234±62    1190±48    1186±66
  Change at 6 mo      10±7      −77±7‡      −23±7      −40±7‡ <0.001 <0.001   0.008 <0.001   0.005 0.21
 Thigh fat (cm3)
  Baseline  1774±132    1700±98    1848±108    1784±125
  Change at 6 mo      −2±36    −260±35‡    −280±35‡    −288±35‡ <0.001 <0.001 <0.001   0.64§   0.61§ 0.97§
BMD at total hip (g/cm2)
  Baseline 1.031±0.025   1.018±0.019   1.047±0.022   1.010±0.025
  Change at 6 mo 0.004±0.004 −0.027±0.004‡ −0.006±0.004 −0.014±0.004 0.001 <.001   0.37   0.005   0.04 0.43
Strength, balance, and gait
 Total 1RM (kg)║
  Baseline    269±21      265±20      288±30      272±16
  Change at 6 mo        2±5       5.0±5        49±5‡        48±5‡ <0.001   0.99 <0.001 <0.001 <0.001 0.82
 Obstacle course (sec)
  Baseline   15.9±0.7     15.5±0.8     16.4±0.6     17.0±1.0
  Change at 6 mo     0.0±0.3     −1.5±0.4‡     −2.2±0.3‡     −2.9±0.3‡ <0.001   0.02 <0.001   0.20   0.01 0.20
 Gait speed (m/min)
  Baseline   75.0±2.6     74.6±2.0     74.3±2.3     68.8±2.2
  Change at 6 mo   −0.5±1.3       8.1±1.3‡       9.3±1.3‡     12.1±1.3‡ <0.001 <0.001 <0.001   0.56   0.03 0.09
*

Plus-minus values for the change scores are the least-squares adjusted means ±SE from the repeated-measures analyses of variance; plus-minus values for the baseline values are the observed means ±SE. Scores on the Physical Performance Test (PPT) (primary outcome) range from 0 to 36, with higher scores indicating better physical function; the minimal clinically important difference is 1.8. Peak oxygen consumption (VO2peak) was assessed during graded treadmill walking. Scores on the Functional Status Questionnaire (FSQ) range from 0 to 36, with higher scores indicating better function. BMD denotes bone mineral density.

†

P values for the changes from baseline to 6 months in between-group comparisons were calculated with the use of mixed-model repeated-measures analyses of variance (with baseline values and sex as covariates) and are reported when the overall P value was lower than 0.05 for the interaction among the four groups over time. In a Bonferroni correction to adjust for the multiple comparisons in the PPT score (in which the P values were multiplied by 5 for the comparison with an alpha level of 0.05), the corrected P values were 0.01 for the combination group versus the aerobic group and 0.02 for the combination group versus the resistance group. In accordance with a gatekeeping strategy, a significant group-by-time interaction (P<0.01) and at least one significant difference between an exercise group and the control group and at least one significant difference among the exercise groups in the change in PPT score were required to continue to testing of the secondary outcomes; comparisons of the exercise groups with the control group were performed with Dunnett’s test and comparisons among the intervention groups were performed with the Fisher–Hayter test. Secondary analyses included a comparison between the combination group and the control group; all P values were less than 0.05.

‡

P<0.001 for the comparison of the value at the follow-up time with the baseline value within the group, as calculated with the use of mixed-model repeated-measures analysis of variance.

§

The between-group differences were not significant in the last three columns because the group-by-time interaction for the three intervention groups (step 1 of the Fisher–Hayter test) was not significant; the family-wise error rate of 0.05 or less is maintained for the Fisher–Hayter multiple comparison procedure.

P<0.01 for the comparison of the value at the follow-up time with the baseline value within the group, as calculated with the use of mixed-model repeated-measures analysis of variance.

║

Total one-repetition maximum (1RM) is the total of the maximum weight a participant can lift, in one attempt, in the biceps curl, bench press, seated row, knee extension, knee flexion, and leg press.