Table 4

Secondary outcomes

Author, publication dateNumber of studies included in synthesisOutcomes
Dietary and physical activity behaviours
Cardona-Morrell, 20103 studies3/12 studies reviewed reported on changes in fat and fibre intake. Substantial improvements demonstrated in one trial only, which reported half the participants meeting fibre and total fat intake goals, and a third achieving saturated fat goal.
Gillett, 20128 studiesAuthors concluded that adherence to lifestyle measures could be problematic and that compliance was variable. Noted that benefits of intervention greatest among those with the highest compliance and highest lifestyle target achievement. In one study (the DPS),1 strong inverse correlation between progression to diabetes and achievement of lifestyle targets noted. Suggested that relatively long duration of intervention (eg, 4 years of DPS) potentially necessary for lasting intervention impact.
Schellenberg, 20134 studiesAuthors concluded that most studies reported positive effects on physical activity and dietary intake. However, results not always statistically or clinically significant or sustained after end of active intervention.
Measures of adiposity
Ashra, 201520 studiesPooled mean weight difference observed in intervention, cf. control arms of 20 RCTs at 12–18 months=−1.57 kg (95% CI −2.28 to −0.86). Weight change difference at >18 months (n=11 RCTs)=-1.26 kg (−2.35 to −0.18).
Baker, 20116 studiesIn the 6/7 studies for which impact on BMI could be calculated, endline BMI differences between intervention and control conditions consistently favoured the intervention, but effect sizes were small: range −0.05 kg/m2 (observed in IDPP) to −0.43 kg/m2 (observed in VIP). No tests for differences between conditions reported.
Balk, 201524 studiesAll studies observed net weight loss associated with intervention, of between 0.2% and 10.5% of initial body weight (summary net change=−2.2%, 95% CI −2.9 to −1.4)
Cardonna-Morrell, 20104 studies
2 studies
Meta-analysis outcomes suggest significant mean weight loss at 12 months associated with intervention: summary difference=−1.82 kg (95% CI −2.7 to −0.99)
Pooled mean waist circumference measurement reduction significantly greater in treated vs control groups: mean difference=−4.6 cm (−5.8 to −3.4)
Gillett, 20125 systematic reviews, 9 RCTsNo summary weight change outcomes reported, but authors note that there was a tendency for weight to be regained soon after end of intervention. This did not occur in one study (DPS), and it was hypothesised that the duration of intervention (DPS=4 years) may be relevant to persistence of weight change.
Glechner, 20153 studiesMeta-analysis results suggest net mean weight difference associated with intervention, cf. control condition at 1 year=−2.44 kg (95% CI −3.45 to −1.43). Results consistent at 3 years: net weight difference=−2.45 kg (−3.56 to −1.33).
Norris, 20056 studiesAt 1-year follow-up, the pooled estimate from four studies suggested additional weight loss of 2.8 kg (95% CI 4.7 to 1.0) in intervention, cf. control scenario, and net difference in BMI (three studies)=−1.3 kg/m2 (−1.9 to −0.8). At two-year follow-up, the net weight difference associated with intervention, cf. control scenario=−2.6 kg (−3.3 to −1.9; 3 studies).
Orozco, 20087 studiesMeta-analysis results suggested net BMI reduction associated with intervention=−1.1 kg/m2 (95% CI −2.0 to −0.2; 6 studies). Results for weight also indicated additional weight loss in the intervention group: summary net change=−2.7 kg (−4.7 to −0.7; 7 studies). No significant between-group difference observed for waist–hip ratio: summary difference=−0.01 (−0.02 to 0.01; 4 studies).
Schellenberg, 20138 studiesAuthors state that most studies reported positive effects on body composition. However, results not always significant or sustained after end of active intervention.
Yoon, 20135 studiesTwo studies showed significant reduction in BMI associated with intervention, cf. control scenario (SLIM study: −0.36±1.47 kg/m2 in the intervention group, 0.08±1.80 in the control group, p=0.014; DPS study: −1.3±1.9 kg/m2 in the intervention group, −0.3±2.0 in the control group, p=<0.0001). In these and two further studies, significant weight loss observed in intervention groups in preintervention and postintervention comparisons. Significant weight increase observed in intervention and control groups of IDPP, but further details not reported.
Microvascular disease
Balk, 20151 studyThe Da Qing study reported a reduction in severe retinopathy at 20-year follow-up associated with intervention, cf. control condition (HR=0.53, 95% CI 0.29 to 0.99). Limited evidence suggested no significant effects on nephropathy or neuropathy.
Schellenberg, 20131 studyThe Da Qing study reported no effect on nephropathy or neuropathy at 20-year follow-up. However, incidence of severe retinopathy was 47% lower in intervention, cf. control participants. Authors commented that loss to follow-up was high and that many participants did not have formal retinal examinations. Hence, they considered the strength of evidence insufficient to draw conclusions.
Macrovascular disease
Balk, 20152 studiesAuthors commented that there was no consistent pattern in cardiovascular mortality outcomes. The Da Qing study observed no difference at 20-year follow-up (HR 0.83; 95% CI 0.48 to 1.40). In the DPP, no significant effect on cardiovascular mortality was observed at 3-year follow-up (RR 0.50; 0.09 to 2.73).
Gillett, 20124 studiesAuthors concluded that studies with long durations of follow-up demonstrated disappointing CVD outcomes.
Hopper, 20112 studiesNon-significant trend towards reduction in cardiovascular mortality in meta-analysis of Da Qing study (20-year follow-up) and DPP (2.8-year follow-up) studies (RR 0.70, 95% CI 0.46 to 1.07).
Schellenberg, 20132 studiesNo differences in CVD event rates between intervention and control groups noted at 10-year follow-up of DPS (RR=1.02, 95% CI 0.73 to 1.42), or the 6-year or 20-year follow-ups of the Da Qing study (at 6-year follow-up, HR=0.96; 0.76 to 1.44; at 20 years, HR=0.98, 0.71 to 1.37). Authors conclude that strength of evidence is insufficient to determine whether lifestyle interventions impact on CVD event rates.
Yoon, 20132 studiesNo differences in CVD event or mortality rates between intervention and control groups noted at 20-year follow-up of Da Qing study (for event rates, HR=0.98; 95% CI 0.71 to 1.37; for mortality rates, HR=0.83, 0.48 to 1.40). Only small number of CVD events observed at the three-year follow-up of the IDPP (n=4 in the intervention group, 2 in the control group).
Quality-adjusted life years
Yoon, 20130 studiesInvestigated and noted that no primary study reported on quality-adjusted life years.
All-cause mortality
Balk, 20153 studiesThe 23-year follow-up data from the Da Qing study were indicative of lower risk of mortality in the intervention vs control arms (HR=0.71; 95% CI 0.51 to 0.99). This effect was restricted to women and not significant at earlier time points. No similar effect was observed at the 20-year follow-up or for men. No impact on all-cause mortality was observed at the 3-year follow-up of the DPP or 10-year follow-up of the DPS.
Hopper, 20114 studiesNo impact of lifestyle intervention on all-cause mortality observed in meta-analysis (RR 0.81, 95% CI 0.61 to 1.09). (Studies considered=DPS 10-year follow-up, Da Qing study 20-year follow-up, DPP 2.8-year follow-up, IDPP 2.5-year follow-up).
Orozco, 20084 studiesAuthors commented that all-cause mortality rates were comparable between the intervention and control groups. (Studies considered=Da Qing study 6-year follow-up, DPP 2.8-year follow-up, IDPP 2.5-year follow-up and the 2-year follow-up of a regional UK-based study).
Yoon, 20131 studyOf studies reviewed, only the Da Qing study reported on mortality rate. No significant difference in overall mortality rate between the intervention and control group observed at 20-year follow-up (HR 0.96, 95% CI 0.65 to 1.41).
  • Results relating to secondary outcomes are listed for each review, as relevant, alongside the number of primary studies drawn on in the associated syntheses. Italicised entries are those assigned AMSTAR scores <8, excluded from sensitivity analyses.

  • BMI, body mass index; CVD, cardiovascular disease; DPP, Diabetes Prevention Programme; DPS, Diabetes Prevention Study; DPS, Diabetes Prevention Study; IDPP, Indian Diabetes Prevention Programme; RCT, randomised controlled trial; RR, relative risk; VIP, Vasterbotten Intervention Programme.