Effect of pedometer-based walking interventions on long-term health outcomes: prospective 4-year follow-up of two randomised controlled trials using routine primary care data
Problem
Data are lacking from physical activity (PA) trials with long-term follow-up of both objectively measured PA levels and robust health outcomes. Two primary care 12-week pedometer-based walking interventions in adults and older adults (PACE-UP and PACE-Lift) found sustained objectively measured PA increases at 3 and 4 years, respectively. Using routine primary care data from trial participants, we aimed to evaluate intervention effects on long-term health outcomes relevant to walking interventions.
Approach
We downloaded primary care data for trial participants who gave written informed consent, for 4-year periods after their randomisation from the 7 PACE-UP and 3 PACE-Lift English general practices (PACE-UP from Oct 23, 2012, to Nov 11, 2017; PACE-Lift from Oct 12, 2011, to Oct 11, 2016). The following new events were counted masked to intervention status for all participants, including those with pre-existing diseases (apart from diabetes, where existing cases were excluded): cardiovascular (myocardial infarction, coronary artery bypass graft, angioplasty, and stroke or transient ischaemic attack, cardiovascular deaths), diabetes cases, depression episodes, fractures, and falls. Intervention effects on time to first event post-randomisation were modelled using Cox regression for all outcomes, except for falls which used Poisson regression to allow for multiple events, adjusting for age, sex, study. Absolute risk reductions (ARRs) and numbers-needed-to-treat (NNT) were estimated.
Findings
1297/1321 (98%) of original trial participants 45-75 years, consented to primary care data linkage and had data downloaded. Events were <20 per group, except fractures and falls. Cox Hazard ratios (95% CI) for interventions vs controls were: non-fatal cardiovascular 0.24 (0.07 to 0.77); total cardiovascular 0.34 (0.12 to 0.91); diabetes 0.75 (0.42 to 1.36); depression 0.98 (0.46 to 2.07); fractures 0.56 (0.35 to 0.90). Poisson incident rate ratio (95% CI) for falls was 1.09 (0.83 to 1.43). ARRs and NNT (95% CI) for cardiovascular events were: non-fatal 1.7% (0.5% to 2.1%), NNT=59 (48 to 194); total 1.6% (0.2% to 2.2%), NNT=61 (46 to 472); and for fractures 3.6% (0.8% to 5.4%), NNT 28 (19 to 125).
Consequences
New cardiovascular events and fractures were significantly decreased at 4 years. Short-term primary care pedometer-based walking interventions can produce long-term health benefits and should be more widely used to help address the public health inactivity challenge. Our study also demonstrates the potential for using routine data to evaluate the outcome of large-scale primary care walking interventions, avoiding expensive objective accelerometry assessment or inaccurate self-report PA data.