Discussion

Summary of the findings

The PACE-UP trial demonstrated that both the postal delivery and the nurse-supported pedometer interventions, based on trying to gradually add in ‘3000 steps in 30 minutes’ on most days, increased objectively assessed PA (step counts by around one-tenth and MVPA in bouts by around one-third) among predominantly inactive 45- to 75-year-olds at 12 months. Although the nurse-led delivery had a greater effect than the postal delivery at 3 months, by 12 months this difference was not sustained. The interventions had no effect on sedentary time, anthropometry or other outcomes and did not increase AEs. No effect modification was demonstrated (by age, sex, taking part as a couple, self-efficacy, disability, socioeconomic status, pain or BMI). Questionnaire-based outcome measures tended to support the conclusions of accelerometer measures, but only if walking was an explicit part of the questionnaire. Thus, the IPAQ MVPA question did not show any intervention effect, but the IPAQ walking question showed a significant effect of both the nurse-supported and the postal delivery interventions with no difference between the interventions, although with less precision than the accelerometry data.

Both interventions were well accepted and the trial had high fidelity; three-quarters of the nurse group attended all three sessions and around 80% of both the postal delivery and the nurse-support groups returned completed step count diaries. Increase in step count was positively associated with both nurse session attendance and completed diary return.

Incremental cost per step was £0.19 and £3.61 per minute in a ≥ 10-minute MVPA bout for the nurse-support group, whereas the postal delivery group took more steps and cost less than the control group. The postal delivery group had a 50% chance of being cost-effective at a £20,000 per QALY threshold within 1 year. The QALY-based conclusion changed to the control group dominating the postal delivery group when four alternative assumptions were made (using the 3-month outcome data, extending the perspective to participants, excluding health service use and using self-reports of AEs), although this was not the case for cost-effectiveness ratios based on step count and MVPA. The postal delivery group was significantly more cost-effective than both the nurse-support group and the control group in the long term and this finding was robust to changes in assumptions.

Nurses and intervention group participants described the intervention in a positive way and confirmed that primary care was an appropriate setting. Nurses believed that participating in the trial, especially in the BCT training, enhanced the quality and delivery of the advice and support they provided within routine consultations. Participants described important facilitators of increasing PA, including the desire for a healthy lifestyle, improved physical health, enjoyment of walking in the local environment, having a flexible routine, appropriate self-monitoring and external monitoring and support from others. Important barriers included physical health problems, having an inflexible routine, work and other commitments and poor weather. Several BCTs were highlighted as having an important impact, including self-monitoring and review of goals and outcomes, planning social support/change and relapse prevention. Although most participants in the postal delivery group were confident in increasing their PA without nurse support, two key caveats were existing health problems and overcoming barriers.

The follow-up of over two-thirds of the trial cohort at 3 years demonstrated persistent increases in both step count and time in MVPA for the nurse-support and postal delivery groups compared with the control group, with no difference between intervention groups. The postal intervention given to the control group at 12 months, with no follow-up telephone call after 1 week and no requirement to post back the diary to be reviewed after 3 months, was not effective at increasing participants’ PA levels. This suggests that these ‘minimal support’ components of the postal delivery intervention, which were not face to face or not provided by a health-care professional, may have been important to its success. Qualitative evaluation found that most participants felt that the PACE-UP trial had increased their awareness of the importance of PA, irrespective of their intervention group and whether or not they felt that their PA had actually increased. Many of the barriers to, and facilitators of, PA maintenance were the inverse of each other and most were similar to those found to be important for increasing PA during the actual trial (health conditions, weather, ageing, social support, time). Participants varied in the resources that they would find most beneficial to help them maintain their PA levels, emphasising the importance of individual tailoring of some aspects of PA interventions.

The PACE-UP trial was novel in clearly separating out the effects of pedometer provision and nurse support in a general population sample of adults and older adults and in demonstrating the effects on both step counts and MVPA in bouts, thus making the outcome assessment relevant to current national and international PA guidelines.

Strengths and limitations

Study strengths

The PACE-UP study had many important strengths. It was large and population based with primary care sampling, allowing response and any bias in response to be assessed, rather than relying on recruiting volunteers. It was designed to have household randomisation, which allowed two members of a couple to take part together if they wished to, enabling a comparison of individual and couple effects. It had three trial arms, allowing the separation of nurse support and pedometer/handbook/diary effects. The intervention was pragmatic, using practice nurses who worked in the practices to deliver the nurse PA consultations, rather than external researchers or exercise specialists. There was a very good uptake rate of nurse appointments and return of completed step count diaries, showing participant engagement with the interventions. The main PA outcomes were objectively measured and were relevant to the PA guidelines. AEs were measured in a number of ways to minimise bias, both self-reported from questionnaires and objectively from primary care records. The trial achieved a follow-up rate of over 90% with complete primary outcome data. The trial also included embedded process, qualitative and economic evaluations, with the economic evaluations using trial results in a simulation of long-term cost-effectiveness. An extended 3-year follow-up allowed maintenance of any intervention effects beyond 12 months to be studied.

Study limitations

There were also some important study limitations. The 10% recruitment into the trial is considered in detail below, in Generalisability. At the baseline assessment, 218 out of 1023 participants (21%) achieved the PA guideline targets based on their accelerometry. These participants were not excluded from the trial because if the intervention were to be rolled out in primary care, self-reported PA levels would define participation. Our nurse-supported intervention group had slightly higher baseline PA levels; however, the trial results were not biased, as analyses were based on individual change, controlling for baseline PA level. It was impossible to mask participants and nurses to the intervention group and, pragmatically, research assistants recruited and followed up the same participants, so were unmasked to group at the outcome assessment. However, all the primary and secondary PA outcomes were assessed objectively by accelerometry. It is possible that participants might have tried harder with their PA when monitored, but this would also have affected control participants and would be reduced by using a 7-day protocol for data collection.31 In addition, our intervention groups increased their MVPA in bouts of ≥ 10 minutes, implying that participants made changes suggested by the programme. Despite recruiting to target and having excellent follow-up, our CIs for the difference between intervention groups cannot rule out a small 12-month difference. Interpretation of our 3-year follow-up findings was potentially limited by the fact that two-thirds of the control group participants received a pedometer, handbook and diary, and 20% of them also received a single nurse appointment after their 12-month follow-up. However, any contamination appears to be minimal, as there was no evidence of a change in the control group and the intervention estimates at 3 years were very similar to those at 12 months. These findings are of potential importance as, in combination, they suggest that the minimal contact with the postal delivery group after participants were sent their pedometer packs (telephone contact after 1 week and encouragement to return their completed step count diaries at 3 months) was important in stimulating an effect. Timing may have been important too; as the control group was offered the intervention 12 months after the participants had initially expressed an interest in participating in the trial, they would have had to have worn the pedometer to measure their step count and set targets, whereas the trial postal delivery group had targets set based on wearing a blinded pedometer at baseline. These added factors may have also been important to the success of the trial postal delivery intervention group.

Generalisability

Overall, only 10% of people invited to participate in the trial ended up being recruited and randomised. This is similar to other primary care PA trials,33,192 but lower than the 30% that we achieved in our recent older adult PA trial.21 However, 10% of a population sample is still a useful percentage to participate in a public health intervention, and this trial shows the potential of primary care to contribute to PA public health goals, particularly within an urban context. As well as monitoring overall recruitment, using primary care as a sampling framework allowed us to look for any selection bias in recruitment to the trial. Primary care record comparisons showed that participation rates were significantly lower in men, in those aged < 55 years, in those who were living in the most socioeconomically deprived quintile and among Asian rather than white or black ethnic groups. Despite selecting practices from deprived, ethnically diverse areas, few participants were from lower socioeconomic and ethnic minority groups, limiting both the subgroup analysis power and the generalisability to more diverse populations. Failure to include socioeconomically deprived or ethnic minority groups, in which PA levels were lower, could also increase health inequalities. In a RCT of an intervention, it is not possible to separate out reluctance to participate in the intervention from reluctance to participate in the trial itself, with requirements for informed consent, randomisation and rigorous follow-up and evaluation. If the intervention were to be rolled out in routine primary care, uptake could be higher and less prone to selection bias. Handing out the intervention materials (pedometer, handbook and diary) in primary care consultations in which advice to increase low PA levels is already being offered may also increase the intervention’s reach (e.g. in relevant chronic disease consultations, or as part of NHS Health Checks, which cover a similar age group and aim to reduce cardiovascular risk59). The intervention could also be a valuable addition to diabetes mellitus prevention strategies, such as the NHS Diabetes Prevention Programme, whereby primary care is being used to identify patients at a high risk of developing diabetes mellitus, many of whom are inactive193 and with higher proportions from ethnic minority groups. Using the PACE-UP trial intervention in these ways would need further evaluation and monitoring, but this may have the potential to improve generalisability and either decrease, or at least not increase, inequalities.

Comparison with other studies

We believe that this is the largest population-based trial of a pedometer-based walking intervention with 12-month follow-up findings and the only pedometer trial with objective PA data on time in MVPA, which is relevant to PA guidelines at 3 years. The results are consistent with, and extend, our findings in 60- to 75-year-olds that were achieved in the smaller PACE-LIFT trial21 and also support the recent change in NICE guidance to promoting pedometers as part of packages including support to set realistic goals, monitoring and feedback.37 The intervention used in the PACE-LIFT trial also included pedometer feedback, use of a step count diary and practice nurse PA consultations based around BCTs. However, the PACE-LIFT trial intervention comprised four longer practice nurse consultations, which also included individual accelerometer feedback on PA intensity. The PACE-LIFT trial was a two-arm trial with only a single intervention arm, and was therefore unable to separate out PA monitor effects from those of the nurse-support group. Despite including a much less intense intervention, the PACE-UP trial has delivered similar levels of effect at both 3 and 12 months in PA outcomes and, furthermore, has shown what can be achieved via a postal route. It is also reassuring that our interventions did not increase sedentary time, given its potential harm,194,195 as compensation can sometimes occur. The absolute step count increase achieved in the PACE-UP trial was modest compared with that reported in systematic reviews.26,31,32 However, most trials with 12-month data have been based on small numbers and recruited either volunteers196 or high-risk groups,35 or reported only self-reported PA data;97 all of these factors are likely to lead to larger effect sizes. Although PA guidelines focus on time in MVPA in bouts, not on step counts, the systematic reviews presented no data on this important outcome.26,31,32 The PACE-UP trial results confirm the PACE-LIFT trial findings,21 with significant 12-month increases in MVPA in bouts. Based on the ‘3000 in 30’ formula, 33–35 extra minutes of MVPA per week in bouts corresponds to approximately 500 extra steps per day. Thus, approximately three-quarters of the extra steps achieved in the PACE-UP trial (650–700 per day) contributed to an increase in MVPA in bouts. We believe that our trial is the first to show that the ‘3000-in-30’ message43 can lead to an approximately one-third increase in weekly MVPA in bouts at 12 months, as was achieved across both intervention groups. The ‘3000 steps in 30 minutes’ formula neatly captures intensity43 and could become an important new public health goal, particularly as many people now have the ability to measure steps easily with their mobile phones. Based on a systematic review, which has quantified the strength of association between walking and the risk of developing CHD,197 the increase of 33 minutes per week in the postal delivery group in our study at 12 months, if sustained, would be expected to reduce the risk of CHD by approximately 4.5% (95% CI 3% to 6%; see Appendix 8 for details). A cohort study that has related pedometer-measured steps to mortality198 has similarly allowed us to estimate that a sustained increase of 642 steps per day would be expected to lead to a decrease in all-cause mortality of approximately 4% (95% CI 1% to 7%; see Appendix 8 for details). Recalculating these estimates for the effect estimates in the postal delivery group at 3 years makes little difference, with the resultant decreases being 4% (95% CI 3% to 5%) for CHD and 4% (95% CI 1% to 5%) for total mortality.

Most pedometer-based interventions have not separated out the effects of the pedometer itself from the effects of the additional support provided.21,24,31 The Healthy Steps trial97 showed that pedometers achieved an additional effect compared with a primary care green prescription, but the PA outcomes presented were based on self-report. The PACE-UP trial demonstrates that although the nurse intervention group had a significantly greater effect on both step counts and time in MVPA at 3 months, by 12 months both the nurse-supported and postal delivery interventions still had a significant effect, but with no evidence of a difference between them. This stronger effect during the period of contact with the nurse, which was not sustainable in the longer term, has also been shown in other interventions with health professionals.199 Both the nurse-support and postal delivery groups received a pedometer, diary and handbook as part of the PACE-UP trial package; therefore, it is not possible to know how much the individual components contributed. A systematic review suggested that step count diaries were common to successful pedometer interventions,31 and approximately 80% of both of our intervention groups returned completed step count diaries. In addition, our process evaluation showed that returning a completed diary was significantly associated with an increase in step counts for both of the intervention groups. Qualitative findings also confirmed that participants from both groups valued the handbook and diary, as well as the pedometer.162 Control group participants provided with the pedometer, diary and handbook by post at 12 months did not significantly increase their PA levels; however, they were not asked to return completed step count diaries after 3 months, which may have contributed to the lack of effect of the materials in this group.

We found no effect of the interventions on anthropometric measures, such as BMI or fat mass; this is consistent with other similar studies.21,196 Our interventions also did not affect anxiety or depression scores, which is consistent with other primary care pedometer-based interventions, suggesting either no effect or insensitivity of these measures to change, particularly when levels are in the normal range for most people.21,33 Although a few participants mentioned that they had negative effects from overdoing walking, most intervention participants talked about feeling fitter, sleeping better, improved mood, having more energy, less pain and keeping more active into older age.162 There is currently a lack of data comparing individual, couple or household participation in walking studies.21,30 Household sampling allowed us to investigate this in the PACE-UP trial, but, unfortunately, only 20% of the participants took part in the study as a couple; therefore, we had reduced power for our subgroup analysis, which showed no effect of taking part as a couple, similar to the findings in our PACE-LIFT trial.21

The self-efficacy differences that we demonstrated between both intervention groups and the control group at 3 months and between the nurse-support group and control group participants at 12 months are consistent with the positive relationship between changing self-efficacy and PA behaviour that others have reported.200 The BCTs most associated with self-efficacy and successful PA outcomes are goal- and action-planning, prompting self-monitoring and feedback and planning of social support/change.200 All of these BCTs were specifically recommended in recent guidance29 and were included in our study in written materials for both intervention groups and as a focus of nurse PA consultations.70 Our qualitative interviews found that more BCT comments were made by the nurse support group than by the postal delivery group, apart from around self-monitoring.162 Increased self-efficacy has also been shown to be important for long-term PA adherence;201 however, we found no difference in 3-year PA maintenance between intervention groups, despite the nurse-support group having a higher level of self-efficacy than the postal group at 12 months.

Walking is a safe intervention, which is indicated in many chronic diseases,1,8 although empirical data on the safety of walking interventions are limited.24 A large trial based on 40- to 74-year-old women, which encouraged a single 30-minute brisk walk 5 days weekly, reported increased falls and injuries.60 Our findings in the PACE-UP trial, showing no increase in AEs, builds on similar evidence from the PACE-LIFT trial,21 using both self-reported and objective primary care data, and highlights the potential importance of building up MVPA gradually, particularly in older adults, those who are inactive or those who have comorbidities.1,11 The suggestion of a protective effect of the interventions in the PACE-UP trial on both falls and cardiovascular events at 12 months is plausible, but not definitive, as it is based on only a small number of events.

We demonstrated a persistent intervention effect at 3 years, in terms of both step counts and time in MVPA in bouts. This adds to the limited evidence from the systematic review by Hobbs et al.,26 who found only two trials with objective PA measurement data beyond 12 months in this age group. One trial reported a significant intervention effect on step counts at 18 months, but suffered high attrition bias,202 and another trial found no effect at the 24-month follow-up on either step count or accelerometry-assessed vector magnitude.203 Recent NICE guidance29 and a Cochrane systematic review23 also called for PA interventions with a longer follow-up period and objective PA measures. Our qualitative evaluations at 3 years also add to the limited evidence base on factors that lead to successful and unsuccessful PA maintenance190 and suggest that many of the factors that were important barriers to, or facilitators of, increasing PA levels in the original trial (e.g. health conditions, weather, ageing, social support, time) are still important when considering maintenance. This provides support for the credibility of our work and suggests that barriers and facilitators may be similar for both PA adoption and maintenance.204 Our findings support others in suggesting that future interventions should focus on techniques to transform PA barriers into facilitators, for example by demonstrating the value of PA for many chronic health conditions, as well as safe ways in which individuals can increase their PA levels to change the presence of chronic health conditions from inhibiting to promoting PA as people age.157,205,206

Our results on cost-effectiveness provide new evidence in a research area that reflects a dearth of primary evidence.139,140 The evidence that the postal intervention has a 50% chance, and the nurse intervention a 5% chance, of being cost-effective within 12 months is new. Although lower than the 95% likelihood of green prescriptions being cost-effective in New Zealand at 12 months,97 it is still a reasonably high percentage for a behaviour change intervention to achieve at 12 months. The expectation that the postal intervention is most cost-effective over a lifetime is very strong and is comparable to other findings from models.141,142

Interpretation of the results

Primary care patients aged 45–75 years can achieve important increases in their PA levels using a 12-week pedometer-based walking intervention, including handbooks and PA diaries (available at www.journalslibrary.nihr.ac.uk/programmes/hta/103202/#/), delivered either by post with minimal support or through practice nurse PA consultations, with both methods achieving similar 12-month effects. An important part of the intervention was to try and gradually add in 3000 steps in 30 minutes most days weekly. The persistent effect at 3 years suggests a long-term beneficial effect. This is a safe intervention that is acceptable to patients and nurses. The postal delivery group was significantly more cost-effective than the nurse-support and control groups in the long term, thus providing a cost-effective way of delivering long-term quality-of-life benefits. The lack of an increase in PA levels at 3 years in the control group, which received a simple postal intervention without further contact after the 12-month follow-up, suggests that contacting participants after posting the intervention components to them and encouraging the return of PA diaries may be important factors for success for the postal intervention route, but this minimal support does not need to be face to face or provided by a health-care professional.

Conclusions

Implications for health care

• A primary care pedometer-based walking intervention, delivered by post with minimal support, could provide an effective and cost-effective approach to addressing the public health physical inactivity challenge.
• The ‘3000 steps in 30 minutes’ formula neatly captures intensity and could become a useful new public health goal, particularly as many people can measure steps easily with their mobile phones.
• The PACE-UP 12-week pedometer-based walking intervention could be considered for inclusion into the NHS Health Check programme, aimed at a similar age group (40- to 74-year-olds), and/or the NHS Diabetes Prevention Programme.

Recommendations for research

• The PACE-UP trial generalisability is limited by the 10% overall recruitment rate and a lower recruitment rate in Asian and socioeconomically deprived patients. Further research into different recruitment methods is needed, as is research assessing the recruitment rate achievable if this programme were offered outside a trial setting over a more prolonged time period.
• Although the overall postal intervention outcomes were as effective and more cost-effective than the nurse-supported intervention outcomes, further research is required to understand who would benefit most from the individual tailoring offered by a nurse-supported intervention.
• There has been a recent dramatic increase in the use of wearables to monitor personal PA levels, including through smartphones, wrist-worn devices, online monitoring and mobile apps. Further research into how the PACE-UP 12-week PA programme could be integrated into the use of these devices (with/without a pedometer) is needed.