STRENGTHS AND LIMITATIONS OF THIS STUDY
-
Efficacy data from the RUFIT-NZ trial, a rigorous, pragmatic, multi-centred randomised controlled trial, allowed for the generalisability of modelled results across Aotearoa New Zealand.
-
The reduction in downstream events of myocardial infarction (MI), stroke and type 2 diabetes mellitus (T2DM) associated with weight reduction was based on the Global Burden of Diseases study, a comprehensive study exploring global disease trends.
-
There was considerable uncertainty around the extrapolation of a modest body weight reduction to subsequent downstream clinical events beyond the trial period.
-
Benefits attributed to weight reduction beyond MI, stroke and T2DM were not explored.
Background
Overweight, defined as a body mass index (BMI) of ≥25 kg/m2, and obesity (BMI of ≥30 kg/m2) are highly prevalent across Aotearoa New Zealand.1 In 2020/2021, approximately 34% of adults aged ≥15 years were overweight, and 38% of adults were obese.1 The health and cost burden attributed to overweight and obesity is correspondingly high, given the established relationship between overweight or obesity and adverse health outcomes.2 In 2021, direct healthcare costs associated with obesity were estimated to be 2 billion New Zealand dollars (NZ$), or 8% of total healthcare expenditure.3 Furthermore, there is a considerable discrepancy in the prevalence of obesity across sexes, with the prevalence of overweight being greater in men compared with women in Aotearoa New Zealand.1 Clearly, the considerable disease and cost burden attributed to overweight and obesity in Aotearoa New Zealand warrants intervention.
The Rugby Fans In Training–New Zealand (RUFIT-NZ) randomised controlled trial (RCT) was conducted in Aotearoa New Zealand to determine the effectiveness of a healthy lifestyle programme on weight loss in men with overweight or obesity.4 The trial was inspired by similar lifestyle interventions conducted in Europe, including the Football Fans In Training (FFIT) and the European Fans In Training (EuroFIT), which were associated with significant reductions in the participants’ body weight.5–7 However, few studies have explored the long-term cost-effectiveness of lifestyle interventions for reducing body weight among males with overweight or obesity. A recently published systematic review of adult health promotion interventions found considerable heterogeneity around the cost-effectiveness of the EuroFIT and FFIT programmes.7 The RUFIT-NZ trial observed that the difference in weight reduction at 52 weeks between participants in the intervention and control arms was −2.77 kg (95% CI −4.92 to −0.61, p<0.05), and a within-trial cost-effectiveness analysis found that RUFIT-NZ was likely to be cost-effective.8 However, as the health consequences of the RUFIT-NZ programme occur beyond the short trial period (through potentially preventing future adverse events), we sought to explore the long-term cost-effectiveness of the RUFIT-NZ programme from the Aotearoa New Zealand healthcare system perspective.
Methods
RUFIT-NZ
The model was profiled on participants in the RUFIT-NZ trial, a pragmatic, two-arm multi-centre RCT that assessed the effectiveness of a healthy lifestyle intervention delivered through three professional rugby clubs across Aotearoa New Zealand.4 9 The RUFIT-NZ trial comprised 200 overweight (BMI of ≥25 kg/m2) males, aged between 30 and 65 years who were able to safely undertake physical activity.4 The primary outcome was body weight changes from baseline to 52 weeks post-randomisation. Secondary outcomes included body weight changes at 12 weeks, waist circumference, blood pressure, fitness lifestyle behaviours and health-related quality-of-life at 12 and 52 weeks. Baseline characteristics of the RUFIT-NZ trial, as well as changes in the participant’s body weight, are presented in online supplemental tables A1 and A2 of appendix A. Full details of the trial have been published elsewhere.8 9
Supplemental material
Economic model
A Markov cohort model was developed to estimate the long-term clinical and cost impacts attributed to the RUFIT-NZ intervention versus no intervention in males with overweight or obesity in Aotearoa New Zealand.10 This analysis was proposed in the RUFIT-NZ trial protocol.4 8 Key clinical events of myocardial infarction (MI), stroke and T2DM were selected due to the strong association between obesity and cardiovascular or endocrine disease and the considerable health and cost burden attributed to these events on the Aotearoa New Zealand healthcare system.2 3 11 As such, the model comprised the following five health states; ‘Alive’, ‘Alive with T2DM’, ‘Alive with MI’, ‘Alive with Stroke’ or ‘Die’ (figure 1).
Structure of the Markov state transition model and possible health states. MI, myocardial infarction; T2DM, type 2 diabetes mellitus.
All modelled individuals began the simulation in the health state ‘Alive’ to reflect the status of males with overweight or obesity with no disease who either participated in the RUFIT-NZ intervention or were in the control group. The model simulated the movement of individual participants across health states based on the onset of MI, stroke, T2DM or death. Participants could remain in their current health state or be at risk of developing a non-fatal event (MI, stroke or T2DM) or death. Participants who developed MI or stroke moved to the health state ‘Alive with MI’ or ‘Alive with Stroke’ and were subsequently at increased risk of death. Participants developing T2DM moved to the health state ‘Alive with T2DM’ and were subsequently at risk of developing MI, stroke or death. Recurrent events of MI or stroke were not modelled to be consistent with the aim of primary prevention.
An annual cycle length with half-cycle correction and a lifetime time horizon was used in the base-case economic model to capture long-term patient outcomes and costs. The health consequences of obesity and the benefits attributed to reducing body weight are likely to occur beyond the trial period of RUFIT-NZ. Therefore, a lifetime time horizon was adopted.12 The model estimated the number of events of MI, stroke or T2DM, quality-adjusted life years (QALYs), years of life lived and costs that would be incurred among a cohort of 10 000 participants over a lifetime time horizon. Decision analysis was applied to compare differences in health and cost outcomes incurred by RUFIT-NZ participants compared with participants in the control arm. The primary outcome of interest was the incremental cost-effectiveness ratio (ICER), in terms of cost per QALY gained and cost per year of life lived for participants in the RUFIT-NZ intervention arm versus participants in the control arm. Although no official willingness-to-pay (WTP) threshold has been established for NZ, contemporary economic analyses have considered a Gross Domestic Production per capita expenditure of NZ$ 45 000 per QALY to be cost-effective.13 Costs were assessed from the perspective of the Aotearoa New Zealand public healthcare system, and all costs were presented in 2021 NZ$.
RUFIT-NZ effectiveness
Key epidemiologic parameters used in the economic model are presented in table 1. Additionally, the incidence of MI, stroke and T2DM across 5-year age bands used to inform the economic model are presented in online supplemental table B1 of appendix B.
Input parameters used in the economic model
The annual incidence of all-cause (background) mortality, MI, stroke and T2DM for the general male Aotearoa New Zealand population was drawn from published sources.14 Data from the latest Global Burden of Disease (GBD) study pertaining to the incidence of MI, stroke and T2DM for each 5-year age group among males aged ≥20 years were used to estimate the age-specific incidence of disease among the general population. The incidence of these outcomes was converted into annual transition probabilities for the general male Aotearoa New Zealand population10 and adjusted to reflect the greater risk of events among males with overweight or obesity, as estimated from previous systematic reviews that explored the association between obesity and T2DM/stroke.11 15 Similarly, the transition probability for MI among males with overweight or obesity was adjusted to reflect the greater hazard of MI among Aotearoa New Zealand European males with overweight or obesity.16 Life table data from the Aotearoa New Zealand Ministry of Health were used to inform age-specific transition probabilities for background mortality.
Subsequently, data from the GBD study were used to explore the impact of the RUFIT-NZ intervention on downstream events of MI, stroke and T2DM.17 First, the relative risk (RR) for events (MI, stroke or T2DM) associated with 5 kg/m2 unit increases in BMI were sourced from the GBD study for individuals aged 45–49 years.17 Second, the mean relative reduction in body weight (2.77 kg) of participants in the RUFIT-NZ intervention arm was transformed into the equivalent reduction in BMI (0.91 kg/m2) using the average height for the total Aotearoa New Zealand adult (aged ≥15 years) male population.1 This was used to estimate the RR for an event based on a decrease in BMI, assuming that the relationship between the risk of health outcomes and BMI was log-linear.18 19 The RRs were subsequently applied to the estimated likelihood of events in the control arm to determine the transition probabilities for MI, stroke and T2DM for participants in the intervention arm. Additional details pertaining to the estimation of the transition probabilities for the RUFIT-NZ intervention are presented in online supplemental table B2 of appendix B. The clinical benefits attributed to weight loss from participating in the RUFIT-NZ intervention were conservatively applied only to the first 2 years of the model in the base-case analysis; that is, following the first 2 years, the risk of MI/stroke/T2DM was equivalent to the control arm.
The incidence of non-fatal MI or stroke events and all-cause mortality following diagnosis of T2DM in Australia (stroke/MI) or Aotearoa New Zealand (all-cause mortality) were used to estimate the transition probability for events of non-fatal MI/stroke or all-cause mortality following the diagnosis of T2DM.20 21 Transition probabilities for fatal MI and stroke events were sourced from recently published studies exploring long-term survival for patients with MI or stroke across Australia and Aotearoa New Zealand.22 23 The likelihoods of mortality at 12 months for MI and stroke were 14.1% and 27%, respectively.22 23 Lastly, to reflect the greater background mortality risk following MI and stroke, the background mortality risk for the Aotearoa New Zealand male population with overweight or obesity was further adjusted using a standardised mortality ratio of 1.56 (MI) and 2.70 (stroke), respectively.24–26
The adjusted incidence of adverse outcomes was used to extrapolate the number of clinical events (MI, stroke or T2DM) occurring as a result of obesity beyond the trial follow-up period.
Cost inputs
Cost inputs used in the economic model are presented in table 2 below.
Key cost and utility inputs used in the economic model
Intervention costs
Participants in the intervention arm were assumed to incur a baseline cost of $1011 (for the first cycle of the model). This was based on micro-costing of the RUFIT-NZ intervention (see online supplemental table B3 of appendix B), which found that the total cost of delivering the intervention among 103 participants was $104 133. This was inclusive of costs associated with programme advertising, personal trainer and nutritionist sessions, participant travel and phone costs, as well as staffing costs and venue hire.4 8 27 As participants in the control arm did not receive the intervention, they were assumed to incur no cost at baseline.
Acute event costs
Costs associated with acute events of MI and stroke were based on Australian Refined-Diagnosis Related Groups (AR-DRGs) and their associated costs for publicly funded casemix hospitalisations in Aotearoa New Zealand in 2020/2021.28 The cost of admission for MI ($9,824) or stroke ($7462) was based on the weighted average of costs for MI-related or stroke-related DRGs, respectively (see online supplemental table B4 of appendix B). In lieu of a robust estimate of the cost of mortality, each death was assumed to be equivalent to the weighted average cost of fatal MI/stroke ($5120). Further, we assumed that only 50% of deaths due to any cause would occur in the hospital, and therefore, only 50% of all deaths would incur hospitalisation costs. No acute event costs were assumed for participants who transitioned to the ‘Alive with T2DM’ health state, as the diagnosis of T2DM does not typically require a hospitalisation.29
Annual treatment costs
The annual management costs associated with MI, stroke or T2DM were estimated using data from Blakely et al, which estimated the excess annual health spending for a variety of non-communicable diseases in Aotearoa New Zealand.30 Specifically, Blakely et al used national-level data tracking all publicly funded non-communicable disease events across uniquely identified individuals over time in Aotearoa New Zealand to estimate the disease costs.30 This includes costs associated with in-hospital and outpatient admissions, and pharmaceutical management following the incident or acute disease events.30 The annual costs of managing MI ($1885), stroke ($1543) and T2DM ($1914) were estimated based on patients aged 45–49 years to reflect participants in the RUFIT-NZ trial.30 Costs were converted from 2016 US dollars (USD) to 2021 USD using the Medical Care Consumer Price Index (MCPI), before adjustment to 2021 NZ$ through applying relevant purchasing power parities provided by the Organisation for Economic Co-operation and Development.31 32
Utility inputs
Key utility values considered in the model were drawn from published sources or participants in the RUFIT-NZ trial and are presented in table 2.
All participants entered the model with a utility of 0.90 (95% CI: 0.88 to 0.92) based on the average utility score for participants in the RUFIT-NZ trial. To capture the temporarily lowered quality of life associated with events, the mean utility value associated with MI (0.80) and stroke (0.70) were drawn from the Valsartan in Acute Myocardial Infarction and a recently published study using data from the Australian Stroke Clinical Registry, respectively.33 34 Participants who transitioned to the health state ‘Alive with T2DM’ were assigned a utility score of 0.82 based on the Action in Diabetes and Vascular Disease: Preterax and Diamicron MR Controlled Evaluation (ADVANCE) trial to reflect the poorer health-related quality of life associated with T2DM.35 Furthermore, decrements of 0.026 for MI and 0.099 for stroke were applied to participants experiencing acute events of MI or stroke following the diagnosis of T2DM.35
Discounting
An annual discounting rate of 3.0% was considered in the base-case economic analysis to discount costs, QALYs and years of life lived.36 This varied between 0% and 5% in sensitivity analyses.
Sensitivity analysis
A variety of deterministic and probabilistic sensitivity analyses (PSA) were performed to assess the uncertainty around key model parameters. In deterministic sensitivity analyses, key input parameters were varied one at a time using ranges reported in tables 1 and 2; this includes varying the time horizon, the assumed duration of efficacy of RUFIT-NZ, the discount rate and the extent to which RUFIT-NZ reduced downstream health events. A second-order Monte Carlo simulation with 2000 iterations was performed for the PSA using the ranges and probability distributions presented in tables 1 and 2.
All analyses were performed using TreeAge Pro 2021 (TreeAge Software, Inc., Williamstown, MA, USA).
Patient and public involvement
No patients or the public were involved in this study.
Results
The results of the base-case economic model are presented in table 3 below.
Results of the base-case cost-effectiveness analyses
The base-case model estimated that over a life-time time horizon, participants in RUFIT-NZ gained 0.02 (discounted) additional QALYs and incurred an additional cost of $863 (discounted), resulting in an ICER of $49 515 per QALY gained. Therefore, from a health economic perspective, RUFIT-NZ was not cost-effective compared with the control arm. In the estimation of first-ever events of MI, stroke or T2DM occurring within the first 5 years, the base-case model predicted that among 10 000 participants, RUFIT-NZ was associated with 14 fewer events of MI (two fewer fatal), one less stroke event (one less fatal) and 43 fewer events of T2DM relative to participants in the control arm (table 3). Over a lifetime, participants in the RUFIT-NZ arm experienced 12 additional events of MI (one additional fatal MI), one additional stroke (one less fatal stroke) and 25 fewer events of DBM.
The results of one-way, deterministic sensitivity analyses for key model drivers are presented in table 4. Additional results of one-way deterministic sensitivity analysis for other modelled inputs are presented in online supplemental table C1 appendix C.
Results of the deterministic sensitivity analysis on key parameters
Based on the results of one-way deterministic sensitivity analyses, the model was most sensitive to the modelled time horizon, the duration of efficacy of RUFIT-NZ, the discount rate, the cost of RUFIT-NZ, the likelihood of T2DM among obese patients and the reduction in the risk of T2DM for participants in RUFIT-NZ. The results of the PSA are presented in online supplemental figure C2 of appendix C. Based on the results of the PSA, RUFIT-NZ was cost-effective in 19% of 10 000 iterations if a WTP threshold of $45 000 per QALY was used.
Discussion
Our cost-effectiveness analysis demonstrated that participation in the RUFIT-NZ lifestyle intervention was unlikely to be cost-effective relative to no intervention. This was as the marginal benefit attributed to participating in RUFIT-NZ (0.02 QALYs gained per person) driven by a reduction in T2DM events was outweighed by the additional costs ($863 per person). To the best of our knowledge, this is the first cost-effectiveness analysis of a lifestyle intervention for reducing body weight from the perspective of the Aotearoa New Zealand healthcare system.
We assessed the within-trial cost-effectiveness analysis of RUFIT-NZ, which estimated an ICER of $40 269 per QALY gained in the base-case analyses.8 The cost-effectiveness of RUFIT-NZ reduced ($40 269 per QALY to $49 515 per QALY) on extrapolation of the body weight reduction between participants in the intervention and control arms to reductions in clinical events (MI, stroke or T2DM) likely to occur beyond the trial period. Notably, in the first 5 years, RUFIT-NZ was associated with 14 fewer events of MI (two fatal), one less stroke event (one less fatal) and 43 fewer events of T2DM relative to participants in the control arm. However, over a lifetime time horizon, participants in RUFIT-NZ experienced additional events of MI and stroke compared with participants in the control arm. This is attributed to the higher proportion of individuals in the health state ‘Alive’ in the RUFIT-NZ arm, who are subject to a greater likelihood of subsequent clinical events of first-ever MI/stroke/T2DM as the incidence of these clinical events increases with increasing age. As such, we report the difference in clinical events between RUFIT-NZ and the control arm accrued at 5 years as well as over a lifetime time horizon (base-case). This allows distinguishing between the benefits of RUFIT-NZ, which are unlikely to continue beyond 5 years, and background clinical events, which occur over a lifetime.
The reduction in the long-term cost-effectiveness of RUFIT-NZ contrasts with economic evaluations conducted for the EuroFIT and FFIT trials.6 7 37 38 Although the within-trial evaluation of EuroFIT suggested that the programme was not cost-effective in the short-term, extrapolation of changes in physical activity levels attributed to EuroFIT to reductions in health events found that the programme was dominant (ie, more effective and less expensive) compared with no intervention over a 10-year time horizon.6 38 Similarly, the cost-effectiveness of FFIT improved from £13 847 per QALY gained (within-trial analysis) to £1790–2200 per QALY gained (lifetime time horizon).5 37 However, the long-term cost-effectiveness analysis of the EuroFIT RCT was based on the assumption that the efficacy of the intervention was sustained over a time span of 5 years, with the model capturing additional benefits attributed to reductions in colorectal cancer and depression, as well as costs attributed to lost productivity for each condition.38 Relative to the long-term economic evaluation of EuroFIT, conservative assumptions were used for the RUFIT-NZ economic analysis on the basis that the programme was associated with a modest (2.77 kg) reduction in the mean body weight over a 52-week period. As such, the efficacy of RUFIT-NZ was assumed to be sustained over 2 years versus 5 years in the long-term evaluation of EuroFIT. Notably, RUFIT-NZ was cost-effective (ICER: $12 042 per QALY) after extending the duration of efficacy to 5 years in deterministic sensitivity analyses. Furthermore, although the long-term weight loss attributed to FFIT (2.90 kg at 3.5 years) was comparable to the weight loss in RUFIT-NZ, the per-person cost of the RUFIT-NZ programme was considerably higher than that used for the long-term cost-effectiveness analysis of FFIT ($1011 vs £164). Importantly, as the marginal costs associated with programme set-up and implementation are likely to decrease considerably due to economies of scale, expansion of the RUFIT-NZ programme across Aotearoa New Zealand will likely coincide with improved cost-effectiveness. This is supported in our sensitivity analyses, with RUFIT-NZ being cost-effective at a per-person programme cost of $859 (deterministic) and cost-effective in 19% of iterated ICERs (PSA). Notably, the benefits attributed to FFIT were maintained in a study exploring the expansion of the FFIT programme across sporting clubs in Scotland, with the Scottish Government funding the expansion of FFIT in light of ongoing evidence pertaining to the clinical and cost benefits of programme delivery.39–41 As such, the base-case ICER estimated in the present economic evaluation of RUFIT-NZ is likely conservative.
A recently published systematic review exploring economic evaluations of non-surgical weight management found considerable variation in the cost-effectiveness/utility attributed to weight management programmes, with ICERs ranging from US$335 952 per QALY to dominant (more effective and less costly versus the comparator).42 In terms of long-term cost-effectiveness/utility analyses, 12 out of 32 identified studies found that weight management programmes were, at minimum, cost-effective (range: US$32 078 to dominant). However, the comparability of results across the identified studies, with the long-term cost-effectiveness analysis of RUFIT-NZ, is limited. This is attributed to the considerable heterogeneity around the type of weight management programme, as well as the methodology adopted across each economic evaluation.42
A key strength of our study was the use of efficacy data from the RUFIT-NZ trial, a rigorous, pragmatic, multi-centred RCT, allowing for the generalisability of results across Aotearoa New Zealand. Furthermore, instead of data pertaining to outcomes of MI, stroke or T2DM from the trial, the long-term impact of reductions in body weight was drawn from the latest GBDs study, which is, to date, the most comprehensive study exploring disease trends at the global level.17 A variety of deterministic sensitivity analyses, and PSAs, supported the robustness of our results.
A key limitation of our study was the uncertainty around the benefit attributed to body weight loss for RUFIT-NZ participants. Although there was a significant mean difference in weight change between the RUFIT-NZ and control arms, participants in both arms had a mean baseline BMI of >35.0 kg/m2, and there was no statistically significant difference in BMI between the groups at 52 weeks. Hence, there was uncertainty around the extrapolation of a modest body weight reduction to subsequent downstream clinical impacts beyond the trial period. As such, we conservatively assumed the clinical benefits of RUFIT-NZ would only be sustained in the first 2 years of the time horizon; thereafter, the risk of events was assumed to be the same for participants between the intervention and control arms. This was consistent with a long-term follow-up study of FFIT, for which participants in the intervention arm had an average weight regain of 2.59 kg (95% CI: 1.61 to 3.58) (p<0.001) over 3.5 years.5 As discussed above, the benefit attributed to the EuroFIT programme was assumed to remain stable over 5 years, despite the trial being conducted over a 12-month period.38 Ultimately, the period over which RUFIT-NZ assumed to be effective (2 years) for the economic evaluation is conservative. Second, other cost-effectiveness analyses of weight-loss interventions explored the impact of the exercise intervention on additional health conditions, including colorectal cancer and depression.38 43 As obesity is associated with a variety of health conditions, limiting our analyses to exploring the impact of body weight reduction on MI, stroke or T2DM likely underestimates the true cost-effectiveness of RUFIT-NZ. However, our model was limited to conditions strongly correlated with body weight (MI/stroke/T2DM) and, therefore, most likely affected by a modest reduction in body weight. Furthermore, our analyses were limited to exploring the impact weight reduction on downstream clinical events. The impact of changes in secondary outcomes on subsequent clinical events, such as improved cardiorespiratory fitness and physical activity levels, were not included in the model. Notably, participants in the RUFIT-NZ programme demonstrated significant improvements in cardiorespiratory fitness, strength (push-ups, sit to stand) and self-reported physical activity score as measured by the Godin Leisure Time Questionnaire.4 8 Studies have demonstrated that cardiorespiratory fitness can mitigate obesity-related cardiovascular disease (CVD risk), with higher levels of cardiorespiratory fitness among individuals with obesity associated with improved outcomes in CVD incidence and mortality.44–47 However, it was not possible to distinguish between the impact of body weight reduction (the primary outcome) and these key secondary outcomes (fitness and physical activity) on downstream events of MI/stroke/T2DM. As such, limiting the economic analysis to exploring the impact of weight reduction likely underestimates the true cost-effectiveness of RUFIT-NZ.
Lastly, it was not possible to discern the impact of RUFIT-NZ among Māori and Pacifika men in a subgroup analysis for the pivotal RUFIT-NZ trial. There is a considerable discrepancy in the prevalence of obesity across ethnic groups, with the prevalence of obesity being 1.7 times higher for Māori (the indigenous people of Aotearora New Zealand) males and 2.3 times higher for Pasifika males relative to non-Māori or non-Pasifika males in Aotearoa New Zealand.1 As such, the cost-effectiveness of a similar lifestyle programme to RUFIT-NZ for Māori and Pacifika in Aotearoa New Zealand is likely to yield different results.21 48
Conclusion
RUFIT-NZ was associated with a reduction in cardiovascular and endocrine events for overweight and obese males. However, based on conservative assumptions, RUFIT-NZ was unlikely to be cost-effective from a healthcare system perspective.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Ethics statements
Patient consent for publication
Ethics approval
The RUFIT-NZ trial protocol was approved by the University of Auckland Human Participants Ethics Committee in accordance with the Declaration of Helsinki (021888) dated 20 September 2018. Approval is granted until 20 September 2024.
Acknowledgments
The authors wish to acknowledge the contribution of Gerhard Sundborn, Sally Wyke and David Revalds Lubans, who contributed to the conceptualisation and design of the RUFIT-NZ RCT and Claire Arandjus, who was involved in the management and running of the RUFIT-NZ project according to the design through NIHI. The authors also wish to acknowledge the rugby clubs who support this study and contribute their time and resources to RUFIT-NZ. The authors also wish to acknowledge the Trainers, and Nutritionists who deliver the program and the Research Assistants who help with the testing sessions. Finally, the authors wish to acknowledge the men who signed up to this program. ‘Rugby Fans in Training New Zealand: A Randomised controlled trial’ builds on the Football Fans in Training (FFIT) program, the development and evaluation of which were undertaken by a research team led by the University of Glasgow with funding from various grants including a Medical Research Council (MRC) grant (reference number MC_UU_12017/3), a Chief Scientist Office (CSO) grant (reference number CZG/2/504) and a National Institute for Health Research grant (NIHR) (reference number 09/3010/06). The development and evaluation of FFIT were facilitated through partnership working with the Scottish Professional Football League Trust (SPFLT). We gratefully acknowledge some source material from the Nutrition and Dietetic Department, NHS Forth Valley and Men’s Health Clinic, Camelon, Falkirk. The programme development is described in Gray et al (2013), and the results of the programme evaluation are reported in Wyke et al (2015) and Hunt et al (2014). These publications (and others relating to the program) are available from https://ffit.org.uk/.
This post was originally published on https://bmjopen.bmj.com