Effectiveness of Adult Health Promotion Interventions Delivered Through Professional Sport: Systematic Review and Meta-Analysis

Studies were eligible for inclusion if they met the following criteria:

Studies reporting on health promotion initiatives delivered in professional sport settings including adult participants were identified through electronic database searches conducted in CINAHL, MEDLINE, SPORTDiscus, Scopus, the Cochrane Central Register of Controlled Trials (CENTRAL), and Google Scholar. The database searches were conducted in August 2019 and an updated search was conducted in July 2021 to ensure all relevant studies were identified and included in this review.

The full search strategy for one database (MEDLINE) is included in ESM Appendix S2. Filters were applied to only include human studies and those published in English, but no date limits were applied. When more than one article reported on the same intervention and the same sample, only the article reporting primary study outcomes was included. Articles reporting on the same intervention but including unique study samples (e.g., a pilot study and a fully powered RCT) were included as separate papers.

Two authors (EG and AR) independently screened each identified study based on the title and abstract (stage 1), followed by full-text screening (stage 2). In the event of a disagreement related to study inclusion or reason for exclusion, consensus was reached through discussion with a third author (AE). Reference lists of all included studies were reviewed for additional studies that potentially met the inclusion criteria.

Data extracted from eligible studies included authors, year of publication, sporting context, study location, study aims, intervention duration and data collection points, intervention details (including comparison/control group condition and frequency of intervention sessions), study incentives, target population (sex, age, and retention rates), outcome measures, and significant changes in study outcomes. Characteristics of the included studies are reported in Table 1. If a study reported results of per protocol and intention-to-treat (ITT) analyses, ITT results demonstrating a significant difference between groups over time were reported. A qualitative synthesis of intervention elements, characteristics of target populations, and study outcomes are reported.

Two authors (ESG and AE) independently assessed the risk of bias in the included studies using the revised Cochrane risk-of-bias tool for randomised trials (RoB 2) [23] (ESM Table S1). The RoB 2 assesses risk of bias across five key domains: the randomisation process, deviation from intended interventions, missing outcome data, outcome measurement, and selection of reported results [23]. For each domain, studies were assigned a rating of ‘low risk of bias’, ‘some concerns’, or ‘high risk of bias’. After all five domains had been assessed, an overall risk-of-bias judgement was assigned. Studies assigned ‘low risk’ are those for which low-risk judgements had been assigned for all domains. An overall judgement of ‘some concerns’ was assigned when there were concerns in at least one domain, while an overall judgement of ‘high risk’ was assigned when there was a high risk of bias in at least one domain or where ‘some concerns’ had been identified in multiple domains [23]. Using Cohen’s kappa [24], the interrater reliability for risk-of-bias ratings between the independent reviewers was 0.91, indicating a high level of agreement [25].

We examined a funnel plot to investigate risk of publication bias [26]. If smaller studies were found to have larger effects, there would be evidence that publication bias was present in the meta-analysis.

All included studies presented data on between-group differences in weight and waist circumference, therefore data on these outcomes were included in the meta-analysis. Authors of all the included studies were contacted to provide any missing data related to sample size, weight or waist circumference, and to verify data we extracted to be used in the meta-analysis. All but one author responded to our data request and we received missing data for four of the six included studies. For the remaining two studies, standard deviations for the change scores for weight and waist circumference were missing. To estimate these missing values, we used the following equation (Eq. 1), which is in line with guidance provided in the Cochrane Handbook [27].

The differences in change in weight and waist circumference between the intervention and control/comparison conditions from baseline to 12-week follow-up were the summary measures. We used Stata version 16 (StataCorp LLC, College Station, TX, USA) to conduct the meta-analyses using random-effects models. As all studies included in the meta-analyses used a common metric for weight (i.e., kilograms) and waist circumference (i.e., centimetres), non-standardised weighted mean differences were calculated. Variation attributable to heterogeneity was assessed using the I² statistic. If heterogeneity (I² > 50%) was found, meta-regression was used to test the impact of potential effect modifiers (i.e., mean baseline age, mean baseline weight/waist circumference). Risk-of-bias score was not investigated as a potential effect modifier due to limited variability.

This is the first systematic review to explore the effectiveness of health promotion initiatives delivered in a professional sporting context. Meta-analyses revealed small but significant changes in weight and waist circumference in favour of the experimental conditions at 12 weeks. Improvements were also observed for outcomes including physical activity [8, 10, 14, 17, 19], sedentary time [14], dietary intake [8, 10, 14, 17, 19] and a range of psychological and social health outcomes [8, 10, 14, 17, 19].

Sport has long been recognised as an institution in which masculinity is constructed and reinforced [30, 31], and sporting environments such as football clubs have been identified as both opportune spaces to engage men in “constructive reflection on their health and wellbeing” and spaces in which “damaging constructions of gender and inequalities” may be reproduced [32]. In recent years, and as demonstrated through this systematic review, sport has been established as a strong entry point through which to engage men for health promotion [8‐10]. This is promising given that, compared with women, men tend to be less engaged with health services and in health promotion and weight loss initiatives [33‐35]. All studies that met the inclusion criteria for this systematic review were designed specifically to engage men, with findings suggesting that the professional sporting context is a suitable and accessible setting through which men’s physical and mental health can be promoted and supported. In line with the original aim and design of the FFIT programme, authors of all studies included in this review discussed the ‘gender-sensitised’ nature of their interventions [8, 10, 14, 17‐19], which are designed to appeal to masculine ideals, preferences and needs. This gender-sensitised approach seeks to work with, rather than against, traditional constructions of masculinity, and key components of gender sensitisation in these studies include delivery of the intervention within the environment of a sporting club, provision of club-inspired merchandise (e.g., team training shirts), male-only groups, simple health messaging, and the use of humour or ‘banter’ to encourage camaraderie and open discussion [28]. There are many complex sociocultural factors influencing men’s engagement with health behaviours [36, 37] that must be considered in the design of gender-tailored health programs. One of the guiding principles for the development of the original FFIT programme was to enhance men’s “physical and symbolic proximity to the club and fellow male supporters” in an effort to counter potential perceived threats to masculinity [38] and masculine capital [37], which may be associated with involvement in a weight management programme. The successful adaptation and implementation of the FFIT programme across multiple demographic and sporting contexts demonstrates the programme’s success in supporting men to achieve lifestyle change [28].

The one programme that was identified through our database searches that exclusively targeted women used a pre-post-test study design to examine the feasibility of an FFIT adaptation for women [39]. Although this study did not meet the inclusion criteria for this review, as it was not conducted as a RCT, this feasibility study found that FFIT for women was feasible and acceptable and resulted in a mean 2.87 kg weight loss post-intervention. Findings from the mixed-methods feasibility study indicated that the programme appealed to women and was viewed favourably in comparison with other weight management programs and commercial dietary programs, which were often quite restrictive in their approaches. Women appreciated the group nature of the programme and the salience of the physical activity content, suggesting minor changes to enhance the programme for future delivery [39]. These findings suggest that recruiting women through a professional sport setting for the promotion of weight loss and healthy lifestyles may be efficacious and warrants further investigation.

In terms of weight loss, the findings from the current systematic review align with those of several other studies. In their review of male-only weight loss and weight maintenance interventions, Young et al. [40] found a significant difference in weight change favouring interventions in comparison with no-treatment controls. Characteristics of successful interventions included face-to-face intervention delivery in group settings, higher frequency of contact (i.e., ≥ 2.7 contacts per month), and inclusion of a prescribed energy restriction. Similarly, all studies included in the current systematic review were delivered in group settings and included face-to-face contact on a weekly basis. Borek et al. [41] also examined the effectiveness of group-based weight loss interventions promoting physical activity and nutrition. The mean difference in weight loss between intervention and control groups was 3.5 kg at 6 months, 3.4 kg at 12 months, and 2.6 kg at 24 months post-baseline. Findings from moderator analyses indicated that interventions that exclusively targeted men, those that explicitly targeted weight loss, and those that incorporated feedback mechanisms were more effective than those without these characteristics [41]. Importantly, most studies included in these reviews were of low study quality, prompting a call for rigorously designed, high-quality studies promoting weight loss and maintenance [40, 41].

All included studies in this systematic review incorporated practical physical activity sessions as a technique to not only increase physical activity participation but to also build a sense of camaraderie and belonging amongst participants. Although the methods used to assess physical activity varied across studies, the majority of studies demonstrated a significant increase in physical activity at the end of intervention in comparison with a waitlist control [8, 10, 14, 17, 19]. In a recent meta-analysis exploring the effectiveness of behaviour change interventions on men’s physical activity, Sharp et al. [42] found a significant intervention effect of 0.35 (Cohen’s d), which is estimated to be equivalent to an increase of approximately 97 min of total physical activity per week, or 980 steps per day. Interventions included in the review by Sharp et al. [42] that were gender-sensitised, underpinned by a theoretical framework, delivered across a period of 12 weeks or less, and involved at least one contact session per week were associated with greater increases in physical activity in comparison with interventions without these characteristics.

Considering the strong psychological connections sport fans form with their team and the strong sense of camaraderie formed between supporters of the same team [5], it is somewhat surprising that none of the included studies directly promoted mental health through their interventions. By nature, all included interventions evoked the sense of belonging and camaraderie offered through sport participation and fandom, and significant between-group differences were observed for a range of psychosocial outcomes [8, 10, 14, 19].

One study that included a specific focus on social connectedness, but did not meet the study design inclusion criteria for the current systematic review, was the HAT TRICK intervention [43]. This study used a pre-post-test study design without a control or comparison group and was delivered through a major junior ice hockey team in Canada. Significant treatment effects were reported for weekly minutes of moderate physical activity (objectively measured), self-reported moderate and vigorous physical activity, and fat scores, but there was no significant change in social connectedness [43]. In a subsequent study exploring the impact of the HAT TRICK programme on men’s mental health, significant positive changes in depression (assessed using the Male Depression Risk Scale) and mental health (assessed using the MH12) [44] were observed.

Study samples were relatively homogenous, with limited ethnic diversity and exclusively male samples. Five of the included studies [8, 10, 14, 17, 19] had a predominantly White sample (range 89.6–99.0%), with only one study reporting a sample comprising 37.75% non-White participants [18]. While the exact reason for the limited ethnic diversity in these samples is unclear, employing strategies to engage men from culturally diverse backgrounds will be key for future interventions designed to engage communities via professional sport. Socioeconomic diversity was also limited in some studies. While Hunt et al. [10] and Gray et al. [8] reported success in engaging men from across the socioeconomic spectrum, Kwasnicka et al. [19] and Petrella et al. [17] both noted limited socioeconomic representation and a need for further research on engaging men from diverse socioeconomic and cultural backgrounds.

All studies that were included in this systematic review were offered at no financial cost to participants. Three studies included the results of a cost-effectiveness analysis of the intervention [10, 14, 19]. One was found to be cost effective [10], one potentially cost effective [19], and one not cost effective in the short-term [14], possibly due to a ceiling effect for the quality-of-life measure on which cost effectiveness was based. The FFIT programme has been successfully scaled up and the original programme has since transitioned to a ‘single-licence franchise model’ that ensures programme fidelity, protects against commercialisation, and supports public health [28]. However, not all adaptations of the programme have been successfully scaled up and translated with greater reach. Considering the need for innovative and evidence-based approaches to engage men in health promotion and weight loss initiatives, it is imperative that these programmes are developed with a view to scalability. One potential approach to support long-term programme implementation, in the absence of ongoing funding, is to employ a paid registration model as is used in the MAN v FAT football programme in the UK and Australia [45]. This programme uses competitive sport, rather than the professional sport context, to support weight loss, and findings from a qualitative evaluation of men’s experiences in the programme reinforce the importance of camaraderie and social connectedness [46]. However, it is worth noting that charging a fee to participate in a health promotion programme may limit participation from some population groups [47], thereby amplifying existing socioeconomic inequalities.

The strengths of this systematic review include the rigorous study design, focus on the professional sport environment, and use of meta-analysis to assess the effectiveness of included RCTs on weight and waist circumference. These anthropometric outcomes are directly associated with chronic disease risk and are of public health importance [48, 49]. However, the findings of this systematic review should also be viewed in light of potential limitations. First, we only included studies that were published in English and did not search grey literature for studies that may not be published in scientific journals. Second, the results of the meta-analyses indicated high levels of heterogeneity (τ² = 2.57, I² = 94.2% for weight; and τ² = 1.13, I² = 76.8% for waist circumference), and none of the hypothesised predictors explained this variation. Third, this review only included RCTs as they are the highest level in the hierarchy of evidence for intervention studies. Although studies using other designs may also provide important evidence on the impact of health promotion initiatives delivered in a professional sporting context, they were not included in this review. Finally, of the studies that met our inclusion criteria, none were conducted in low- or middle-income countries, and ethnic and socioeconomic diversity was limited.