News Feeds

HealthFolk's News section is a health news aggregator with a focus on Nigeria, and Africa as a whole.

The effect of a programme to improve men’s sedentary time and physical activity: The European Fans in Training (EuroFIT) randomised controlled trial

Tags: None


Methods and findings

A total of 1,113 men aged 30–65 with self-reported body mass index (BMI) ≥27 kg/m2 took part in a randomised controlled trial in 15 professional football clubs in England, the Netherlands, Norway, and Portugal. Recruitment was between September 19, 2015, and February 2, 2016. Participants consented to study procedures and provided usable activity monitor baseline data. They were randomised, stratified by club, to either the EuroFIT intervention or a 12-month waiting list comparison group. Follow-up measurement was post-programme and 12 months after baseline. EuroFIT is a 12-week, group-based programme delivered by coaches in football club stadia in 12 weekly 90-minute sessions. Weekly sessions aimed to improve physical activity, sedentary time, and diet and maintain changes long term. A pocket-worn device (SitFIT) allowed self-monitoring of sedentary time and daily steps, and a game-based app (MatchFIT) encouraged between-session social support. Primary outcome (objectively measured sedentary time and physical activity) measurements were obtained for 83% and 85% of intervention and comparison participants. Intention-to-treat analyses showed a baseline-adjusted mean difference in sedentary time at 12 months of −1.6 minutes/day (97.5% confidence interval [CI], −14.3–11.0; p = 0.77) and in step counts of 678 steps/day (97.5% CI, 309–1.048; p < 0.001) in favor of the intervention. There were significant improvements in diet, weight, well-being, self-esteem, vitality, and biomarkers of cardiometabolic health in favor of the intervention group, but not in quality of life. There was a 0.95 probability of EuroFIT being cost-effective compared with the comparison group if society is willing to pay £1.50 per extra step/day, a maximum probability of 0.61 if society is willing to pay £1,800 per minute less sedentary time/day, and 0.13 probability if society is willing to pay £30,000 per quality-adjusted life-year (QALY). It was not possible to blind participants to group allocation. Men attracted to the programme already had quite high levels of physical activity at baseline (8,372 steps/day), which may have limited room for improvement. Although participants came from across the socioeconomic spectrum, a majority were well educated and in paid work. There was an increase in recent injuries and in upper and lower joint pain scores post-programme. In addition, although the five-level EuroQoL questionnaire (EQ-5D-5L) is now the preferred measure for cost-effectiveness analyses across Europe, baseline scores were high (0.93), suggesting a ceiling effect for QALYs.

Author summary


Physical activity is important in preventing chronic diseases, including cardiovascular disease, type 2 diabetes, and several cancers [1,2]. Global recommendations from the World Health Organisation (WHO) advise at least 150 minutes per week in moderate-to-vigorous physical activity. Recent estimates show that nearly one third of adults worldwide do not meet these recommendations and around 9% of premature deaths worldwide in 2008 can be attributed to lack of physical activity [2]. Not meeting the WHO physical activity recommendations costs healthcare systems globally 53.8 billion international dollars (INT$), with an additional indirect cost of INT$13.7 billion [3].

Sedentary behaviour has recently been shown to be associated with all-cause and cardiovascular mortality, independently of physical activity [4]. Sedentary behaviour is defined as any waking behaviour in a sitting, reclining, or lying posture with energy expenditure ≤1.5 metabolic equivalent tasks (METs) [5]. A meta-analysis has shown that interventions focusing primarily on physical activity have little effect on sedentary behaviour [6], and a specific focus on sedentary behaviour is needed to achieve substantial improvements in sedentary behaviour. Combining such a specific focus on sedentary behaviour in a lifestyle intervention programme with a focus on both physical activity and diet is novel, and given the contribution of all three behaviours to the burden of the world’s leading noncommunicable diseases, such a programme could have a substantial public health impact.

Men are often underrepresented in behavioural lifestyle interventions and are considered a hard-to-reach and underserved group [7]. However, many men lead an unhealthy lifestyle and are at high risk for developing noncommunicable diseases. It has been suggested that of all facets of health promotion, physical activity might be the most likely behaviour to engage men with their health. A systematic review has identified gender-sensitised physical activity programmes as a key development in men’s health promotion, with the potential to engage hard-to-reach men. The review also reported that all four identified studies that involved men engaging in physical activity with other men through professional sports resulted in increased physical activity [8]. Gender-sensitised physical activity programmes for men may also provide useful strategies in promoting other areas of men’s health. Another systematic review concluded that weight loss and maintenance for men is best achieved with interventions increasing physical activity and improving diet while using behaviour change techniques [7].

Achieving sustainable health behaviour change is challenging, and at-risk population groups, including overweight and/or inactive men, are difficult to engage and underserved. The Scottish Football Fans in Training (FFIT) lifestyle programme was designed to attract overweight men and enable them to lose weight through improvements in physical activity and diet. FFIT was shown to be cost-effective in supporting clinically significant weight loss. It also significantly improved self-reported physical activity and diet at 12 months [9], and improvements were partially maintained 42 months after baseline [10]. The multi-country European Fans in Training (EuroFIT) programme shifted the focus from weight loss to improving physical activity and sedentary time [11]. Like FFIT, EuroFIT uses the allegiance many fans have to their football club to attract at-risk men to a group-based lifestyle change programme delivered in their clubs.

This paper describes the results from the randomised controlled trial that aimed to evaluate the effectiveness of the EuroFIT lifestyle programme. The primary aim of the trial is to determine whether EuroFIT can help men aged 30–65 years with a self-reported body mass index (BMI) ≥27 kg/m2 to increase objectively assessed physical activity and decrease objectively assessed sedentary time over a 12-month period. Secondary outcomes of the trial include cost-effectiveness, food intake, body weight, BMI, waist circumference, resting systolic and diastolic blood pressure, cardiometabolic blood biomarkers, well-being, self-esteem, vitality, and quality of life.

Materials and methods

Recruitment and participants

Football clubs were selected by contacting clubs known by the study team to be likely to be interested in taking part. We sought a minimum of three and a maximum of five in each country, and the first 15 clubs that signed up were included. Clubs were Arsenal, Everton, Newcastle, Manchester City, and Stoke (England); Ado den Haag, Groningen, Philips Sport Vereniging (PSV), and Vitesse (the Netherlands); Rosenborg, Strømsgodset, and Vålerenga (Norway); and Benfica, Porto, and Sporting (Portugal).

Football clubs led recruitment of participants using emailed invitations to fans, the club website, social media posts, features in local press, and match-day recruitment.

Participants registered interest online, providing contact details, age, self-reported height and weight, and preferred football club. A follow-up telephone call included the adapted Physical Activity Readiness Questionnaire-Plus questionnaire (PAR-Q+) [12], previous participation in health promotion programmes at the club, and asking if men were willing to consent to randomisation and to wearing an activity monitor for 1 week at baseline and again at both follow-up assessments. On the consent form, men had the opportunity to opt into providing blood samples at the baseline and the 12-month follow-up measurements.

Men were eligible if they were aged 30–65, had a self-reported BMI of ≥27 kg/m2, and consented to study procedures. Men were excluded if they reported a contraindication to moderate intensity physical activity in the PARQ+ or participation in an existing health promotion programme at the club, or did not provide at least 4 days of usable activity monitor data at baseline.


EuroFIT was primarily designed to support men to become more physically active, reduce their sedentary time, and maintain these changes to at least 12 months after baseline. Dietary change was also introduced for those who wanted to lose weight. The programme was delivered at club stadia to groups of 15–20 men over 12 weekly, 90-minute sessions that combined interactive learning of behaviour change techniques with graded group-based physical activity. A reunion meeting was scheduled 6–9 months after the start of the programme. To facilitate group bonding and team spirit, the same group of 15–20 men were expected to attend at the same time each week.

Details of the EuroFIT programme are published, including a description of the programme in the template for intervention description and replication (TIDieR) [13]. In brief, we developed detailed manuals for coaches and participants, and trained club coaches over 2 days to deliver programme content in an appropriate and accessible style. This included encouraging positive banter, making sessions enjoyable, promoting a ‘team’ environment, and using interactional styles congruent with other (predominantly) male contexts [14]. The programme aimed to work with rather than against predominant constructions of masculinity [9,14] whilst supporting lifestyle change. Some elements (e.g., tips to change diet or increase physical activity) were adapted to country-specific norms. Coaches were taught about the importance of warm-up activities to prevent injuries, and the programme included the Fédération Internationale de Football Association (FIFA) 11+ programme [15]. Coaches taught participants to choose from a ‘toolbox’ of behaviour change techniques (including setting and reviewing goals for behaviours and outcomes, action planning, self-monitoring, and information about health and emotional consequences of change) and to emphasise personally relevant benefits of behaviour change (e.g., being better able to fulfil valued activities and roles). These behaviour change techniques were offered as tools for men to use for however long they found them useful and to encourage men to develop internalised and self-relevant motivation for becoming more active, sitting less, and eating a healthier diet [16].

We developed a novel pocket-worn, validated device (SitFIT) [17] to allow self-monitoring of sedentary and nonsedentary time (time spent upright [18]), in addition to daily steps (S1 Appendix). In the first week of the programme, men were taught how to measure the time they spent upright and the number of steps they take each week as a baseline. In the second and subsequent weeks, they were encouraged to follow an incremental programme to set weekly goals to slowly increase the number of steps and time spent upright each week, and to use the SitFIT to monitor their progress to these goals. Evidence on the use of self-monitoring devices for physical activity after participation in the FFIT programme suggests that although some continued to find them useful in the long term, others do not, as walking and other physical activity was embedded in everyday life without self-monitoring being necessary [19].

EuroFIT also explicitly encouraged between-session and post-programme peer support for changing behaviour through interacting with each other using a social media platform most of them were familiar with (e.g., WhatsApp, Facebook Groups). They were not given specific instructions on the content of interaction; coaches could decide whether or not they participated in the interactions. Between-session group social support was also encouraged using game-based social interaction with the MatchFIT app (S1 Appendix). MatchFIT allowed participants to contribute their weekly steps to their group’s collective average step count and compare it with that of a virtual competitor team. Coaches encouraged the use of MatchFIT as a means for participants to support one another as they pursued increases in their step counts, but did not themselves participate. Programme materials are available through request at


A fieldwork team collected outcome data at baseline, post-programme, and after 12 months in club stadia. They scheduled separate measurement sessions for intervention and comparison groups post-programme to minimise contamination. For participants who consented to biomarker assessment, we took a venous blood sample at baseline and 12 months, after 6 hours fasting.

To maximise attendance and retention, we made appointments by telephone, confirmed by email or letter, and sent short message service (SMS) reminders. We scheduled additional measurements either in stadia or at home as needed, but almost all men attended the regular measurement sessions. We recorded sociodemographic characteristics (age, ethnicity, education, marital status, current employment status, and income) at baseline.

In thanks for their participation in the research, we offered a club store voucher for the equivalent of €25 at post-programme and €75 at the 12-month measurements.

Self-reported behavioural outcomes.

Self-reported physical activity was assessed using the International Physical Activity Questionnaire (Short Form) (IPAQ) [23], self-reported sedentary time using the Marshall questionnaire [24], frequency of physically active choices using the Activity Choice Index [25], self-reported diet using an adapted Dietary Instrument for Nutrition Education (DINE) [26], and alcohol intake using a 7-day recall questionnaire.

Self-reported health and psychosocial outcomes.

Participants rated their well-being using the Cantril ladder, self-esteem using the 10-item Rosenberg self-esteem questionnaire, vitality using the subjective vitality scale, and health-related quality of life using the five-level EuroQoL questionnaire (EQ-5D-5L). EQ-5D-5L utility weights were estimated using the English value set [28]. Quality-adjusted life-years (QALYs) were calculated by multiplying the utility weights with the amount of time a participant spent in a particular health state. Transitions between health states were linearly interpolated. Participants also reported joint pain and any long-standing illnesses, disabilities, or infirmities. Questionnaires are available in S3S5 Appendices.


Costs were measured from the societal perspective and included programme delivery, healthcare utilisation, medication use, and absenteeism from work. Unit costs (£, 2016) from the UK were used to value healthcare utilisation and absenteeism [29,30]. Programme delivery costs were calculated using costs reported by participating football clubs (i.e., preparation, coordination and administration, recruitment, programme delivery and staffing, and materials). Costs for non-UK football clubs and universities were converted to British pounds using purchasing power parities [31].

Statistical analysis

Continuous data are summarised as mean and SD, median and interquartile range (IQR), or mean and standard error (SE) for multiply-imputed data in the cost-effectiveness analyses. Categorical data are summarised as frequencies and percentages. Outcomes post-programme and at 12 months were analyzed using linear mixed-effects regression models, adjusted for randomised group and baseline value of the outcome measure as fixed effects, and football club and country as random effects. Model residual distributions were examined graphically, and data were transformed as necessary. All analyses were intention-to-treat.

Baseline data were summarised by randomised group and for those who did or did not provide outcome activPAL data at the post-programme and 12-month assessment points, to assess the representativeness of those who provided outcome data for analysis.

Sensitivity analyses were carried out for analyses of the two primary outcomes and for body weight: (a) multiple imputation of missing baseline data, (b) repeated measures analysis, using data from all three time points in the same model, and (c) analyses to account for waking wear time (the duration for which the activPAL device was worn whilst the participant was awake).

For repeated measures analyses, data from all three time points (baseline, post-programme, and 12 months) were included as outcomes; fixed effects were included for randomised group, time point, football club, and a randomised group-by-time interaction. A random participant effect was included, and a general (unstructured) covariance structure was allowed for model residuals across the three time points. Intervention effects at post-programme and 12 months were estimated using the interaction terms from these models.

Two methods were used to account for waking wear time. First, the primary analysis models were repeated using the mean number of steps per hour and the percentage of waking time spent sedentary as outcome variables. Second, the repeated measures analyses described above were repeated with waking wear time included as a fixed effect.

For the primary outcomes and weight at 12 months, intervention effect heterogeneity was assessed by extending the regression models to include group-by-moderator interaction terms. Moderating factors considered were age, marital status, years of education, employment status, income, club, country, baseline BMI, long-standing illness, and pain in upper and lower joints.

All p-values are two-sided. For the primary outcomes, p-values <0.025 are considered statistically significant. For all other analyses, no adjustment has been made for multiple comparisons, and p-values <0.05 are considered suggestive of true associations. The statistical analysis plan is provided in S1 Analysis Plan.


Participants were recruited between September 19, 2015, and February 2, 2016. Participant flow through the trial is shown in Fig 1. Main reasons for exclusion for men who showed interest in the trial were BMI <27 kg/m2 (42.4%), inability to reach men after they expressed interest, men not being approached because the study had reached the maximum number of participants at a club (39.3%). Participants spanned all sociodemographic groups, but a majority were ‘native’ to the study country (meaning each of the participant, their mother, and their father was born there), had at least 12 years of education, were in full-time work, and were married or living with a partner (Table 1). At baseline, participants’ mean daily step count was 8,372 steps/day, sedentary time was 625 minutes/day, and BMI was 33.2 kg/m2.


Table 1. Baseline sociodemographic characteristics of participants allocated to the EuroFIT programme immediately (Intervention) or after 12 months (Comparison).

Data are mean (SD), or N (%). N (missing) are reported where necessary.

Those who provided outcome data (i.e., those who returned activPAL monitors with at least 4 valid days of measurements) were, on average, approximately 2 years older than those who did not, and slightly more likely to be married (S1 Tables, Table A). There was no clear difference in income in those who provided outcome data, nor in ethnicity, education, employment, or prevalent long-standing illness. In terms of baseline measures of study outcomes (S1 Tables, Table B), those who provided outcome were generally more active and less obese at baseline, compared with those who did not provide outcome data. This is a common feature of lifestyle intervention studies, in which those with the poorest lifestyle are hardest to engage in research.

We observed deliveries of the fourth session in 14/15 clubs. In these, coaches delivered 221 of 252 (88%) key tasks. Coaches in each of the 15 clubs provided attendance records for 553 programme participants: of these, 473 men (85.6%) attended at least 6 of the 12 sessions; 296 (53.5%) attended 10 or more sessions; and 85 (15.3%) attended all 12 sessions. Intervention participants rated their overall experience of the EuroFIT programme positively, producing a median score of 9 on a 10-point scale (IQR 8, 10; 70 missing). Asked to report their use of the SitFIT and MatchFIT, 65.1% of intervention participants reported they used the SitFIT ‘a great deal’ (score 4 on a scale of 0–4) and 36.8% reported they used MatchFIT ‘a great deal’.

The intervention group had a higher mean daily step count at 12 months than the comparison group (estimated difference: 678 steps/day [97.5% confidence interval (CI), 309–1,048], p < 0.001). There was no evidence of a difference between groups in sedentary time (estimated difference: −1.6 minutes/day [97.5% CI, −14.3–11.0], p = 0.77) (Table 2). In post-programme measurement, larger between-group differences in step counts (estimated difference: 1,208 steps/day [95% CI, 869–1,546]) and sedentary time (estimated difference: −14.4 minutes/day [95% CI, −25.1 to −3.8]) were observed (Fig 2). Sensitivity analyses using multiple imputations, adjusting for activPAL wear time and repeated measures analysis, showed broadly similar results (S1 Tables, Tables C, D and E).


Table 2. Objectively assessed physical activity and sedentary time outcome measures for participants allocated to the EuroFIT programme immediately (intervention) or after 12 months (comparison).

Data are mean (SD). Intervention effects estimated are mean differences (95% CI), derived from mixed-effects regression models**.

Data summaries for participants who provided data at both baseline and post-programme, or baseline and 12 months, are provided in S1 Tables, Tables F and G.

There was no evidence that improvement in physical activity at 12 months varied by age, marital status, years of education, employment status, income, club, country, baseline BMI, long-standing illness, or pain in upper and lower joints. There was a significant interaction between the effect of the programme on sedentary time at 12 months and limiting long-standing illness (p = 0.034), so that those with limiting long-standing illness increased their sedentary time. There was no evidence of any other intervention effect differences between subgroups (S1 Fig, Figure A and B).

Mean body weight, BMI, waist circumference, and the proportion of participants with BMI ≥30 kg/m2 all improved significantly in favor of the intervention group (Table 3). The intervention effect on body weight varied by baseline BMI (interaction p < 0.001), with greater effects seen in those with larger BMI at baseline (S1 Fig, Figure C).


Table 3. Physical measures for participants allocated to the EuroFIT programme immediately (Intervention) or after 12 months (comparison).

Data are mean (SD) or N (%). Intervention effects estimated are mean differences or odds ratios (95% CI), derived from mixed-effects regression models**.

All self-reported behaviours, including diet, improved post-programme and at 12 months in favor of the intervention, except alcohol intake, which improved only at 12 months (Table 4). In contrast to objective measurements, self-reported sitting time at 12 months significantly decreased in the intervention group compared with comparison.


Table 4. Self-reported behavioural outcomes for participants allocated to the EuroFIT programme immediately (intervention) or after 12 months (comparison).

Data are mean (SD) or N (%). Intervention effects estimated are mean differences or odds ratios (95% CI), derived from mixed-effects regression models***.

The intervention also improved several cardiovascular risk biomarkers at 12 months. Systolic and diastolic blood pressure were both improved; fasting insulin and HOMAIR were reduced by 15%; and fasting triglycerides, and ALT and GGT concentrations were reduced by 7%–8% (Table 5).


Table 5. Metabolic biomarkers for participants allocated to the EuroFIT programme immediately (intervention) or after 12 months (comparison).

Data are mean (SD) or N (%). Intervention effects estimates are mean differences (with 95% CIs), derived from mixed-effects regression models, or geometric mean ratios (with 95% CIs) (95% CI estimates derived from mixed-effects regression models of log-transformed biomarkers).

The intervention significantly improved self-reported well-being, self-esteem, and vitality, but not quality of life, as measured by the EQ-5D-5L at 12 months (Table 6).


Table 6. Self-reported psychosocial outcomes for participants allocated to the EuroFIT programme immediately (intervention) or after 12 months (comparison).

Data are mean (SD). Intervention effects estimated are mean differences (95% CI), derived from mixed-effects regression models.

The intervention group reported more recent injuries and higher upper and lower joint pain scores post-programme, and a higher lower joint pain score at 12 months (Table 7).


Table 7. Self-reported injuries and joint pain for participants allocated to the EuroFIT programme immediately (intervention) or after 12 months (comparison).

Data are N (%). Intervention effects estimated are odds ratios (95% CI), derived from mixed-effects regression models.

Prices per cost item and unadjusted mean differences in costs between the two groups are presented in Table 8. Costs of the EuroFIT programme differed between countries, ranging from £189.5 to £267.5 per participant. There were no significant differences in any other cost categories between intervention and comparison groups except for visits to physiotherapists. There was no statistically significant difference in total societal costs.


Table 8. Multiple imputed, unadjusted costs used by participants allocated to the EuroFIT programme immediately (intervention) or after 12 months (comparison), and their unit costs (£, 2016) over 12-month follow-up.

The mean difference in QALYs between the intervention and comparison group was small and not statistically significant (Table 9). One QALY lost in the intervention group was associated with an incremental cost of £126,119 compared with the comparison group. The probability of EuroFIT being cost-effective compared with the comparison group was at most 0.13 for ceiling ratios up to 30,000 £/QALY (S2 Fig).

One additional step/day in the intervention group was associated with an incremental cost of £0.41 compared with the comparison group (equating to £410 per 1,000 extra steps/day). There was a 0.95 probability of EuroFIT being cost-effective compared with the comparison group at a

Read more at the original link...


Read On:

Are you a doctor? Click here and signup to join medical discussion forums, and receive medical job alerts! Free!

Medical Jobs



Most Recent

Doctor? Click here to signup to receive news & research in your specialty