Physical activity referral schemes : socio-demographic patterning of exposure, uptake and attendance

Background. Despite the prevalence of Physical Activity Referral Schemes (PARS) in Britain, to date, there has been no attempt to elucidate their public health role. The nature of this role will depend upon who schemes are accessible to, and appropriate for. Objective. The aim of the present research was to provide a rigorous account of sociodemographic bias in referral to, and attendance of a county-wide PARS by tracking participant progress, from point of referral to the end of their involvement with the scheme. Methods. Epidemiological methods were applied to analyse data from people referred to the Somerset-wide ProActive PARS over a three-year period. For all Somerset residents referred, age, gender, and the deprivation level (measured by the Townsend score) and urban-rural character of their area of residence were used to make socio-demographic comparisons with the county population as a whole. Mann-Whitney and Kruskal-Wallis difference tests were used to compare socio-economic characteristics. Binary logistic regression analysis was used to identify sociodemographic characteristics associated with referral uptake (attending ≥1 session) and completion of physical activity programmes (≥80% attendance). Results. The proportion of referred participants (n=3569) who were female was above the county average (61.1 vs. 51.4%). Referrals increased markedly with age from 6.6% (≤30 yrs) to 22.7% (50-59 yrs), dropping off sharply thereafter (≥60 yrs). The mean Townsend deprivation score for participant area of residence was greater than the county average (0.33 vs. 0.00, p<0.001). These patterns indicated that, with the exception of older adults, those groups most likely to consult primary care were referred most frequently. Regression analysis (n=2864) revealed that increasing age (Exp(B)=1.014, p<0.001) and urban residency (Exp(B)=1.317, p<0.001) increased the likelihood of referral uptake. Uptake was less likely for those living in more deprived areas (Exp(B)=0.933, p<0.001). In participants who took up referral, the likelihood of completion was lower in women than men (Exp(B)=0.818, p=0.041) but increased with age (Exp(B)=1.018, P<0.001). Conclusion. Younger people, residents of deprived and rural areas, and women appear less likely to progress through schemes or complete PARS programmes, regardless of relative scheme exposure. PARS appear more appropriate for adults of middle-to-old age who are more likely to require supervision and should be targeted accordingly. To promote physical activity in a preventive capacity and to increase activity across the socio-economic strata is likely to require broader and multi-faceted environmental and policy-led interventions to promote habitual activity, rather than attempting to increase recreational activity through individual-orientated such as PARS.