Physical activity follows a social gradient: the more disadvantaged a group, the less active it tends to be, and the fewer opportunities it has to change that. For a sport and exercise medicine (SEM) clinician this matters twice over, because unequal opportunities to be active contribute to avoidable differences in health and wellbeing, and because well-intentioned promotion can widen the gap it was meant to close. A campaign that works best for people who are already motivated, resourced and well served will increase inequality even while improving the population average. This page covers who is least active and why, the social and cultural factors that shape participation, and how to design promotion that narrows rather than widens the gradient.
Who is least active, and why?
Activity levels differ systematically by socioeconomic position, sex, ethnicity, disability, age and geography, and these characteristics interact rather than simply adding together, which is why an intersectional view matters. The detailed participation figures quoted in this area come from national surveillance in England, which continues to find persistent differences by deprivation, disability, age, ethnicity and sex. People in the most deprived areas are, on average, least likely to meet the guidelines and most likely to live where being active is hardest: less access to high-quality, safe and usable green or blue space, poorer air quality, higher perceived crime, fewer safe walking and cycling routes and fewer affordable facilities. Note that deprivation is more consistently associated with the quality and accessibility of green space than with the simple amount of it. Women are on average less active than men across most age groups, with participation falling sharply during adolescence, although not every individual follows that trajectory. Disabled people are on average markedly less active and face both physical inaccessibility and the assumption that activity is not for them. Participation also differs between and within minority ethnic populations, and these groups should not be treated as one: patterns vary by generation, sex and local area, and the drivers include cost, provision that does not accommodate cultural or religious requirements, and services that have never asked what would work.
Barriers operate at individual, social and structural levels, though many cross levels. Individually focused interventions may widen the gradient when uptake or benefit is greater among more advantaged groups.
It helps to sort barriers into three levels, while treating that classification as a heuristic rather than a fixed taxonomy, because many barriers cross levels: cost is both economic and structural, caring responsibilities are both interpersonal and structural, and safety is both social and environmental. Individual factors include health conditions, confidence, previous experience, skills and available time. Social factors include cost, caring responsibilities, cultural norms about who exercises and what is appropriate, and whether people feel safe and welcome. Structural factors include the built environment, transport, facility provision, lighting and air quality, and these are largely outside the individual's control. The clinical significance is the phenomenon of intervention-generated inequality: individually focused interventions such as advice, motivational messaging and apps may generate wider inequalities when uptake or benefit is greater among more advantaged groups. That is why an intervention can improve average population activity while widening the gap between the most and least advantaged.
How should promotion be designed to narrow the gap?
The organising principle is proportionate universalism: provide for everyone, but with intensity and resource scaled to need. A purely targeted programme risks stigma and misses the many people with moderate need, while a purely universal one predictably benefits the advantaged most. In practice that means universal messaging combined with additional, better-resourced provision in the areas and groups where activity is lowest. Co-production matters more here than anywhere else in this subcategory: services designed with the intended participants rather than for them are far more likely to address the barrier that actually binds, which is frequently not the one clinicians assume. Asking what would make this possible is a more productive question than explaining why activity is beneficial.
Proportionate universalism provides for the whole population while scaling additional resource to need, avoiding both the stigma of purely targeted programmes and the drift of purely universal ones towards the advantaged.
Several practical design choices follow. Remove cost where possible, since price is among the most consistently reported barriers. Locate provision where people already are, and consider transport explicitly. Address safety and lighting for anything delivered outdoors or in the evening. Where local co-production identifies them as barriers, options such as single-sex sessions, culturally appropriate facilities and flexible dress requirements can be offered, recognising these are locally identified preferences rather than assumed requirements of an entire community, and provide childcare or family-inclusive sessions where caring responsibilities are the constraint. Make accessibility genuine rather than nominal for disabled participants, including staff trained to adapt activity. Use trusted local messengers and community organisations rather than assuming clinical channels reach everyone, since many of the least active are also least engaged with health services. Structural advocacy sits alongside all of this: active travel infrastructure, protected green space and planning decisions influence population activity far more than any clinic ever will, and clinicians have a legitimate voice in those conversations. Finally, evaluate by group rather than in aggregate, measuring invitations and reach, uptake, retention, experience, outcomes and adverse effects across relevant groups and combinations of characteristics, because a service can only demonstrate that it has narrowed the gradient if it looks at distribution rather than averages.
Exam Tips
•Physical activity follows a social gradient, and unequal opportunities to be active contribute to avoidable differences in health and wellbeing; characteristics interact rather than simply adding together.
•Barriers operate at individual, social and structural levels, but the classification is a heuristic: cost, caring and safety all cross levels, and access to high-quality usable green space matters more than the amount of it.
•Intervention-generated inequality: individually focused interventions may widen inequalities when uptake or benefit is greater among more advantaged groups, raising the population average while widening the gap.
•Proportionate universalism means providing for everyone with intensity scaled to need, avoiding both the stigma of purely targeted programmes and the drift of purely universal ones.
•Practical measures include removing cost, locating provision where people already are, addressing safety and transport, and offering culturally appropriate or single-sex options where local co-production identifies them as barriers rather than assuming them.
•Evaluate invitations and reach, uptake, retention, experience, outcomes and adverse effects across groups and combinations of characteristics, because narrowing the gradient cannot be demonstrated from aggregate figures.