The study, published in SSM – Population Health, drew on the UK Household Longitudinal Study, a nationally representative panel that has followed roughly 40,000 households since 2009. Researchers led by Enrico Pfeifer of University College London analysed data from tens of thousands of participants aged 16 and over, tracking four key behaviours: physical inactivity, fruit and vegetable intake, problematic alcohol consumption, and smoking. Unlike most previous work, which focused on older caregivers in cross-sectional snapshots, the new analysis captured people across the entire adult lifecourse, from young carers in their teens and twenties to retirees, and distinguished between entering caregiving and leaving it behind.

Methodologically, the team deployed two complementary statistical strategies. The primary approach used fixed-effects models, which exploit only within-person change over time and thereby control for all stable characteristics, observed or not, such as genetics, early-life experiences and personality. As a robustness check, the researchers fitted piecewise growth curve models on a propensity score matched sample, comparing each new caregiver with up to three matched non-caregivers and modelling behavioural trajectories before, during and after the transition. Sensitivity analyses using continuous measures, cigarette counts and multiple imputation broadly reproduced the main findings.

The headline result is a striking behavioural double edge. Entering caregiving was associated with a 16 percent reduction in the odds of being physically inactive, an effect that held for both low-intensity caregivers providing fewer than 20 hours of care per week and high-intensity caregivers providing 20 hours or more. The association was strongest among adults aged 65 and over and those in early mid-adulthood. Yet the same transition was linked to a 16 percent increase in the odds of being a smoker, an effect that proved remarkably consistent across sexes and age groups and, according to the trajectory models, was sustained rather than transient.

The smoking finding carries particular weight because it was most pronounced among younger caregivers, those aged 16 to 29 and 30 to 49, precisely the groups that most caregiving research has overlooked. The authors interpret this through the lens of stress and coping theory and the COM-B behaviour change model: role strain may erode caregivers’ capability and opportunity to cope in healthier ways while boosting the motivation to seek immediate relief through cigarettes. Lifecourse theory adds a second layer, since caregiving in early adulthood collides with education, career building and family formation, amplifying strain in ways that caregiving at later, more socially normative stages does not.

Not every behaviour moved. Fruit and vegetable consumption showed no significant change following either entry into or exit from caregiving, and the evidence on problematic drinking remained inconclusive. Fixed-effects models hinted at a marginal increase in problematic drinking after caregiving onset, driven by low-intensity caregivers, but the trajectory models found no significant overall association. Intriguingly, high-intensity caregivers consistently displayed lower levels of problematic drinking than matched controls before, during and after the transition, a pattern the researchers attribute not to caregiving itself but to selection, consistent with the so-called healthy carer hypothesis, in which relatively healthier individuals become intensive caregivers in the first place.

Leaving caregiving told a different story. Exiting the role was associated with a 9 percent increase in the odds of physical inactivity, strongest among adults aged 65 and over, and the trajectory models showed exited caregivers diverging steadily from those who continued to provide care, particularly among women. Crucially, no compensating improvements appeared in smoking, drinking or diet after caregiving ended. The authors caution that the apparent activity boost on entering care may partly reflect physically demanding caregiving tasks, such as manual handling and mobility assistance, rather than health-promoting leisure exercise, since the questionnaire used cannot distinguish the two, and occupational physical activity is known to carry adverse cardiovascular effects.

The public health implications are concrete. The researchers argue that caregiving transitions represent identifiable windows of behavioural change, and that smoking cessation support could be integrated into existing caregiver support programmes, with particular attention to younger caregivers vulnerable to stress-related smoking. Because smoking behaviours did not improve after caregiving ceased, such support should extend beyond the caregiving period itself. At the same time, the modest effect sizes suggest that individual-level interventions alone will not suffice; structural supports such as respite services, flexible working arrangements and accessible formal care may be needed to create conditions in which caregivers can sustain healthy behaviours.

The study’s strengths lie in its scale, its nationally representative sample, its separation of caregiving entry from exit, and the rare combination of fixed-effects and matched trajectory modelling, approaches seldom applied together in this field. Its limitations are those inherent to observational self-report data: possible recall and social desirability bias, a two-year gap between most outcome assessments, the inability to fully rule out differential pandemic effects in later waves, and the ever-present possibility of residual confounding. The authors are careful to note that causality cannot be assumed.

Nevertheless, the message is clear and quietly urgent. Caregiving does not act on health through a single mechanism; different behaviours respond to different blends of role strain, role enrichment, stress coping and selection, and the consequences depend on when in life the caring begins. With demand for unpaid care rising as populations age, the habits formed in the months after a loved one falls ill may echo through carers’ own health for decades. Supporting the supporters, the study suggests, is not only a matter of compassion but of preventive medicine.

Subject of Research: Longitudinal changes in health behaviours among UK adults transitioning into and out of unpaid caregiving

Article Title: Health Behaviour Changes Around Transitions Into and Out of Unpaid Caregiving in the UK: A Longitudinal Study

Article References: Pfeifer, E., Lacey, R., Xue, B., Pikhart, H., & McMunn, A. (2026). Health Behaviour Changes Around Transitions Into and Out of Unpaid Caregiving in the UK: A Longitudinal Study. SSM – Population Health, Article 101975. https://doi.org/10.1016/j.ssmph.2026.101975

Image Credits: AI Generated

DOI: 10.1016/j.ssmph.2026.101975

Keywords: unpaid caregiving, health behaviours, smoking, physical activity, UK Household Longitudinal Study, lifecourse theory, caregiving intensity, alcohol consumption, diet, public health, young carers, longitudinal study