Background Aspects of community severance (the separation of people from goods, services, and each other by busy roads or other transport infrastructure) have been linked to poor health and wellbeing, but few studies have examined the overall relationship between community severance and health. We created a novel index for community severance and estimated its association with the self-rated health of adults in Great Britain.

Methods Data were collected from a nationally-representative online panel survey of 4,111 participants, February-July 2016. To construct an index, polychoric factor analysis (suitable for ordinal variables), was conducted on four survey items related to the perceived impact of roads on ability to walk locally: 1) traffic volume, 2) traffic speed, 3) availability of crossing points, and 4) adequate crossing time. Community severance index scores were negatively skewed, and were thus categorised into four groups (lowest 40%, second, third and highest). Logistic regression was used to examine the association of community severance with self-rated health (‘good’ (very good/good) vs. ‘poor’ (fair/bad/very bad)), adjusting for potential confounders (age, income, employment status). We also examined effect modification of the severance and self-rated health association by environment type (urban/not urban) and the presence of a car in the household.

Results Polychoric factor analysis confirmed that it was appropriate to combine the four survey items into a single index (Cronbach’s Alpha=0.86; Keiser-Meyer-Olkin measure of sampling adequacy= 0.76, all factor loadings >0.74). After controlling for confounding factors, being in the highest community severance index group was associated with higher odds of reporting poor self-rated health (Odds Ratio: 1.79, 95% Confidence Interval: 1.48–2.17) compared with the lowest scoring group. There was a dose-response gradient, with those in the second and third highest groups having increased odds of reporting poor self-rated health, though of lower magnitude ((OR 1.21, 1.01–1.45) and (OR 1.41, 1.16–1.71) respectively).

Conclusion We found an inverse association between community severance index and self-rated health. This suggests that to improve health, local governments and road authorities should take steps to reduce community severance through traffic reduction and calming, pedestrian prioritisation, and the installation of well-designed crossing points.