Background In European countries, the self-rated health (SRH) item has been widely used to measure respondents’ self-reported overall health, because of its satisfactory reliability and validity and its usefulness. India is an emerging country and has implemented the SRH item in many of its population and health surveys. However, the validity of SRH between high and low income countries is not known: does SRH measure the same thing in India and in a European context? The objective is to assess the construct validity of SRH in India and Switzerland.

Methods Data on 9,228 Indians residents (World Health Survey 2003) and 16,651 Swiss residents (Swiss Health Interview Study 2007) were used. SRH was the dependent variable and health predictors – comparable between the two surveys – were grouped into five dimensions: physical (four variables), mental (five variables) and functional (two variables) health, chronic diseases (three variables) and health behaviours (five variables). Regression models were used to assess the contribution of each predictor separately for women and men, in both countries. The robustness of findings was checked by using different coding of SRH (linear, binary, etc.). All analyses were adjusted for sociodemographic and socioeconomic factors and conducted with SPSS 22.

Results Analyses showed that almost all health predictors – representing the five health dimensions – were associated with SRH, for men and women, in both countries. Among men, SRH variance was best explained by functional health (India coefficient of determination (R²) 0.375, Switzerland R² 0.214), then by mental health (India 0.262, Switzerland 0.123), physical health (India 0.215, Switzerland 0.175), chronic diseases (India 0.170, Switzerland 0.103) and health behaviours (India 0.145, Switzerland 0.088). Among women, the pattern was similar. R² was systematically higher in India compared to Switzerland, among men and women.

A few inter-country differences were observable. For example, obesity was negatively associated with good SRH in Switzerland among men (B −0.172 95% CI [−0.220, −0.124]) and women (B −0.162 95% CI [−0.207, −0.116]), while it was positively associated in India among men (B −0.111 95% CI [0.026, 0.195]) and not associated among women.

Conclusion We found emerging evidence of similar construct validity of SRH between India and Switzerland, suggesting that SRH measures the same dimensions of health. Moreover, explained variance was systematically higher in India compared to Switzerland, suggesting a better performance of SRH in this country. However, due to the limited number of comparable health variables in these surveys, further research needs to confirm these preliminary findings.