The association between health literacy and subjective and objective physical activity levels: a cross-sectional analysis

Studies show that high levels of health literacy (HL) can support the promotion and maintenance of healthy behaviour such as physical activity (PA). However, most of these studies rely on subjective data. Hence, the aim of the present study was the investigation of the relationship of HL and PA, not only with subjectively (questionnaire) but also with objectively (accelerometry) measured PA data. Methods: The present study is a pooled analysis of baseline data from the research association TRISEARCH (2015-2018), which focused on HL of working adults. HL was measured by Lenartz’ questionnaire, which consists of six subscales (self-regulation, self-control, self-perception, proactive approach to health, communication and cooperation, and dealing with health information). Subjective level of PA was assessed using the Global Physical Activity Questionnaire (n=1056). A subsample (n=124) also received accelerometers (Actigraph GT3X+) to provide additional objective PA data. Partial correlations (covariates: age, sex) were used to investigate the relationship of HL and questionnaire- and accelerometer-derived PA. Two linear regression models were calculated using subjectively and objectively assessed PA as dependent variables, respectively. The six subscales of HL, sex (male vs. female) and age were used as independent variables. Results: Daily moderate to vigorous PA was 102.4 (±141.9) min according to the questionnaire and 57.8 (±38.8) min measured by accelerometry. Very weak partial correlations could be found for HL subscales and daily PA by questionnaire (self-regulation, self-control; each r=-.06, p<.05) and accelerometer (proactive approach to health; r=.26, p<.01). Linear regression models revealed no statistically significant contribution of any subscale of HL or individual related variables (age and sex) to the amount of daily PA (all p>.05, GPAQ: adjusted R²=.003, Actigraph GT3X+: adjusted R²=.05). Conclusion: In contrast to previous findings, only weak partial correlations and no statistically significant influence of HL on PA could be found in the present study. Hence, it seems that not all aspects of HL influence the occurrence of healthy behaviour – such as PA – in the same way. This should be considered when HL-based interventions are designed. Further studies which use objectively measured HL as well as PA data are needed for a better understanding of the relationship of HL and PA. Support/Funding Source: This work was supported by the German Federal Ministry of Education and Research [reference: 01EL1425A].