Validation of the 24-h perceived exertion recall survey in women in rural Tigray, Ethiopia

BACKGROUND: Physical activity affects nutritional status and health. Currently, there are few validated survey tools for estimating physical activity in rural areas of low-income countries, including Ethiopia, which limits the ability of researchers to assess how physical activity affects nutritional status. OBJECTIVES: This study used accelerometry to validate 2 in-person questionnaires, the global physical activity questionnaire (GPAQ) and the 24-h perceived exertion recall survey (PERS). METHODS: This study recruited 180 women aged between 18 and 45 y living in rural Tigray, Ethiopia. Participants had previously participated in an impact evaluation of a public work safety net. They wore an accelerometer for 8 d and responded to perceived exertion questionnaires twice. Data were collected on 89 women during the short rainy period and 91 women during the main rainy season. A survey method was considered valid if the proportion of time spent in moderate or vigorous physical activity (MVPA) levels had a Pearson’s correlation coefficient of >0.40 to the proportion of time spent in MVPA recorded by accelerometry. RESULTS: The GPAQ had high reliability, but the overall validity was poorer than accelerometry. The proportion of time spent in MVPA according to the accelerometer was associated with discordance between GPAQ and accelerometry. MVPA levels, as measured by the 24-h PERS, had a fair agreement with accelerometry. The agreement increased to moderate/acceptable when adjusted for season and BMI. CONCLUSIONS: The 24-h PERS is a valid tool for estimating the physical activity of women living in rural highland Ethiopia. It can be used in future research to understand the physical activity demands of living in rural highland Ethiopia, enabling more targeted programs to address undernutrition.