METHODS: In the first stage, a preliminary list of items measuring socio-cultural and religious beliefs during menstruation was generated. In the second stage, exploratory factor analysis was performed. Finally, confirmatory factor analysis using reflective measurement model and structural equation modelling was performed using partial least squares. The practices of these beliefs were included as mediating effect. Biological symptoms of menstruation were added in as another factor.

RESULTS: A total of 400 female students from the Universiti Malaysia Sarawak, Malaysia were recruited. A preliminary list of 22 items was first generated. From the confirmatory factor analysis, two factors were iteratively removed due to poor factor loadings. Four factors were retained, i.e., i) "religious beliefs"; ii) "unpleasant (or dirty) nature of menstruation"; iii) "personal restrictions (dietary and behavior)"; and iv) "restrictions of interactions with male gender". In structural equation modelling, only 2 factors, i.e., the practices of "personal restrictions (dietary and behavioural)" and "restriction of interactions with males" had significant negative impact on quality of life.