- Potentially confounding variables were identified from a review of all previous NTD case-control studies

- Only maternal age, education, obesity, and previous still birth or miscarriage proved to be important risk factors or confounders in preliminary analysis

- These were later included in the adjustment of estimates