Menstrual cycle variations in the BOLD-response to a number bisection task: implications for research on sex differences.

Numerical processing involves either number magnitude processing, which has been related to spatial abilities and relies on superior parietal regions, or arithmetic fact retrieval, which has been related to verbal abilities and involves the inferior parietal lobule. Since men score better in spatial and women in verbal tasks, we assume that women have advantages in fact retrieval, while men have benefits in number magnitude processing. According to findings on menstrual cycle variations in spatial and verbal abilities, fact retrieval should improve during the luteal phase and magnitude processing during the follicular phase. To dissociate sex- and menstrual cycle-dependent effects on fact retrieval and number magnitude processing, we applied a number bisection task in 15 men and 15 naturally cycling women. Multiplicative items (e.g. 12_15_18) are part of a multiplication series and can be solved by fact retrieval, while non-multiplicative items (e.g. 11_14_17) are not part of a multiplication series and require number magnitude processing. In men and women in their luteal phase, error rates were higher and deactivation of the medial prefrontal cortex and the bilateral inferior parietal lobules was stronger for non-multiplicative compared to multiplicative items (positive multiplicativity effect), while in the follicular phase women showed higher error rates and stronger deactivation in multiplicative compared to non-multiplicative items (negative multiplicativity effect). Thus, number magnitude processing improves, while arithmetic fact retrieval impairs during the follicular phase. While a female superiority in arithmetic fact retrieval could not be confirmed, we observed that sex differences are significantly modulated by menstrual cycle phase.