3.2.4. Short sleep effect: by cognitive domain and study type The associations between self-reported short sleep and performance on various cognitive domains were still observed when we
considered cross-sectional studies only (Table 2).
Self-reported short sleep was associated with higher odds for deficits in multipledomain tasks, executive functions, verbal memory, and working memory capacity, with ORs ranging from 1.30 to 1.38.
A signifi-cant association was not present for speed of processing.
In prospective cohort studies, self-reported short sleepers showed elevated risk for poor performance only in multiple-domain tasks (OR = 1.44, 95% CI = 1.02–2.04). No significant association was found in other domains (Table 2), and this was likely due to the small number of prospective cohort studies which assessed performance in these cognitive domains (k = 2–4).