Are cognitive deficits similar in remitted early bipolar I disorder patients treated with lithium or valproate? Data from the STOP-EM study

In bipolar disorder (BD), lithium and valproate are both reportedly associated with mild cognitive deficits with impaired psychomotor speed and verbal memory ascribed to both while impairments in learning and attention are mainly attributed to valproate. However, there are few direct comparisons of the impact of lithium and valproate on cognitive function in early BD. Using data from the STOP-EM study, we compared neurocognitive functioning in BD patients, who had recently recovered from a first episode of mania, and were on treatment with lithium (n=34) or valproate (n=38), to a comparable sample of healthy controls (HC; n=40), on the domains of processing speed, attention, verbal memory, nonverbal memory, working memory and executive functions. The three groups were comparable on socio-demographic (all p>0.12) and clinical variables (all p>0.08). MANOVA revealed a significant difference between the three groups on overall cognitive functioning (Wilk׳s lambda=0.644; F= 3.775; p<0.001). On post-hoc Tukey test, the valproate group performed poorer on working memory compared to the lithium (p=0.001) and HC groups (p<0.001). There was no significant difference between the lithium and valproate groups on other cognitive domains (all p>0.13). Treatment with valproate and not lithium may be associated with working memory deficits early in the course of BD.