In 87 older non-demented adults with BD who presented for treatment of mania, overall cognitive performance was lower than expected based on age-adjusted norms. The cognitive domains most severely affected were memory and visuospatial ability, while the domains most frequently affected were memory and abstraction. Congruent with previous reports, individuals with psychotic symptoms had significantly worse cognitive performance than those without psychosis. However, contrary to previous reports, the cognitive performance of subjects with or without a history of substance use (largely alcohol use disorder) did not differ significantly. We found that later age of first manic episode and greater vascular disease burden were related to lower memory performance. To our knowledge, our report is based on the largest sample of elderly adults experiencing mania.
In accord with certain published reports, later age of first manic episode was related to vascular disease burden in this sample (26
). Later age of first manic onset was also related to lower memory performance. Worsening cerebrovascular disease may reasonably be related to both later age of first manic episode and memory impairment. Our findings, in context of the current literature, suggest that through different pathophysiological mechanisms greater lifetime BD severity (BD Specific Factors; see
), via hypercortisolemia, and later age of onset, via vascular disease (BD Co-Morbidity Factors; see
), may impair memory. There is arguably a BD “vascular subtype,” similar to the vascular unipolar depression hypothesis of Alexpoulos (40
), where vascular brain changes are related to mania and cognitive decrements in late-life (41
Based on reports in middle-aged adults (8
), we expected, but did not find a relationship between vascular disease burden and executive dysfunction in the DRS Initiation/Perseveration subscale. Further, we did not observe a synergistic relationship between lifetime duration of BD and vascular disease burden. BD lifetime duration in itself does not appear to render individuals with BD at greater risk for overall cognitive impairment due to vascular disease burden. Lifetime duration of BD may not capture the critical illness features over the lifespan that affect cognition, such as frequency, duration, and intensity of mood episodes. In older adults, lifetime duration may be a more limited proxy for BD severity than in middle age.
Certain limitations need to be considered in interpreting our findings. First, the DRS does not provide the highly-detailed assessment of performance across cognitive domains that would be available through more comprehensive neuropsychological assessment: it does not provide a measure of information processing speed and it is not especially sensitive for examining executive dysfunction (43
). Second, while only 4.3% of screened subjects were excluded due to cognitive impairment, the generalizability of our findings to older adults with BD outside of an RCT is presumably limited by the selection criteria necessary for the study (see Methods). Hence, our findings may under-represent cognitive dysfunction in older adults with BD because the sample may have been skewed towards healthier older adults who could tolerate lithium or divalproex. Third, serum anticholinergicity, which can negatively impact cognitive performance, was not measured (44
). Last, although the positive associations found in the subscale scores are biologically plausible, the possibility that multiple exploratory hypotheses testing led to type I error needs to be considered.
In addition to the relatively large sample, the strengths of our report include a well-characterized group of patients with BD I, all over the age of 60 years, who underwent a careful psychiatric and general medical assessment. A notable difficulty in conducting research in older adults with BD, who are not acutely ill, is making certain that the study group consists entirely of subjects with BD. Focusing on elders who are acutely manic identifies a group of patients that all have BD and mitigates this concern about diagnostic heterogeneity. Further, while cognitive dysfunctions are present in all BD mood states, they may be particularly evident in individuals who are manic (5
Our report shows that older adults experiencing mania exhibit a pattern of executive and memory dysfunction similar to what is found in younger adults with mania. Noting the above limitations, we find that longer lifetime duration of BD is not related to greater overall cognitive impairment in non-demented elders with BD. This may be a reassuring message for individuals with BD, who are struggling with a lifelong relapsing and remitting illness. We did not observe an interactive effect of lifetime duration of BD with vascular disease burden. While we could not detect an impact of prior lithium or divalproex exposure on cognitive function, our study did not examine duration or intensity of treatment with either agent.
Our findings preliminarily argue for the necessity of individuals with BD to take care of their general medical problems and in particular, risk factors for vascular disease. Patients with BD and their health care providers generally view their psychiatric care as the most important aspect of medical care (46
). This view has led to a relative under-recognition and inattention to many of the physical diseases these patients experience. Appropriately addressing medical problems related to vascular disease, such as hypertension, diabetes mellitus, and hypercholersterolemia, may protect against some elements of cognitive dysfunction observed in older adults with BD. Future studies will need to carefully examine the relationships among severity of illness (numbers, durations, and intensities of mood episodes), treatment exposure, medical comorbidity, and cognitive dysfunction to develop interventions intended to prevent, halt, or reverse cognitive dysfunction in patients with BD.