This preliminary study examined the associations between clinical symptoms, social function and neurocognitive function among 81 euthymic BD patients stratified by occupational status (employed/unemployed)and subgroups of DSM-IV BD ( BD-I and BD-II). In line with earlier findings (DelBello et al., 2007; Wingo et al., 2010), our findings showed that employed BD patients displayed greater social functioning (autonomy, occupational functioning, and interpersonal relationships) and performed better in verbal learning and processing speed than unemployed patients, indicating that employed BD patients could develop stronger social functions and some neurocognitive functions. In a previous cluster analysis study with euthymic BD patients, the group with lower functioning, measured by FAST scores, showed the highest unemployment rate, which indicated the main areas of functional loss in autonomy, occupational functioning, cognition and interpersonal relationships (Solé et al., 2018). This is basically consistent with our findings. Some studies reported that employed BD patients showed greater neurocognitive functioning measured by the Assessment of Neuropsychological Status (RBANS), especially in the verbal memory domain (Dickerson et al., 2004). Another study clearly showed that BD patients who have jobs may show better executive function than those without jobs (Altshuler et al., 2007), which is consistent with our results.
Further correlation analysis under stratification showed that social function outcomes had close ties with neurocognitive function among employed BD patients. There was a stronger correlation between neurocognitive function and the occupational and interpersonal relationship domains in the FAST, followed by the financial issues and autonomy domains. Verbal learning measures were predominantly associated with the occupational domain of the FAST in the employed group. Previous studies have demonstrated a strong correlation between verbal learning ability and occupational status (Dickerson et al., 2004), and our results confirmed this association. In addition, processing speed was also significantly correlated with the leisure time and interpersonal relationship domains in employed BD patients. we considered that employed person is more likely to participate in social activity, to facilitate self-management, and to promote learning and information-processing capacity. Many papers (Andreou et al., 2013; Bonnin et al., 2014; Bora, 2018; Duarte et al., 2016; Konstantakopoulos et al., 2016; Sanchez-Moreno et al., 2018; Szmulewicz et al., 2018) have reported that social functional impairment could be associated with neurocognitive measures. Remarkably the measures of processing speed, visual memory and verbal learning were powerful determinants of functional impairment in these studies, consistent with our findings for the employed group. Bearden et al. also evidenced that baseline cognitive impairment across multiple domains, particularly working memory and speed of processing, were significantly associated with concurrent occupational function impairment (Bearden et al., 2011). Jaeger et al. found that baseline attention and speed of processing domains could predict functional outcomes (including occupational function) over a 12-month period (Jaeger et al., 2007). Most jobs require relatively strong abilities in the areas of learning, memory and processing speed. Thus, the improvement of these functions may benefit occupational performance. Unfortunately, we found that the correlation between neurocognitive function and social function significantly weakened in the unemployed group compared with the employed group (see Figs. 3 and 4), demonstrating the poorer association between social function and neurocognition in unemployed BD patients. We presume that occupational status may be a core factor in promoting overall functional development in euthymic bipolar patients.
Our findings also testified the neurocognitive differences between the BD subtypes. There were no differences in neurocognitive measures between the BD subtypes, which is in line with earlier findings (Dittmann et al., 2008). Dittmann and his colleagues evaluated psychomotor speed, working memory, verbal learning, visual ⁄ constructional abilities and executive functions in euthymic bipolar patients by using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) (Randolph et al, 1998) and the Trail Making Test (TMT) A and B (Lezak et al., 2004, Lu and Bigler., 2000). They also concluded that patients with both BD subtypes had similar levels of neurocognitive deficits. Similarly, another meta-analysis reported that neurocognitive differences between clinical BD subtypes were very subtle and not significant (Bora, 2018). However, some studies have obtained different conclusions; for example, one study indicated that BD I patients performed worse than BD II patients in all areas of cognitive function except for working memory (Schenkel et al, 2012). Another meta-analysis revealed that BD-II patients may present deficits in working memory and executive function, and more than half of the studies showed worse verbal memory. This moderate difference between BD subtypes may be complicated due to residual symptoms, number of episodes, age at illness onset, etc. (Solé et al, 2011).
Furthermore, significant differences in FAST scores and its six domains between the BD subtypes showed that BD-I patients had excellent total functioning, especially in occupational functioning and interpersonal relationships. Our findings are consistent with previous studies that showed that BD-I patients had better social function than BD-II patients did (Vinberg et al., 2017; Kupka et al., 2007; Vieta, 2010). We considered that BD-I patients with mania were more likely to seek opportunities for social activity and interpersonal communication. In contrast, Dell'Osso and his colleagues found that bipolar patients with remitted BD and BD-I had significantly lower GAF scores than those with BD-II (Dell'Osso et al., 2007). The reasons for the discrepancy in functional outcomes may be related to functioning measures, duration of illness, education levels, and pre-disease cognitive performance (Chen, et al., 2019; Baune, et al., 2015). Besides, our correlation analyses did not suggest functional differences between the BD subtypes, indicating the same associations between processing speed and occupational functioning and interpersonal relationships both in the BD-I and BD-II patients (Duarte et al., 2016). We speculate that both BD-I and BD-II patients may exhibit similar functional impairments even in euthymia.
Our results did not show significant differences in demographic characteristics, clinical measures or social or neurocognitive functioning measures between genders. Some studies have reported that male patients could show poorer social and neurocognitive functional outcomes than female patients (Sanchez-Moreno et al., 2018; Vaskinn et al., 2011), but Bücker et al. reported that males had better working memory and sustained attention than females (Bücker et al, 2014). Further study is needed to verify this gender difference in total functioning based on a larger sample size.
Several limitations of this study should be taken into consideration. First, cognitive functional assessments are scarce and do not reflect all dimensions of neurocognitive functioning in BD patients. The outcomes with no significant differences in neurocognitive measures between BD subtypes need to be tested using full-scale instruments in future research. Second, although occupational status can be viewed as an advantage of this study, current occupational status does not reflect the quality or stability of job performance, which is a particularly important area for understanding functional recovery and deserves further clarification. In addition, some studies have pointed out that the definition of occupation may influence the final assessment results (Duarte et al., 2016; kozma et al., 2010). A prior study defined employment as including fulfilling domestic responsibilities at home or attending school (Altshuler et al., 2007), thus this association between total functioning and occupational status should be cautiously elucidated. Third, the possible effects of psychotropic medications on social function and neurocognitive function should be considered, and typical antipsychotics and partial mood stabilizers could hinder total functioning performance (Dias et al., 2012). Finally, the sample size was not sufficient, which could reduce the statistical power when conducting comparison and correlation analysis.