When each variable was entered into the first step of hierarchical linear regression analysis, perceived stress explained more than 40% of the variance in depressive symptoms, followed by felt stigma, at about 20%.
Using a cohort of adults with chronic epilepsy, findings showed that psychological factors such as perceived stress and felt stigma were the most significant correlates of depressive symptoms, followed by social and epilepsy-related factors.1
Epilepsy-related felt stigma refers to the shame of living with epilepsy and the fear of encountering actual episodes of social discrimination based solely on this condition. In this multicenter, cross-sectional study, 316 adults with epilepsy had depression assessed using the Beck Depression Inventory (BDI) and perceived stress evaluated using the 10-item Perceived Stress Scale (PSS-10). Felt stigma was determined using the 3-item Stigma Scale for Epilepsy (SSE), with individuals categorized as positive for felt stigma if they answered “yes” to at least 1 of the 3 items.
Findings on both crude and adjusted models showed that interactions between SSE scores of at least 1 and PSS-10 scores significantly affected BDI scores (P <.05 each). Specifically, in an adjusted model, BDI scores were more strongly associated with PSS-10 scores (ß = 1.699; P <.001; partial eta2 = 0.225) in individuals with SSE scores of at least 1 than those with scores of 0 (PSS-10 scores, ß = 1.115; P <.001; partial eta2 = 0.205).
Lead author Sang-ahm Lee, MD, PhD, Department of Neurology, University of Ulsan College of Medicine, and colleagues excluded patients if they had a seizure during the 48 hours prior to being asked to complete the questionnaire, if they had a neurological deficit that affected their quality of life, and if they were taking regular medication for medical disorders or psychiatric illnesses. BDI scores indicative of depression were not excluded, but only if they were not taking medication to treat depression.
WATCH NOW: Resources for Dravet Syndrome and LGS
Most of the cohort had focal epilepsy (87%) and most experienced at least one seizure recurrence during the previous year (72.5%). Abnormal BDI scores were recorded in 216 individuals (68.4%), with mild-to-moderate depression as the most commonly reported (30.1%). Scores of at least 30, considered severe depression, were found in 40 subjects (12.7%). Of the cohort, 109 individuals (34.5%) reported feelings of stigma, with 47 (14.9%) feeling highly stigmatized, as represented by answers of "yes" to all 3 items.
With a mean BDI score of 16.6 (±11.2), univariate analyses showed that BDI scores correlated significantly with higher PSS-10 scores, SSE scores of at least 1, older age, lower educational level, unemployment status, higher seizure frequency, recurrence of generalized tonic-clonic seizures of FBTCS, and antiseizure polytherapy. By contrast, BDI scores did not correlate with sex, marital status, age at seizure onset, or duration of epilepsy.
The linear regression analysis, which highlighted the most important correlates of depressive symptoms, explained 54% of the variance of BDI scores. Differential effects of significant factors were examined by entering each variable into the first step of the regression, thereby determining its independent contribution to the dependent variable, and by entering the same factor into the last step of the regression. Sociodemographic factors, entered in step 1, explained 11.9% of the variance. Here, investigators found that unemployed status (ß = 3.622; P = .001) and being married (ß = 2.182; P = .049) were associated with higher BDI scores.
Epilepsy-related factors, entered in step 2, explained an additional 5.4% of the variance. In the final step, the addition of psychological factors explained an additional 36.7% of the variance. Both PSS-10 scores (ß = 1.306; P <.001) and SSE scores of at least 1 (ß = 4.680; P <.001) were independently associated with a higher BDI score. PSS-10 scores accounted for the largest amount of the variance when each factor was entered in the first (42.4%) or final (23.5%) step of the hierarchical linear regression analyses.