|Home | About | Journals | Submit | Contact Us | Français|
Epilepsy research efforts have primarily focused on medical treatment and physical management of epilepsy; however, to provide comprehensive care, efforts cannot focus solely on physical manifestations of epilepsy. Research findings show that people with epilepsy face many challenges that can negatively affect quality of life (QOL). In this descriptive study, we examined the individual relationships between depressive symptoms, stigma, social support and regimen-specific support and QOL in adults with epilepsy. Study data were obtained from a subset of patients (N=147) who participated in a longitudinal study of adult patients with epilepsy. Measures of QOL, depressive symptoms, stigma, social support and regimen-specific support were analyzed to answer the research questions. The results of correlational analyses revealed statistically significant negative correlations between depressive symptoms, stigma and sometimes regimen-specific support and QOL and statistically significant positive correlations between social support and QOL. A hierarchical multiple linear regression model revealed that depressive symptoms accounted for the most variance in QOL. Psychosocial variables measured 3 months prior to QOL were entered into a hierarchical multiple linear regression model, revealing that depressive symptoms, stigma and social support can be used to predict QOL at a later time.
Epilepsy, a chronic neurological condition in which recurrent seizures originate from abnormal electrical signals in the brain, is one of the most common brain disorders [1, 2]. Although primary treatment goals focus on the physical complications of epilepsy through seizure control [1, 3, 4], people with epilepsy face many other challenges, including problematic psychological manifestations (e.g. poor self-esteem, fear and anxiety) and social complications (e.g. driving restrictions, unemployment and social isolation) [3, 5–9], all of which can impact quality of life (QOL).
The World Health Organization  conceptually defines QOL as the perception of a person regarding his or her life, including the aspects of his or her ‘physical health, psychological state, level of independence, social relationships, personal beliefs and their relationship to salient features of the environment’ (p. 3). Many researchers agree that QOL is a multidimensional concept involving one's physical, social and psychological life perceptions [5, 6, 9, 11]. A little over a decade ago, researchers in epilepsy began to study QOL in persons with epilepsy . Researchers have found, almost without exception, that QOL is compromised in persons with epilepsy [5, 8, 9, 12–15] and that seizure control, both in frequency and in severity, seems to play a dominant role [5, 8, 9, 14–17]. Those with well-controlled seizures tend to have a QOL level near that of the general population [13, 14, 18–20], whereas QOL appears to diminish as seizure control decreases.
While the association between seizures and QOL is well understood, investigators have not thoroughly assessed factors affecting QOL. The present study builds upon previous research, continuing to explore factors associated with QOL. In this study, we focused on the psychosocial factors of depressive symptoms, stigma, social support and regimen-specific support.
The rationale for selecting depressive symptoms, stigma, social support and regimen-specific support was based on the literature attesting to their importance in understanding QOL. Psychiatric disorders, especially mood disorders, are commonly recognized as comorbidities of epilepsy; depression is the most commonly diagnosed mood disorder among people with epilepsy [21–24]. Researchers have also found that, compared with the general population, people with epilepsy are more likely to have depressive feelings or be diagnosed with depression . Prevalence of depression in patients with intractable seizures, ranging from 9 to 55% and 20 to 25%, is significantly higher than the prevalence of depression in the general population, ranging from 1 to 3% in men and 2 to 9% in women [22, 25–27]. Investigators studying the relationship between depression and QOL in people with epilepsy have found that epilepsy diagnosis and depression are independent, significant predictors of health-related QOL . Although investigators have only recently evaluated depression in light of QOL, they have found that adult patients with epilepsy who were also identified as having depression, whether major or minor, reported a much lower QOL, especially among patients with uncontrolled seizures [24, 28, 29].
Researchers have long since recognized that epilepsy carries with it a certain stigma among society or at least a perceived stigma among persons with epilepsy [30–33]. Stigma can be defined as a label assigned to an individual or a feeling of being designated as different from the normal population due to a life condition, invoking feelings of dread or fear among others . Although not all people with epilepsy feel stigmatized, studies have shown that 30–51% of participants reported feeling stigmatized, and 18% reported feeling highly stigmatized [26, 33, 35]. Perceived stigma has been found to be significantly associated with higher levels of psychological distress symptoms  and lower QOL .
Interactions with society, as well as relationships between friends and loved ones, are often negatively affected by the consequences of epilepsy [32, 37, 38]. Social interactions, however, have the opportunity to carry a positive impact when resulting in supportive actions and behaviors. Supportive relationships create an environment of encouragement, sharing positive life events and cushioning negative situations . Among people with epilepsy, support can also include helping people with their treatment regimens, such as reminding one to take his or her medications. Investigators have specifically noted that social support and its related concept regimen-specific support positively relates to self-management practices and correlates with a lower level of unpredictability in epilepsy [40, 41]. Moreover, relationships that act supportively in the lives of those with epilepsy have been known to positively impact QOL perceptions [9, 42]. In addition to its direct relationship with QOL, social support is indirectly related to QOL through the person's mental health status, where mental health is positively correlated with QOL . QOL is also higher among people with epilepsy who feel more comfortable in their social environment [42, 43].
Due to the lack of research regarding psychosocial factors and QOL in persons with epilepsy, the purpose of this descriptive study was to examine the relationships between depressive symptoms, stigma, social support and regimen-specific support and QOL among adults with epilepsy. We first asked which variables are associated with QOL at the same point in time (cross-sectional analysis) and which variables are the best predictors of QOL when QOL was measured at a future point in time (longitudinal analysis). We sought to provide a better evidence-based understanding of these relationships to inform the development of interventions and education materials for persons with epilepsy.
The sample for the present analysis consisted of a subset of patients from the total number of participating patients (N=320) in the longitudinal study, Project EASE (Epilepsy Awareness, Support and Education). The specific aims of Project EASE were to examine self-management practices of patients with epilepsy and to test a social cognitive model of self-management.
Project EASE participants were recruited from three epilepsy clinic sites—two in Atlanta, GA, and one in Boston, MA. Patients attending regularly scheduled visits were asked by their clinicians to consider participating in the study. Interested patients were referred to a study nurse who assessed the individuals for eligibility. More detailed information about inclusion and exclusion criteria and study procedures for Project EASE is available in DiIorio et al. . Participants began Project EASE with a baseline interview and then participated in two follow-up interviews at 3 months (Interview II) and 6 months (Interview III) after baseline. During Interview III, participants were administered a QOL scale. Due to its late incorporation into the study, as well as its administration only at Interview III, not all 320 Project EASE participants completed the QOL scale. Since this study pertains to QOL, only those participants who completed the QOL measure during Interview III (n=147) were included in this analysis. Depressive symptoms, stigma, social support and regimen-specific support were measured at both Interviews II and III. The research design and procedures of Project EASE and this study were approved by the Institutional Review Boards of the investigators' institutions and the appropriate research committees; participants signed informed consent forms.
QOL was measured using the 31-item Quality of Life in Epilepsy (QOLIE-31) scale, adapted from the more comprehensive 89-item scale and designed to address key epilepsy QOL issues [6, 45]. The QOLIE-31 contains seven subscales and one question addressing overall health; subscale concepts are emotional well-being, social function, energy/fatigue, cognitive function, seizure worry, medication effects and overall QOL. Scores are obtained by calculating a weighted average of the subscales using a scoring manual provided by the QOLIE-31 development group . The QOLIE-31 instrument has been reported to be both reliable and valid [6, 46, 47]. The scale was found reliable for the present sample (α=0.74–0.88 for subscales and α=0.94 for the overall scale).
The 20-item Center for Epidemiologic Studies Depression Scale (CES-D scale) was used to measure symptoms of depression . Items are scored on a four-point Likert scale according to how frequently an individual experiences the item. An overall score is obtained by summing individual responses; higher scores indicate more depressive symptoms. Previous study has revealed supportive evidence that the CES-D scale is both reliable and valid among adults . The alpha coefficient for the present sample was 0.94 and 0.92 for Interviews II and III, respectively.
Stigma was measured by a 10-item scale adapted from the Parent Stigma Scale developed by Austin et al. . Items from the original scale were reworded to direct statements toward adults with epilepsy, while maintaining the same concept. Individual item responses on a seven-point Likert scale are summed to yield a total score; higher scores are associated with higher stigma perception. The original scale was found to be reliable in a study population of parents of children with epilepsy . Results from the stigma scale used in Project EASE at Interview I (N=314) indicate that it is a reliable scale . In this study, the alpha coefficient was 0.89 and 0.91 for Interviews II and III, respectively.
Social support was measured by the second part of the Personal Resource Questionnaire (PRQ), developed as a measure of perceived social support [50, 51]. The PRQ85-part 2 is a 25-item scale with items scored individually on a seven-point Likert scale. Overall scores are obtained by summing the individual item scores; higher scores are associated with higher levels of perceived support. Investigators have found the PRQ85-part 2 to have been a reliable measure within samples of adults with epilepsy [9, 41]. In the present study, the alpha coefficient was 0.93 for both Interviews II and III.
Regimen-specific support was measured using a nine-item Epilepsy Regimen-Specific Support scale . The original scale consisted of a seven-item, five-point Likert scale, measuring a patient's perceived available support for assisting in the completion or maintenance of epilepsy-related tasks. In 1999, one item was slightly modified and two items were added. Scores are obtained by adding individual item responses. Higher scores relate to a higher level of regimen-specific support. In this study, alpha coefficients were 0.89 and 0.91 for Interviews II and III, respectively.
Data on demographic characteristics of age, gender, race, marital status, education, employment, income and study site and epilepsy-related information of mean age that seizures began, seizure occurrence in the past year, activity restriction and other chronic conditions were measured at each interview for data analysis.
Descriptive statistics were conducted for demographic characteristics and epilepsy-related variables (Interview II) and for the study variables—QOL, depressive symptoms, stigma, social support and regimen-specific support (Interviews II and III). Correlation analyses using Pearson product moment correlation were conducted to assess associations among the study variables at Interviews II and III. Because QOL data were only collected at Interview III, QOL was only included in cross-sectional analyses at Interview III. Correlation analyses were also conducted using depressive symptoms, stigma, social support and regimen-specific support measured at Interview II with QOL measured at Interview III. Finally, hierarchical multiple regression analyses were used to assess QOL. The first model (cross-sectional) included all demographic variables (first block), epilepsy-related medical history variables (second block) and the Interview III psychosocial variables demonstrating a significant correlation with QOL (third block). The second model (longitudinal) included all demographic variables (first block), epilepsy-related variables (second block) and the Interview II psychosocial variables demonstrating a significant correlation with QOL (third block). Results of initial model testing for both cross-sectional and longitudinal models were used to refine the models and create final models.
The study sample consisted of 147 adult patients with epilepsy from Boston and Atlanta study sites. Sample demographics and epilepsy-related variables are displayed in Table I. Participants' ages ranged from 19 to 75 years, with a mean of 45 years of age. There were slightly more males than females, most were Caucasian, were married or living with a partner and had received at least some college education. Less than half were employed full- or part-time; most reported an annual income of $30000 or less. The mean age of participants at the time of their first seizure was about 23 years. Slightly over half reported having at least one seizure during the year prior to the interview, and most reported that their seizures had prevented them from doing at least some of the activities that they had wanted to do. Most participants reported having an additional chronic medical condition; hypertension, depression and hypothyroidism were the most common conditions reported.
Descriptive analyses of the study variables revealed that the mean overall score for the QOLIE-30 was 51.21 (SD=11.27). The mean scores for depressive symptoms on the CES-D were 14.54 (SD=13.02) for Interview II and 9.50 (SD=11.78) for Interview III. Both CES-D scores were slightly lower than the score of ≥16.0, indicative of possible clinical depression. However, about 35 and 38% of participants (Interviews II and III, respectively) did show symptoms of possible clinical depression by a score of ≥16.0.
Results of the tests of association for Interview II variables indicate that depressive symptoms (−0.711, P<0.001), stigma (−0.513, P<0.001) and regimen-specific support (−0.219, P=0.009) demonstrated a significant negative association with QOL. Conversely, social support (0.458, P<0.001) demonstrated a significant positive association with QOL. At Interview III, similar statistically significant associations between depressive symptoms (−0.829, P <0.001), stigma (−0.558, P<0.001) and social support (0.513, P<0.001) with QOL were observed. However, in Interview III, regimen-specific support showed a very weak negative association (−0.082, P=0.326) with QOL that were not statistically significant. It was hypothesized a priori that depressive symptoms and stigma would be negatively correlated with QOL and that social support and regimen-specific support would be positively correlated with QOL; thus, the significant negative correlation of regimen-specific support obtained at Interview II and no relationship at Interview III were not expected.
The initial regression model for Interview III (cross-sectional) accounted for 80.7% of the variance in QOL, with the psychosocial study variables accounting for an additional 35.1% of the variance in QOL beyond the demographic and epilepsy-related variables. In this model, age, employment, seizures in the past year, activity restriction, depressive symptoms and stigma were significant. A final model was created using these variables. The final model, accounting for 78.6% of the variance in QOL, revealed that not being employed, having seizures in the past year, reporting greater activity restriction and having higher levels of depressive symptoms and stigma were all statistically significant predictors of lower QOL (Table II).
The initial regression model for Interview II (longitudinal) accounted for 72.4% of the variance in QOL, with the psychosocial study variables accounting for an additional 19.8% of the variance in QOL beyond the demographic and epilepsy-related variables. In this model, education, seizures in the past year, activity restriction, depressive symptoms, stigma and regimen-specific support were statistically significant. A final model was created using these variables. The final model, accounting for 68.8% of the variance in QOL, revealed that having seizures in the past year, reporting greater activity restriction and having higher levels of depressive symptoms, perceived stigma and regimen-specific support were all statistically significant predictors of lower QOL (Table II). For both models, the variable of depressive symptoms was the psychosocial variable accounting for the most variance in QOL.
Due to the unexpected result of social support not remaining in the models, we performed an analysis of the collinearity statistics of the regression models. This analysis revealed that depressive symptoms and social support had a high collinear relationship by tolerance values of 0.46 and 0.52 for depressive symptoms and 0.42 and 0.49 for social support (Interview II and III, respectively). We also ran an additional hierarchical multiple linear regression model for QOL that included social support but not depressive symptoms as a possible predictor to better understand if the variable of depressive symptoms was masking the role of social support in the overall regression model. The model accounted for 61.8% of the variance in QOL, and the psychosocial variables accounted for 16.4% of additional variance beyond demographics and epilepsy-related medical history variables. Higher levels of social support were found to be significant predictors of higher QOL.
The purpose of this study was to examine the relationships between four psychosocial variables and QOL in order to more adequately understand factors associated with QOL in adults with epilepsy. Analyses revealed that a significant portion of the variance in QOL scores was explained by depressive symptoms and stigma in the cross-sectional analysis and by depressive symptoms, stigma and regimen-specific support in the longitudinal analysis.
Results showed that higher levels of depressive symptoms were the most important predictor of lower QOL. These findings support previous studies' results in which depression was a significant factor in predicting QOL in adults with epilepsy [24, 28, 29], even though these studies used different instruments to measure depression and QOL. These results provide consistency and strengthen the evidence for this relationship between depressive symptoms and QOL. In this study, more than 30% of participants received scores on the CES-D consistent with possible clinical depression. This percentage of possible clinical depression greatly exceeds the 1–9% of clinical depression reported in the general population  and adds further support that depression and depressive symptoms are a common comorbidity of epilepsy [21–24]. Care for people with epilepsy should include an assessment of depressive symptoms as a way to treat a psychological manifestation of epilepsy and as a way to predict the level of a person's QOL.
A second finding of this study was that higher levels of perceived stigma were an important predictor of lower QOL in both the cross-sectional and the longitudinal analyses. These findings parallel those of other studies analyzing the relationship between stigma and QOL [31, 43] and provide further evidence to support the hypothesis that stigma levels are significant among people with epilepsy and are negatively correlated with QOL scores.
The findings related to regimen-specific support were mixed and opposite of the hypothesized relationship. Based on theory, it was anticipated that regimen-specific support would be positively correlated with QOL and would enter the model as an important predictor of QOL. In the cross-sectional analysis, regimen-specific support was not significantly related to QOL and therefore not included in the regression analysis. In the longitudinal analysis, regimen-specific support was negatively correlated with QOL. In the model, regimen-specific support was a strong predictor of QOL, suggesting that as regimen-specific support increased, QOL declined. Although not as originally predicted, these findings are similar to the findings of DiIorio et al.  who reported that higher levels of regimen-specific support correlated with higher anxiety levels and decreasing self-management efforts in patients with epilepsy. The authors suggested that regimen-specific support may indicate that the regimen-specific supportive role could be seen as nagging, such as someone constantly reminding a person with epilepsy to take his or her medications or worrying rather than encouraging. Thus, regimen-specific support could be a source of negative tension and stress in a relationship. Another possibility is that having regimen-specific support could indicate a more severe epilepsy condition requiring more assistance and attention, thus negatively associated with overall QOL. Further research in this area is necessary to better understand the interaction between these two variables.
The final variable included in the analysis was social support. Surprisingly, social support demonstrated a statistically significant positive relationship with QOL both cross-sectionally and longitudinally but did not enter into either regression model. The high collinearity found between depressive symptoms and social support suggests that there may be little additional variance in QOL explained by social support.
We ran the additional hierarchical multiple linear regression model for QOL that excluded depressive symptoms to determine whether or not the depressive symptoms variable was masking the predictive relationship between social support and QOL. The results suggest that social support is a significant predictor of QOL when examined independently from depressive symptoms and supports previous studies displaying a positive relationship between social support and QOL [9, 42, 43]. Studies indicate that the condition of epilepsy and associated seizures often make a person with epilepsy feel isolated or less socially connected [32, 37, 38]. Since the perceived social connectedness of a person with epilepsy is an important part of his or her life, encouraging people with epilepsy to seek social relationships and teaching them how to better cope with their condition in social settings may help to improve their QOL.
Although the component of emotional well-being of QOL may overlap with the constructs of the depressive symptoms variable, other unique constructs are measured in QOL. Similarly, social support and the social function component of QOL may overlap. Although these variables are strongly correlated, the construct of QOL seems to measure more than the subject of any potential overlap with the different variables.
Results from these analyses were also used to address the final research question of the predictability of QOL over time. Based on the correlation tests and the regression model results, evidence from this study supports the predictability of QOL from these four psychosocial variables over a 3-month time period. Combining the results from both final regression models reveal that depressive symptoms, stigma and restriction of activity in the past year are the most influential variables affecting levels of QOL, cross-sectionally and longitudinally. Highlighting their importance enables a closer examination of these variables in the future.
A purposive convenience sampling method was used to recruit participants for Project EASE by approaching patients in the clinics associated with the investigators. The study sample used in this analysis was further narrowed by whether or not the QOL scale had been completed in Interview III. Due to some of the demographic characteristics of the sample, such as the lack of ethnic diversity and large proportion of participants with at least some college education, the sample presents some possible biases. Although there were no statistically significant differences in most background variables, participants completing the QOLIE-31 at the Boston site were, on average, 5 years older than those not completing the QOLIE-31. Additionally, the sample may be biased toward adults with epilepsy who have access to and who seek medical care at specialized epilepsy or neurology clinics in large metropolitan areas in the eastern United States. Therefore, it is important to consider that the sample is not representative of all adults with epilepsy, and the results are not generalizable outside the study population. Further research with a more diverse sample population is needed to determine if similar results would emerge.
Because QOL was not measured at Interview II, prior QOL could not be statistically controlled in the analysis of predictability of the psychosocial variables (measured at Interview II) on QOL measured at Interview III. Future research must determine the extent to which the psychosocial variables measured at an earlier point in time add to understanding current QOL.
Epilepsy is a public health concern affecting the lives of those possessing this chronic condition. As the scope of understanding epilepsy and its impact continues to grow, more applied public health efforts can be achieved. Topics of future studies should include evaluating the intensity of epilepsy's impact by comparing QOL and psychosocial factors of people with epilepsy to the general population, studying QOL over time in patients with epilepsy and examining other factors that directly influence QOL or influence the psychosocial measures mediating the relationship with QOL.
An enhanced knowledge and understanding of factors influencing QOL in adult patients with epilepsy is valuable to provide more comprehensive and effective care for these individuals through proper treatment regimens, interventions and educational materials. Realizing the psychosocial aspects affecting QOL in a person with epilepsy enables public health professionals to identify and apply interventions that are both appropriate and effective. Also, educating the public about epilepsy and the lives of those who have epilepsy would dispel public misconceptions, fear and anxiety surrounding epilepsy and seizures.
Additionally, the strong predictive relationship between depression or depressive symptoms and QOL in patients with epilepsy provides a point of clinical intervention where a psychological manifestation of epilepsy may be treated. Recognizing depressive symptoms in people with epilepsy should be of great concern for health care providers and health educators and may be predictive of lower QOL in these individuals. If these factors can be identified, people with epilepsy can be referred to counseling or other types of mental health interventions can be used to give the person more comprehensive individualized care.
The original study, Project EASE, was supported by the National Institute of Nursing Research [R01-NR04770 to Project EASE study group, M01-PR01032 to Beth Israel Deaconess Medical Center-GCRC].
We acknowledge the coinvestigators on the original study: T. R. Henry, R. Letz (posthumously) and S. Clements (Emory University, Atlanta, GA site); D. L. Schomer and P. O. Shafer (Beth Israel Deaconess Medical Center, Boston, MA site) and all Project EASE study group members.
Disclaimer: The findings and conclusions expressed in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention (CDC) or the US government. The authorship and contributions of A.D.W. to this study were performed in private capacity, and no official support or endorsement by CDC is intended or should be inferred.