SEARCH

SEARCH BY CITATION

Keywords:

  • Epilepsy;
  • Mental health;
  • Quality of life;
  • Normal activity limitations;
  • Social support

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References

Although seizure control is a critical goal for persons with epilepsy (PWE) from the biomedical perspective, there is growing support for approaches that take into account the effect of epilepsy on the whole person (the biopsychosocial perspective). We hypothesized that PWE who report normal activity limitations due to epilepsy and poor social/emotional support would report poor mental health regardless of whether they report having seizures in the past 3 months or currently taking medication for their epilepsy. In 2005 and 2006, twenty-two states asked about epilepsy in the Behavioral Risk Factor Surveillance System (BRFSS). A set of survey weight-adjusted logistic regression analyses were conducted to examine self-reported poor mental health. Our findings reveal that both normal activity limitations due to epilepsy and poor social/emotional support are significant predictors of reporting poor mental health. After controlling for social support, demographics, comorbidities, and health behaviors we determined that those reporting normal activity limitations due to epilepsy are more likely to report poor mental health (odds ratio [OR] 3.05, 95% confidence interval [CI] 1.41–6.58). Similarly, after controlling for the same factors, PWE with poor social/emotional support are more likely to report poor mental health (OR 3.22, 95% CI 1.84–5.61). In contrast, reported seizures in the past 3 months (OR 1.12, 95% CI 0.56–2.25) or current use of medication (OR 0.58, 95% CI 0.30–1.10) were not significantly associated with poor mental health. These population-based findings support the need for clinical approaches that address psychosocial issues in addition to seizure control and treatment.

It was long ago suggested that epilepsy treatment go beyond purely a biomedical focus of medication and surgery to a biopsychosocial one that also integrates the physical, mental, and social aspects of health (Lennox & Markham, 1953). Although seizure freedom is an important treatment goal, evidence is accumulating that poor mental health has a greater influence than seizures on quality of life (Perrine et al., 1995; Boylan et al., 2004). Despite greater recognition by neurologists of the mental health complications in epilepsy (Blumer et al., 2004), a large proportion of persons with epilepsy (PWE) remain unscreened (Shneker et al., 2009) and untreated for psychiatric comorbidities (Wiegartz et al., 1999; Kanner et al., 2000). Based on the clinical literature, we hypothesized that PWE with normal activity limitations due to epilepsy and poor social/emotional support would report poor mental health regardless of whether they report having seizures in the past 3 months or currently taking medication for their epilepsy.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References

The Behavioral Risk Factor Surveillance System (BRFSS) is an ongoing state-based, random-digit–dialed telephone survey of noninstitutionalized persons in the United States older than the age of 18 that is designed to assess chronic disease burden and monitor health risks. BRFSS data are weighted to reflect the age, gender, and racial/ethnic distribution of the state’s estimated population during that survey year. In 2005, 19 states asked about epilepsy as part of their supplemental module using questions developed by the Centers for Disease Control and Prevention (CDC). In 2006, 11 states asked about epilepsy yielding a combined sample of 1,835 epilepsy cases. The Ohio State University Institutional Review Board deemed the research exempt.

For this study, respondents were defined as having epilepsy if they responded “yes” to ever having been told by a doctor that they had a seizure disorder or epilepsy. A recent U.S. study of self-reported epilepsy in New York City used a similar method of case ascertainment and yielded a positive predictive value for epilepsy of 81.5% (Kelvin et al., 2007). This is comparable to the positive predictive value of 84% for an epilepsy diagnosis based on ICD-9 coding used in administrative records (Jette et al., 2010).

To address the biomedical perspective, two questions were used as independent variables: seizures in the last 3 months and current medication use for epilepsy. To address the biopsychosocial perspective, we used two measures as independent variables: normal activity limitations due to epilepsy and social/emotional support. The dependent variable for this study, poor mental health, was based on an assessment of self-reported mentally unhealthy days in the past month.

Data analysis

All analyses were conducted with Stata version 9.2 (Stata Corp LP, College Station, TX, U.S.A.) using calculated variables and weights provided by the Centers for Disease Control and Prevention (CDC). Data were weighted to obtain appropriate estimates, standard errors, and confidence intervals. The 2 years of data were merged. Weights for samples from the same state were adjusted proportionally to the total weights of each year consistent with previous analyses of epilepsy data from the BRFSS. To account for state-level differences, states are included in the analysis as strata. A statistically significant difference in results was determined if p-values were <5%. Survey weight-adjusted proportions and 95% confidence intervals (CIs) are reported. Because age, gender, race/ethnicity, income disparity, educational attainment, employment status, comorbid health conditions, and health behaviors (obesity and smoking) are also associated with poor mental health in the general population we control for these factors in our analyses. A statistically significant difference in results was determined in these analyses when the 95% CIs of odds ratios do not overlap with value 1.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References

The examination by poor mental health status revealed differences in seizures in the past 3 months, normal activity limitations due to epilepsy, poor social/emotional support, and marital status (see Table 1).

Table 1.   Bivariate associations of epilepsy and support by poor mental health status— 2005–2006 Behavioral Risk Factor Surveillance System
VariablePoor mental health
YesNo
n% (95% CI)n% (95% CI)
  1. *0.01 < p < 0.05, **p < 0.01.

  2. Poor mental health is defined as ≥14 days in response to the question “Now thinking about your mental health, which includes stress, depression and problems with emotions, for how many days during the past 30 days was your mental health not good?”

Seizures in the past 3 months**    
 Yes16831.8 (23.5–41.5)26119.5 (15.9–23.6)
 No29568.2 (58.5–76.5)1,11180.5 (76.4–84.1)
Epilepsy medication    
 Yes26257.3 (46.7–67.3)73445.9 (40.6–51.2)
 No22742.7 (32.7–53.3)73754.1 (48.8–59.4)
Epilepsy interfering with normal activities**    
 Yes13434.9 (24.6–46.9)18216.3 (11.7–22.1)
 No33465.1 (53.1–75.4)1,20983.7 (77.9–88.3)
Quality of support    
 Social/emotional support**    
  Yes22052.1 (41.3–62.3)1,11177.0 (71.6–81.7)
  No26047.9 (37.4–58.7)32823.0 (18.4–28.4)
 Married**    
  Yes18543.7 (33.3–54.5)77860.8 (55.8–65.6)
  No30456.3 (45.5–66.7)68539.2 (34.5–44.2)
Quantity of support in the household    
 Number of adults    
  022327.5 (20.3–36.1)51718.5 (15.5–22.0)
  119351.6 (40.4–62.7)74457.5 (51.9–63.0)
  24215.6 (7.93–28.5)12015.3 (10.7–21.4)
  3 or more185.23 (2.69–9.94)548.66 (6.08–12.2)
 Number of children    
  033158.9 (47.3–69.5)1,02160.1 (54.7–65.2)
  17821.1 (12.3–33.7)21719.0 (15.4–23.3)
  25210.9 (6.66–17.3)15413.9 (10.0–18.8)
  3 or more309.17 (3.65–21.2)817.07 (4.96–9.98)

From a biomedical perspective, there was a statistically significant association with seizures in the past 3 months and poor mental health (OR 1.93, 95% CI 1.19–3.14) in the unadjusted regression model only (see Table 2). This association did not remain once medication use for epilepsy, normal activity limitations due to epilepsy, social support (quality and quantity), demographics, comorbid health conditions, and health behaviors were controlled for in the final model (OR 1.12, 95% CI 0.56–2.25). Reported use of medication for epilepsy was not associated with poor mental health in any of the regression models.

Table 2.   Associations between the biomedical and the biopsychosocial perspectives on self-rated poor mental health in persons with epilepsy
OutcomeBiomedical perspectiveBiopsychosocial perspective
Seizures in past 3 monthsCurrently taking medication for epilepsyEpilepsy or its treatment limited normal activitiesSocial/Emotional support
Yes vs. NoYes vs. NoYes vs. NoPoor vs. Good
nOR (95% CI)nOR (95% CI)nOR (95% CI)nOR (95% CI)
  1. Poor mental health is defined as ≥14 days in response to the question “Now thinking about your mental health, which includes stress, depression and problems with emotions, for how many days during the past 30 days was your mental health not good?”

  2. Epilepsy or its treatment limited normal activities is based on the question “During the past 30 days, to what extent has epilepsy or its treatment interfered with your normal activities like working, school, or socializing with family or friends?” The responses “moderately, quite a bit or extremely” were recoded as “yes” versus “not at all or slightly,” which were recoded as “no.”

  3. Social/emotional support is based the question “How often do you get the social and emotional support you need?” The responses “sometimes, rarely or never” were recoded as “poor” and the responses “always or usually” were recoded as “good.”

  4. Model 1 is the unadjusted models for having seizure in the past 3 months.

  5. Model 2 adjusts for having seizure in the past 3 months and currently taking medication for epilepsy.

  6. Model 3 adjusts for model 2 covariates and epilepsy or its treatment-limited normal activities, number of adults, number of children, marital status, and social/emotional support.

  7. Model 4 adjusts for model 3 covariates and demographics: age, gender, employment status, race/ethnicity, educational status, health insurance, and annual income.

  8. Model 5 adjusts for model 4 covariates and comorbid health conditions (type 2 diabetes, asthma, and heart disease) and health behaviors (obesity and smoking).

Poor mental health        
 Model 11,8351.93 (1.19–3.14)      
 Model 21,8331.77 (0.98–3.18)1,8331.26 (0.70–2.26)    
 Model 31,7111.07 (0.52–2.22)1,7111.15 (0.61–2.16)1,7111.83 (0.81–4.14)1,7112.69 (1.52–4.77)
 Model 41,3371.05 (0.55–1.99)1,3370.68 (0.38–1.24)1,3372.88 (1.36–6.06)1,3373.11 (1.86–5.19)
 Model 51,2121.12 (0.56–2.25)1,2120.58 (0.30–1.10)1,2123.05 (1.41–6.58)1,2123.22 (1.84–5.61)

The regression models examining the biopsychosocial perspective found that both normal activity limitations due to epilepsy and social/emotional support were significant predictors of poor mental health. After controlling for seizures in the past 3 months, medication use for epilepsy, social support (quality and quantity), demographics, comorbid health conditions, and health behaviors, those reporting normal activity limitations due to epilepsy were significantly more likely to report poor mental health (OR 3.05, 95% CI 1.41–6.58). Similarly, after controlling for normal activity limitations due to epilepsy, social support, comorbid health conditions, and health behaviors, those with poor social/emotional support were significantly more likely to report poor mental health (OR 3.22, 95% CI 1.84–5.61); see Table 2.

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References

For PWE, normal activity limitations due to epilepsy and poor social/emotional support strongly predict poor mental health after controlling for biomedical factors such seizures in the past 3 months and treatment with medication. Our results provide support for a biopsychosocial approach to epilepsy that treats the whole person—not just their seizures. This multifaceted approach recognizes that epilepsy and its treatment can create significant normal activity limitations, especially in the areas of work, school, or socializing with family or friends. The current study also supports the suggestion that the persons most debilitated by epilepsy are those who lack social support—not those with poorly controlled seizures (Droge et al., 1986).

Small scale clinical studies of quality of life favor the importance of poor mental health over seizures. Because these studies have been limited to clinical populations, the current study adds to this literature from a population perspective. Although the current study lacks specificity with respect to type, dosage, or number of antiepileptic drugs (AEDs) used by study participants, the results demonstrate that the association between active seizures and poor mental health is moderated by medication use for epilepsy. However, this association was reduced much further in the regression models that controlled for various biopsychosocial factors, thus providing more evidence in support of a holistic treatment approach.

Despite numerous studies that examine the neuropsychiatric components of epilepsy and the resulting psychosocial issues, most medical-based tertiary epilepsy centers in North America do not provide comprehensive intervention programs that address the psychosocial, educational, vocational, and family support needs of PWE and their families (Sherman, 2009). Although epilepsy patient associations offer some of these services to PWE and their families, the development of programs focused on appropriate screening and treatment of mental health issues may help improve quality of life in PWE.

Limitations

It is important to acknowledge limitations in the BRFSS data. The BRFSS is a cross-sectional study, and this greatly limits the ability to assess any causal or temporal relationships between epilepsy and poor mental health. Data presented in this study were from a handful of states and therefore do not represent PWE from all 50 states. The BRFSS also excludes institutionalized individuals. Self-report of epilepsy by participants is also limited, since no confirmation is made by health professionals. The case definition of epilepsy in the BRFSS does not allow for specific examination of persons with intractable epilepsy, the length of epilepsy diagnosis, seizure type, or number/type of antiepileptic drugs used. In addition, the poor mental health measure used in the BRFSS does not allow for discrimination between the various mood and anxiety disorders prevalent in PWE.

Conclusions

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References

Our findings reveal that both normal activity limitations due to epilepsy and poor social/emotional support are significant predictors of self-reported poor mental health. These associations remained significant after controlling for seizures in the past 3 months, medication use for epilepsy, demographics, comorbid health conditions, and health behaviors. These findings provide support for interventional approaches that take into account the biomedical aspects of epilepsy (seizures and medication treatment) as well as the psychosocial aspects of epilepsy such as normal activity limitations and poor social/emotional support.

Acknowledgment

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References

There was no study sponsor as part of this research.

Disclosure

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References

Bo Lu, PhD, and John O. Elliott, MPH have no disclosures to report. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. Acknowledgment
  8. Disclosure
  9. References