Illness Perceptions Mediate the Relationship Between Depression and Quality of Life in Epilepsy Patients

Summary

The current study examined whether negative illness perceptions help explain the link between depression and quality of life. Seventy patients with epilepsy completed standardized self-report questionnaires measuring depression, illness perception, and quality of life (QOL). Illness perception statistically mediated the relationship between depression and QOL [Indirect effect = -.72, LL = -1.7, UL = -.22, P<.05]. Results held with and without adjusting for potential confounding variables (age, sex, ethnicity, income, and seizure frequency) and when operationalizing depression as a continuous variable that indexed severity of symptoms or as a dichotomous variable that indexed criteria consistent with a diagnosis of major depressive disorder. This study is the first to suggest that illness perceptions may be a useful target in screening and intervention approaches in order to improve QOL among low-income, racially/ethnically diverse patients with epilepsy.

Epilepsy is the most common chronic and serious brain disorder worldwide.¹ The prevalence of depression in persons with epilepsy (PWE) is twice that of the general population.² Depression is also a major cause of premature mortality and a primary predictor of impaired QOL.³ Although the association between epilepsy and depression may be bidirectional, depression frequently precedes a diagnosis of epilepsy⁴ and contributes to greater impairment in QOL than seizures.⁵ Despite the prevalence and prognostic significance of depression in PWE, very little is known about the factors that explain how depression leads to important clinical outcomes like diminished QOL. Medication adherence and seizure activity have been examined as potential explanatory factors but have not been supported as mechanisms underlying the relationship between depression and QOL.⁶ Low socioeconomic status and low social support are each associated with depression and with poorer QOL in epilepsy patients.⁷ Yet, no evidence exists that these variables explain the link between depression and impaired QOL in PWE. Additionally, the clinical significance of these relationships is limited given that these variables are not easily targeted by therapeutic interventions. It is imperative to examine candidate mediators such as psychological factors to identify novel intervention targets to improve QOL.

Preliminary data in PWE and substantial evidence across a range of disease conditions indicate that illness perception, patients' beliefs about their illness, is highly correlated with depression and is a modifiable factor that independently predicts QOL.⁸ These findings suggest that negative illness perception, which is especially high in PWE due to the stigma, chronicity, and complexity of epilepsy, may be an important link between depression and QOL.⁹ Also, racial/ethnic minorities may have more negative illness perceptions due to greater concerns about understanding their diagnosis and interpreting their symptoms,¹⁰ which may help explain lower QOL among these groups. Characterizing the role of illness perception as a potential mediator is critical to determining whether this modifiable factor is a viable target for interventions aimed to improve QOL in PWE, particularly those from racial/ethnic minority groups.

The aim of this study was to test whether illness perceptions statistically mediate the relationship between depression and QOL in a diverse patient sample with epilepsy.

Methods

Study design and patient selection

We enrolled 70 low-income and ethnically diverse English and Spanish speaking patients over the age of 18 from NYU-affiliated neurology and epilepsy clinics at Bellevue Hospital. A bilingual (English/Spanish) member of the research team approached all patients identified by the Bellevue Epilepsy Clinic Director as eligible candidates based on medical chart review. Not all such patients could be contacted during clinic visits due to scheduling. Less than 5% of patients who were asked to participate declined.

Patients were recruited between August 2013 and June 2014 and had either focal or generalized epilepsy and at least one seizure within the past six months. Diagnosis of epilepsy was confirmed via review of medical records and/or presence of epileptiform activity on an electroencephalogram. Patients who had known or suspected psychogenic non-epileptic events were excluded. The Rapid Estimate of Adult Literacy in Medicine- short form (REALM-SF) was used to ensure the patients were able to read the questionnaires. Patients who scored > 4 on the REALM-SF were included in the study. The study was approved by the NYU Langone Medical Center and Bellevue Hospital institutional review boards. After obtaining written informed consent, patients answered questions about their epilepsy management and treatment and completed a series of validated self-reported assessments.

Measures

Depressive symptoms were measured with the 6-item Neurological Disorders Depression Inventory for Epilepsy (NDDI-E).¹¹ A cut-off score of > 15 was used to identify patients with symptoms consistent with a diagnosis of major depressive disorder (MDD). Patients' cognitive and emotional representations of their illness was assessed with the 8-item Brief Illness Perception Questionnaire (BIPQ), a validated and reliable measure of illness representations in PWE.¹² The BIPQ measures perceived consequences, timeline, personal control, treatment control, identity, coherence (i.e., understanding of their illness), concern, and emotional representations of their illness. Higher scores reflect more negative or threatening illness perceptions. Quality of Life was measured with the Quality of Life in Epilepsy-Patient-Weighted (QOLIE-31-P).¹³ Higher scores reflect better QOL.

Statistical analyses

Mediation analysis was conducted using the PROCESS module in SPSS 21, designed to assess direct and indirect effects in mediation models. A bias-corrected 95% bootstrap-confidence interval (10,000 iterations) was used to determine statistical significance of the mediator (i.e., the indirect effect of depression on QOL through illness perceptions). Age, sex, ethnicity, income, and seizure frequency were included as control variables.

Results

Descriptive statistics

Forty percent of participants reported elevated depressive symptoms consistent with a diagnosis of MDD. Table 1 summarizes sample descriptive statistics for the entire sample, as well as for patients meeting criteria consistent with a diagnosis of MDD vs. non-MDD.

Table 1. Demographic and clinical characteristics of study population.

Characteristics	Entire Sample	Depresseda (N=28)	Not Depressed (N=42)	P valueb
Age (in years), mean ± SD	38.4 (10.9)	40.1 (10.2)	37.3 (11.4)	.31
Gender
Female	34 (47.9)	21 (50)	12 (46.4)	.81
Race/Ethnicity				.07
Non-Hispanic, White	14 (20.3)	9 (32.1)	5 (11.9)
Hispanic	38 (55.1)	11 (39.2)	26 (61.9)
Black/African-American	8 (11.6)	5 (17.8)	3 (7.1)
Asian	7 (10.1)	1 (3.5)	6 (14.2)
Other/unknown	2 (2.9)	2 (7.1)	2 (4.7)
Marital status				.80
Single (never married, widowed, or divorced)	44 (62.0)	17 (60.7)	27 (64.3)
Married/living with partner	27 (38.0)	11 (39.3)	15 (35.7)
Years of education				.20
Less than high school	16 (22.5)	6 (21.4)	9 (21.4)
High school or GEDc	22 (31.0)	6 (21.4)	16 (38.1)
Some college/tech school certificate	11(15.5)	4 (14.3)	7 (16.7)
College or graduate degree	22 (31.0)	12 (42.9)	10 (23.8)
Professional or graduate school
Family income per yeard
≤ $40,000	43 (82.7)	18 (81.8)	24 (82.8)	1
Seizure frequency				.45
Less than once per month	16 (25.4)	8 (33.3)	8 (20.5)
More than once per month	32 (50.8)	10 (35.7)	22 (56.4)
More than once per week	15 (23.8)	6 (25.0)	9 (23.1)
Illness perceptions, mean ± SD	48.2 (12.3)	51.7 (10.2)	45.9 (13.2)	.08
Consequences	7.1 (2.8)	7.4 (2.7)	6.9 (2.9)	.50
Timeline	8.1 (2.9)	8.21 (2.9)	7.93 (3.0)	.69
Personal control	5.6 (3.2)	6.3 (3.2)	5.0 (3.2)	.11
Treatment control	2.8 (2.7)	3.3 (3.2)	2.3 (2.2)	.17
Identity	6.1 (2.8)	6.3 (2.7)	5.9 (2.9)	.62
Concern	8.7 (2.2)	8.1 (2.5)	9.0 (1.9)	.18
Understanding	3.3 (2.9)	4.1 (3.3)	2.8 (2.6)	.07
Emotional response	6.6 (3.3)	7.4 (3.0)	6.0 (3.4)	.08
Quality of Life	59.2 (14.9)	54.3 (14.9)	62.7 (14.1)	.02

Note. Data are presented as number (percentage) unless otherwise indicated.

^aNDDI-E scores > 15=symptom criteria consistent with diagnosis of MDD.

^bFisher's exact test for proportions and Mann–Whitney U-test for continuous and ordinal variables; statistical tests are between depressed and non-depressed groups.

^cGeneral Education Development credential.

^dTotal N=52 due to missing data for income variable.

Hypothesis testing

Figure 1 shows the path model and associated standardized beta weights for the relationships between depressive symptoms, illness perception, and QOL. Greater depressive symptoms were associated with more negative illness perceptions and poorer QOL. The direct relationship between depression and QOL was no longer significant when illness perception was included in the model. The bias-corrected bootstrap confidence-interval that does not include zero indicates that illness perception statistically mediated the relationship between depressive symptoms and QOL [Indirect effect = -.72, LL = -1.7, UL = -.22, P<.05].

Figure 1.

*P<.05. The relationship between depression and quality of life mediated by illness perception controlling for age, sex, ethnicity, income, and seizure frequency. Numbers represent standardized betas; parenthesized numbers represent betas when predictors were entered into regression model simultaneously. The variance of QOL explained by this model is 43%, R² = 0.43, P<0.01.

Discussion

This study tested the hypothesis that negative illness perception may be an important link between depression and QOL in a diverse sample of PWE. Results indicated that illness perception statistically mediated the relationship between depression and QOL. This is a novel finding in patients with epilepsy and parallels evidence from other studies demonstrating that illness perceptions are an important and modifiable link between depression and QOL.⁹ Notably, results remained robust when controlling for potentially confounding variables (e.g., age, sex, ethnicity, income, and seizure frequency). Additionally, the magnitude and significance of results were unchanged when examining depression dichotomized to reflect diagnostic criteria consistent with MDD (vs. non-MDD). This supports that findings are generalizable to individuals who meet clinical criteria for MDD.

Since illness perceptions are modifiable, these results point to important considerations relevant to screening and intervention approaches that aim to reduce suffering in epilepsy patients with depression who are at greatest risk for impaired QOL. For example, cognitive-behavioral and acceptance-based interventions that target negative cognitive evaluations may be especially beneficial for depressed patients with elevated scores on the illness perception questionnaire. Although the vast majority of psychotherapeutic interventions for PWE have not examined active ingredients of such treatments, one study found that decreases in negative epilepsy-related cognitions mediated pre-post changes in QOL following treatment with Acceptance and Commitment Therapy.¹⁴ This preliminary evidence supports the validity of our findings and additionally underscores the importance of interventions that may break the link between depression and QOL by reducing negative illness representations. Results also inform the potential utility of targeting elevated illness perceptions prophylactically in individuals with elevated depressive symptoms who are at risk for developing major depressive disorder. In addition to informing the development and refinement of clinical interventions, the statistical mediation results from this study contribute to the empirical base in support of neurocognitive and psychotherapeutic models to be further tested. Such models may generalize to additional outcomes and other disease conditions.

Results also contribute to the growing evidence that psychosocial factors– not clinical variables (i.e., seizure frequency) – have the greatest impact on quality of life in PWE. Overall, findings converge with other investigations demonstrating that a psychological model explains significantly greater variance in QOL compared with a social and biological-biomedical model alone.¹⁵

Limitations And Future Directions

Several limitations of the current study merit further investigation. First, as a cross-sectional study, causal conclusions about the relationship between depression and QOL and the underlying mechanisms cannot be drawn. Additionally, our small sample size precluded the use of superior techniques like structural equation modeling that account for measurement error and provide model fit indices that inform the plausibility of the causality assumptions in our hypothesized mediational model. The present results support patterns of relationships between depression, illness perceptions, and QOL and demonstrate statistical meditation only. However, our data converge with evidence from prospective and experimental designs that support causal inferences.^5,14 Further, although some theoretical and empirical considerations indicate that the relationship between depression and illness perceptions may be bi-directional, an alternative model whereby illness perceptions lead to diminished QOL through depression was not empirically supported by our data because the direct relationship between illness perception and QOL remained highly significant after controlling for depression. Overall, while our meditational results must be interpreted with caution, the pattern of relationships presented here, together with converging evidence from longitudinal and experimental studies, offers novel, meaningful, and theoretically supported evidence for the relationships between depression, illness perceptions, and QOL in patients with epilepsy. ^5,14 The present research lays the groundwork for future investigators to safely invest in longitudinal studies that can address causal relationships between these variables. Another limitation is that our sample was predominately low-income and results may not generalize to PWE with higher incomes. Finally, due to limited variability within some of the subscales of the IPQ, this study reports mediation results using a total score from the IPQ, rather than subscales. Future studies with larger sample sizes are needed to test specific subcategories of illness representations as mediators to determine which aspects of illness perception should be targets for intervention.

Conclusion

The prevalence of depression and its causal role in impaired QOL in PWE warrants concern and scientific investigation to identify the mechanism underlying this association. This study is the first to indicate that negative illness perception helps to explain the link between depression and QOL in PWE. Results highlight that illness perception may be an important target for screening and interventions aimed at improving QOL in PWE. Findings also contribute to the growing evidence in support of a psychotherapeutic model to reduce suffering in patients with comorbid depression and a range of chronic diseases.