Factors associated with impaired quality of life in younger and older adults with epilepsy

Introduction 

The elderly represent the fastest growing segment of our population. Incidentally, this group also has the highest incidence of acute symptomatic seizures and epilepsy (Stephen and Brodie, 2000, Devinsky, 2005, Cloyd et al., 2006). Etiology, clinical presentation, and management of epilepsy in the elderly are distinct. Nearly half of all elderly epilepsy is cryptogenic, but among the known causes stroke, tumor, and dementia constitute a much higher fraction than at other ages. Diagnosis can be particularly challenging given the high frequency of comorbid disorders, social isolation, underrecognition of complex partial seizures, and prolonged post-ictal confusion that often mimics dementia. Additionally, medication compliance is an enormous challenge due to a particular vulnerability to side effects, memory impairment, among other factors. Moreover, the elderly are particularly prone to seizure sequelae, including physical injury, loss of confidence, and reduced independence (Devinsky, 2005, Cloyd et al., 2006).

Although epilepsy is characterized by episodic attacks, it exerts a significant psychological burden on the patients, their families, and on society in general. Furthermore, epilepsy is associated with high rates of psychiatric comorbidity, which additionally affect the quality of life (QoL) even in the absence of active seizures (Kanner, 2006, Pulsipher et al., 2006, Schachter, 2006). Over the past decade, several studies have focused on how epilepsy and its treatment affect QoL in adolescents, young adults, and middle aged patients, but only few have examined the impact of epilepsy in older adults (Martin et al., 2003, Devinsky, 2005). Martin et al. (2005), for instance, assessed concerns of living with recurrent seizures in 33 seniors and recognized a substantial burden of epilepsy in those over the age of 60. Recently, Laccheo et al. (2008) used both seizure-specific and generic QoL instruments to assess health-related QoL in 23 older adult epilepsy patients, and compared the results to those of earlier studies addressing epilepsy in the general population. The authors suggest that aging-related clinical and psychosocial variables (e.g., cognitive impairment, change in expectation), comorbid conditions (e.g., mood disorders), and epilepsy-induced complications (e.g., medication side effects) may uniquely affect health-related QoL in older adults with epilepsy. Despite the recognition of the need for QoL assessment in epilepsy patients, only limited information is available at this time to help address these particular concerns in older adults. A clear understanding of QoL issues is critical to optimal care of epilepsy.

Previous studies looking at the impact of epilepsy on the lives of elderly people have reported different findings as to how age may influence QoL. A recent study in the United States (Pugh et al., 2005) reported important differences in patients’ perception of their health status by age and epilepsy chronicity, while findings from research conducted in the United Kingdom (Baker et al., 2001) did not support the notion of an increased impact of epilepsy in older age, although patients first diagnosed late in life were found to have a comparatively more impaired QoL. These studies assessed the impact of epilepsy on health status in relation to psychological, psychosocial, and seizure-related variables. However, factors contributing to QoL ratings, specifically in older adults with epilepsy, remain largely unexplored. The purpose of this study was to use an epilepsy-specific instrument to identify and weigh demographic and clinical factors that predict QoL in older adults with epilepsy compared to younger adult patients.

Results 

Table 1 shows the demographic and clinical characteristics of the total sample, and the younger and older adult groups. Of a total of 146 patients with localization-related epilepsy, 114 patients (mean age: 46.9±17.8 S.D.) without mental retardation or dementia participated in this study. Forty-five patients aged 60 or older (66.5±5.7 years; range: 60–80) made up the older adult group, while the younger adult group consisted of 69 patients (34.2±9.3 years; range: 18–57). Sixteen patients (35.5%) in the older adult group had new-onset seizures. About one-third of patients (36.8%) reported no seizures in the preceding year (older: 40%; younger: 34.8%). Medical comorbidity was reported in 18.4% of the patients, and it was significantly more frequent among the older adults than among younger patients. The primary etiology for the symptomatic cases in the older adult group was cerebrovascular disease (older: 17.8%; younger: 4.3%, P<0.05), followed by head trauma (older: 13.3%; younger: 4.3%; P=0.17). Hypertension was reported more frequently among the older adults (13.3% vs. 1.4% in younger adults; P<0.05). Other common medical conditions reported in this sample, either alone or in combination, were asthma (3.5%), diabetes (2.6%), hypercholesterolemia (2.6%), and arthritis (1.8%). The QOLIE-31 total score was 62.7±20.6 in the total sample, and was found to be similar across patient groups (Table 2). Similarly, there was no statistical difference in terms of QOLIE-31 subscale scores, except for the medication effects subscale, where older adults scored significantly higher than younger patients. No significant difference was observed in BDI-II scores across patient groups. Although at the time of recruitment, only 2 patients (4.4%) in the older adult group and 10 (14.5%) in their younger peers had an established diagnosis of depression, BDI-II scores indicated depressive symptoms in 17 (37.7%) and 34 (49.3%) patients in the older and younger adult group, respectively. Four young adults out of the 12 patients suffering from clinical depression had been diagnosed before the onset of seizures.

Table 1. Demographic and clinical characteristics of 114 epilepsy patientsa.

		Total sample	Younger adults	Older adults
		N=114	N=69	N=45
	Age, mean±S.D.	46.9±17.8	34.2±9.3	66.5±5.7***
	Sex: female, N (%)	49 (43.4)	32 (47.1)	17 (37.8)
	Education, mean±S.D.	13.4±2.5	13.9±2.3	12.7±2.6*
	Age at onset, mean±S.D.	26.1±19.7	16.6±11.2	41.6±21.1***
	Epilepsy duration, mean±S.D.	20.7±15.2	17.6±11.7	25.5±18.6*
	Number antiepileptics, mean±S.D.	1.7±0.9	1.9±0.8	1.5±0.9
	Medical comorbidity, N (%)	21 (18.4)	6 (8.7)	15 (33.3)**
	Generalized T-C seizures, N (%)	63 (55.3)	39 (56.5)	24 (53.3)
	Epilepsy type: TLE, N (%)	69 (60.5)	47 (68.1)	22 (48.9)
	Seizure frequency, mean±S.D.	14.0±27.1	16.2±29.7	10.6±22.2
	BDI-II score, mean±S.D.	15.2±13.8	16.7±14.1	13.0±11.4
	BDI-II depression, N (%)	51 (44.7)	34 (49.2)	17 (37.7)
	Mild (14–19)	15 (13.1)	9 (13.0)	6 (13.3)
	Moderate (20–28)	18 (15.8)	13 (18.8)	5 (11.1)
	Severe (29–63)	18 (15.8)	12 (17.4)	6 (13.3)

BDI-II, Beck Depression Inventory-II; TLE, temporal lobe epilepsy.

a Statistical differences are calculated between the younger and older adult patient groups. * P value<0.05. ** P value<0.01. *** P value<0.001.

Table 2. Scores on the QOLIE-31 Inventorya.

		Total sample	Younger adults	Older adults
		N=114	N=69	N=45
	QOLIE-31 total score	62.7±20.6	61.6±21.0	64.8±19.6
	Seizure worry	55.5±29.5	53.7±28.7	58.2±30.8
	Overall QoL	56.1±23.9	56.5±24.6	55.4±23.1
	Emotional well-being	61.7±25.7	60.7±25.1	63.1±26.8
	Energy/fatigue	58.5±24.3	58.2±23.8	58.9±25.3
	Cognitive functioning	64.7±23.7	64.0±24.4	65.7±22.8
	Medication effects	59.6±27.4	54.1±23.6	67.9±30.7*
	Social functioning	70.8±26.9	68.3±27.3	74.7±26.2

QOLIE-31, Quality of Life in Epilepsy Inventory for Adults-31; values represent mean±S.D.

a Statistical differences are calculated between the younger and older adult patient groups. * P value<0.05.

Linear regression analyses for the total sample revealed that severity of depressive symptoms, as indicated by the BDI-II score, as well as seizure frequency were significant predictors of the QOLIE-31 total score. This regression model (F=13.7; d.f.=11,102; P<0.001) accounted for 57% of the variance in the QOLIE-31 total score, with BDI-II explaining 32% of the variance, and seizure frequency only 3%, as reflected by the squared semipartial correlation. BDI-II also uniquely contributed to the variance in all QOLIE-31 subscale scores, and this contribution was substantive, except for the medication domain, where only 4.7% of the score variance was uniquely explained by the BDI-II. Seizure frequency contributed to the variance in three subscale scores in the total sample, namely seizure worry, cognitive functioning, and social functioning, although in all cases less than 10% of the total variance was explained. Type of epilepsy was a predictor of emotional well-being and energy/fatigue domains. On these subscales, TLE was associated with lower scores, but the contribution to the total variance in each score was less than 5%. Female gender and number of antiepileptic drugs were positively related to the medication effects and the social functioning domain, respectively, although the effect of these predictors in terms of variance explained was weak (<5%). The predictors of the QOLIE-31 total score and subscale scores in older and younger adults are given in Table 3.

Table 3. Multivariate regression analyses for QOLIE-31 total score and subscale scores in younger and older adult epilepsy patients.

	Significant predictors	Younger adults			Older adults
		N=69 (models d.f.=10,58)			N=45 (models d.f.=10,34)
		Adj. R2/F	β	SR2	Adj. R2/F	β	SR2
	QOLIE-31 total score	0.54/8.98***			0.62/8.06***
	BDI-II		−0.63***	0.299		−0.69***	0.292
	Seizure frequency		−0.20*	0.034		−0.33**	0.077
	Seizure worry	0.22/2.97**			0.47/4.94***
	BDI-II		−0.34**	0.101		−0.65***	0.235
	Seizure frequency		−0.32*	0.080		−0.35**	0.088
	Overall QoL	0.41/5.89***			0.51/5.54***
	BDI-II		−0.53***	0.220		−0.35**	0.089
	Epilepsy duration		0.22	0.019		0.47**	0.070
	Emotional well-being	0.53/8.74***			0.40/3.90**
	BDI-II		−0.58***	0.255		−0.50**	0.156
	Epilepsy type (TLE)		−0.32**	0.091		0.03	0.001
	Energy/fatigue	0.35/4.77***			0.46/4.59**
	BDI-II		−0.52***	0.199		−0.52**	0.163
	Epilepsy type (TLE)		−0.27*	0.065		−0.04	0.001
	Cognitive functioning	0.39/5.37***			0.41/4.09**
	BDI-II		−0.49***	0.177		−0.70***	0.301
	Medication effects	0.21/2.81**			0.22/2.26*
	BDI-II		−0.21	0.045		−0.47**	0.137
	Sex (female)		−0.38**	0.119		0.09	0.005
	Social functioning	0.48/7.27***			0.42/4.13**
	BDI-II		−0.64***	0.331		−0.69***	0.288
	Seizure frequency		−0.15	0.017		−0.36**	0.092

QOLIE-31, Quality of Life in Epilepsy Inventory for Adults-31; BDI-II, Beck Depression Inventory-II; TLE, temporal lobe epilepsy; β, standardized beta regression coefficient; SR2, squared semipartial correlation.

* P value<0.05. ** P value<0.01. *** P value<0.001.

Discussion 

In the present study, the QOLIE-31 was used to determine the relative contribution of clinical and demographic factors to QoL in older and younger adults with epilepsy. Our findings revealed that severity of depressive symptoms, as indicated by the BDI-II score, as well as seizure frequency were predictors of QOLIE-31 total score in both the younger and older adult patients. However, in both cases, the impact of BDI-II outweighed more than 3-fold that of seizure frequency, which contributed modestly to the variance in QOLIE-31 total score. BDI-II contributed substantively to the variance in most subscale scores, being the sole predictor for several domains among the older adults. In addition, epilepsy duration was positively related to the overall QoL subscale only among older adults, whereas TLE diagnosis was associated with lower levels of emotional well-being and energy/fatigue among younger patients. Female gender emerged as a strong predictor for the medication domain among younger adults, although not to the level as BDI-II generally predicted QoL in both younger and older adult epilepsy patients.

The QOLIE-31 total score in our sample (62.7±20.6) was comparable to QoL scores reported in other studies in epilepsy patients, including the study validating the QOLIE-31 questionnaire (Cramer et al., 1998, Boylan et al., 2004, Meldolesi et al., 2006, Laccheo et al., 2008). The BDI-II score in this work was also similar to that of previous analyses of epilepsy (Boylan et al., 2004, Loring et al., 2004). In addition, there was no significant difference in the QOLIE-31 total score and the majority of the subscale scores across patient groups, indicating that older adults did not experience poorer QoL than younger adults. This finding, which is consistent with the results of a recent community-based study conducted in the United Kingdom, does not support the notion of an increased impact of epilepsy in old age per se (Baker et al., 2001). Laccheo et al. (2008) reported that, while older adults with epilepsy were more likely to have limited physical reserve and energy retention, as indicated by scores on a generic QoL instrument, their emotional and social well-being were at par with that of younger adult patients. These data suggest that aging does not exert a negative effect on overall QoL in the epilepsy population.

Summary 

Some limitations are acknowledged in our study including its cross-sectional design, which does not allow demonstration of causal relationship between predictors and QoL measures. Further to be considered are the relatively small sample size and the application of subjective self-report instruments. Nevertheless, both the QOLIE-31 and BDI-II have good psychometric properties and are widely used to measure QoL and depressive symptoms, respectively. It is important to point out, however, that the BDI-II and QOLIE-31 have shared items regarding mood ratings that likely influence association between the two instruments. In addition, previous studies have demonstrated interaction between seizure control and the levels of depressive symptoms (Tracy et al., 2007). Then, the relation of seizure frequency to QoL scores may be somewhat negligible when mood is taken into account. Because the QOLIE-31 as well as other current QoL instruments in epilepsy were developed and tested almost exclusively in adults under age 65 (Devinsky, 2005), we cannot ensure that QOLIE-31 is as sensitive to QoL issues in older adults as it is in younger adults or middle aged patients, and hence that the similarity in QoL scores between younger and older adults in this study is partly due to limitations related to the sensitivity of the questionnaire. The development of health-related QoL instruments specifically for older adults may be necessary to confirm these findings. Moreover, despite significant difference in mean age between younger and older adults in the present study, this sample did not include very old patients (>80 years), who may have particular QoL concerns. Further studies of QoL at very old or extreme ages, which pose unique challenges, are awaited to clarify the main QoL issues in the elderly population.

It is noteworthy that the present study in neuropsychiatry clinics involved an unselected population of patients with epilepsy, and not a volume of epilepsy patients referred for the presence of particular comorbid psychiatric conditions. Therefore, we assume that this sample is representative of the general epilepsy population, and that the data do not show a bias toward high rates of depression. A reflection of this fact is that BDI-II score in this study was comparable to that of previous investigations of patients with frequent seizures in Epilepsy Centers (Boylan et al., 2004, Loring et al., 2004, Griffith et al., 2005); and in addition, the percentage of patients with depressive symptoms (44.7%) was similar to that of earlier studies using the BDI in general populations of epilepsy patients (Beghi et al., 2002, Grabowska-Grzyb et al., 2006).

In summary, our data suggest that older adults with epilepsy do not experience poorer QoL than younger adult patients, as assessed by a seizure-specific QoL instrument, the QOLIE-31. Furthermore, we provide evidence that depressive symptoms are the strongest predictor of QoL, having greater effect than seizure frequency in both younger and older adults, even when substantial differences are known to exist in epilepsy clinical features and psychosocial factors between younger and older individuals (Devinsky, 2005, Cloyd et al., 2006). Quality of life, as measured by the QOLIE-31, is hence more determined by psychological than by seizure-related factors, regardless of age. We also found that the duration of epilepsy correlates positively with overall QoL only among older adults, suggesting that long-term adaptation to changes in health and life in general may enable better acceptance of the condition. Depression is the commonest psychiatric comorbidity in epilepsy and its persistence leads to poor QoL, increased risk of suicide, and greater use of health services (Boylan et al., 2004, Kanner, 2006). It is therefore imperative to screen for depression, and timely diagnose and treat this condition, not only in younger adults, but also in the elderly, who are not only the fastest growing segment of our population, but also a subgroup heavily underdiagnosed in terms of depressive symptoms and clinical depression.