Introduction

People living with epilepsy experiences social consequences, cognitive impairment, mental health conditions and poor Quality of Life (QOL) [1, 2]. The QOL has been described as pragmatic end point in the management of people living with epilepsy [3, 4]. Seizure frequency, seizure severity, level of education, depression, polytherapy and adverse events profile are major determinants of QOL in Patients with Epilepsy (PWE). In addition, women with epilepsy (WWE) faces unique challenges throughout their lifespan owing to the interplay between seizures, anti-epileptic drugs (AEDs) and sex steroid hormones [5]. This tri-directional complex relationship results in the disruption of the hypothalamic pituitary ovarian axis which affects the determinants of QOL like seizure frequency, depression and adverse events profile in WWE [6, 7]. In a cross-sectional study to evaluate the use of AEDs patterns and their impacts on QOL among 200 patients, there were more cases of adverse drug reaction among those on polytherapy compared to those on monotherapy and the study concluded that selection of rational and safer AED treatment options plays a major role in achieving better QOL in patients with epilepsy [8]. However, previous studies had shown that positive effects of newer AEDs like levetiracetam (LTM) as add-on therapy on QOL but just few studies have been able to look into the actions of LTM as a sole agent on QOL of PWE [9]. Furthermore previous studies have linked AEDs especially those that interfere with cytochrome P450 to reproductive endocrine with alteration of the level of circulating gonadotropins and sex steroid hormones [5, 10]. These changes have attending effects on seizures control and frequency which are some of the major determinants of QOL especially in WWE. Previous studies have evaluated the effect of CBM as monotherapy on sex steroid hormones and its effect on QOL, but there are few data on effect of LTM as monotherapy on sex steroid hormones and attending effect on overall QOL in our environment and sub-Saharan African (SSA) by extension. This study aims to identify the predictors and compare QOL in WWE on carbamazepine (CBM) and LTM monotherapy.

Methods

This is a medical out-patient hospital-based cross-sectional study between August 2015 to August 2016. A total of 100 age-matched WWE of reproductive age group with 50 each on CBM and LTM monotherapy were randomly selected from available records in the clinic which serve as sampling frame. We excluded WWE that were currently or previously on any other AEDs apart from CBM or LTM, and those with any form of endocrinopathies, mental health issues and primary/secondary amenorrhea. The choice of CBM, was driven by the fact that it is mostly prescribed/used of the first-generation AEDs but limited by significant drug interaction and induction hepatic microsomal enzyme [11, 12]. While LTM is a newer generation drug with increasing usage, lesser drug interaction and effect on hepatic microsomal enzyme [11]. The participants on CBM were 200 mg of twice daily, while those on LTM were on 250 mg twice daily. Sample size was calculated using the formula for comparison of means average and standard deviation of Quality of Life Inventory Scale 31(QOLIE-31) among WWE from previous studies was used [13]. However, because of the non-parametric distribution of variables and the availability of small data set bootstrapping re-sampling was done. Participants for this study were fully informed on the research protocol detailing the purpose, method, risks, and benefits of the research. The risk concerning adverse reaction related to CBM and LTM, pain from blood sampling withdrawal, possible reaction to contrast during neuroimaging, and possible allergy with gel used during attachment of Electroencephalography (EEG) electrode were explained to the patients. A Phoenix digital 32-channel EEG machine made in Austria was used. Each of the participant gave a written and well understood informed consent. The consent was translated to the local language for those who did not understand English language and the services of interpreters were employed. Participants were free to decline participation or withdraw from the study at any time without reprisal or loss of benefit. After clinical diagnosis of epilepsy, seizure classification and definition of epileptiform activity on EEG were in accordance with International League Against Epilepsy (ILAE) and Clinical Neurophysiology Society’s Standardized Critical Care EEG Terminology by two different Neurologists [14,15,16,17,18]. An interviewer based pre-established questionnaire was used to obtain clinical information with regard to socio-demographic characteristics and medical history related to epilepsy of interest which include age of onset, aetiology, duration of epilepsy, frequency and types of AED used [19]. We used QOLIE-31, an instrument previously validated in clinical studies among Nigerian cohort to assess the QOL [13, 20]. The QOLIE-31 has good psychometric properties with test–retest reliability ranging from (0.64–0.85) and internal consistency reliability co-efficient ranging from α = 0.77 to α = 0.85 for social functioning scale and cognitive functioning scale, respectively [3, 4, 13]. The instrument measures QOL by conversion of preceded numeric value from raw data to a scale of 0 to 100 with higher scores indicating a better QOL and places emphasize on seven domain which includes seizure worry, emotional well-being, energy/fatigue, medication effects, cognitive functioning, social functioning and overall quality of life [13]. The Zung Self-Rating Depression Scale (ZSRDS) was used to screen for depression [21, 22]. It consist of a 20 item with Likert-type scale after each item and the score for each of item ranges from 1 to 4. The minimum possible score is 20 and 80 is the maximum possible score. A score > 50 is taken as depression [22, 23]. Hormonal sample collection, handling analysis and menstrual characteristics was as we previously described [24]. Catamenial epilepsy as a twofold increase in daily seizure frequency during specific phases of the menstrual cycle [25,26,27]. The Statistical Package for Social Science for windows version 22 acquired by IBM in 2013 and manufactured in Armonk, New York, was used for analysis of the data obtained after initial entering and cleaning on Microsoft Excel. The Shapiro–Wilk test was used to check for normality of data and the significant value was greater than 0.05 for all the test thus making the data a normal distribution. Socio-demographic characteristics was tested using Chi-square statistics while the association between the socio-demographic characteristics and QOLIE-31 score were tested using the independent Student’s t-test and one-way analysis of variance (ANOVA) with a post hoc test using Tukey's honest significance test. The Tukey's honest was used to find means of the correlates (socio-demographic and clinical variables) that are significantly different from each other. A multiple linear regression model was used to assess the independent predictors of the QOLIE-31 total score. A level of statistical significance was set at p-value of less than 0.05 for all statistical analysis.

Results

Table 1 shows that the mean age of patients was 29.07 ± 7.55, while the mean age of onset was (20.01 ± 11.57). The largest percentage of participants had secondary education and a total of 14% had depression. Furthermore, the duration of epilepsy was more than 2 years in 77% and less than 2 years in 23% among WWE. Considering, QOLIE 31 total score, the mean score was statistically significantly higher among WWE that have postgraduate education, use of LTM, epilepsy duration of < 2 years, slow background frequency on EEG and those without depression. With regard to menstrual history, both groups had comparable age of menarche and presence of catamenial epilepsy; p-value: p (0.094, 0.092), respectively. A higher number of the LTC group had hirsutism, 10 (20.0%), inter-menstrual bleeding, 9 (18.0%), dysmenorrhea 40 (80.0) and dyspareunia, 13 (26.0%) as opposed to the CZP group (see Table 2).

Table 1 relationship between socio-demographic and clinical characteristics and QOLIE-TS
Full size table
Table 2 Showing comparison of menstrual characteristics of WWE on CBM and LTM monotherapy
Full size table

Table 3 shows that the mean total score of QOLIE-31 was 49.69 ± 13.45, with the highest seen in overall QOL domain 54.56 ± 21.64 and lowest in cognitive functioning domain 46.98 ± 15.98. Higher scores was seen in LTM group with statistically significant differences across all domains except seizure worry (p = 0.051). The level of education is associated with seizure worry (p = 0.020), overall QOL (p = 0.002), energy (p = 0.032), cognitive functioning (p = 0.000) and social functioning (p = 0.001). Depression is associated with overall QOL (p = 0.009), energy (p = 0.022), cognitive functioning (p = 0.016), medication effects (p = 0.009), and social functioning (p = 0.013).

Table 3 Relationship between socio-demographic characteristics and QOLIE-31 domains
Full size table

There is an association between duration of epilepsy and overall QOL (p = 0.005), medication effects (p = 0.019) and social functioning (p = 0.015). Family history of epilepsy is associated with seizure worry (p = 0.022) and cognitive functioning (p = 0.038). There is no significant association between last episode of seizure and epileptiform pattern with domains of cognition except for emotional well-being (p = 0.005, p = 0.028, respectively). There is significant association between types of AED and domains for cognition except for seizure worry (p = 0.051). EEG frequency is associated with overall QOL (p = 0.004), emotional well-being (p = 0.001), cognitive functioning (p = 0.009) and medication effects (p = 0.023) (see Table 4).

Table 4 Relationship between seizure characteristics and QOLIE-31 domains
Full size table

Logistic regression analysis showed that the medication (p = 0.000), EEG frequency (p = 0.005), duration of epilepsy (p = 0.017), depression (p = 0.008) and level of education (p = 0.003) were significant predictors of poor QOL (see Table 5).

Table 5 Showing logistic analysis of the correlates of QOLIE-TS
Full size table

As shown in Table 6, there was a significant correlation between prolactin and all the domains of QOLIE-31 in both follicular and luteal phase. Progesterone (p = 0.040), oestradiol (p = 0.011) and prolactin (p = 0.002) in the follicular phase showed a statistically association with QOLIE-total score. While FSH (p = 0.015), prolactin (p = 0.000), LH-FSH ratio (p = 0.009) in the luteal phase and testosterone (p = 0.015) showed a significant association. However, none of the hormones independently predict the quality of life on linear regression analysis (see Table 7).

Table 6 Relationship between hormones and QOLIE-31 domains
Full size table
Table 7 Showing linear analysis of the correlates of QOLIE-TS
Full size table

Discussion

To the best of our knowledge, this is one of the very few studies aimed at identifying determinants of QOL in WWE. Identified predictors of poorer QOL in WWE from this study include medication effects, educational level, depression, background EEG frequency and duration of epilepsy. The level of education has been a consistent indicator found to be associated significantly with QOL [13, 28,29,30], thus it was not surprising that this study demonstrated that low QOLIE-31 total score was seen among people who had little formal education and higher among those with postgraduate education. Education is an important indicator that may directly or indirectly influence QOL through its association with employment, higher social class, and economic status. The role of education in ensuring medication adherence should not be underestimated [13, 30, 31]. Clearly the choice of anticonvulsants plays a significant role in the determination of the QOL in epilepsy [28]. Previous studies have demonstrated an improvement in QOL with use of LTM as add-on therapy. In this present study, LTM performed better than CBM across all domains of QOL except seizure worry. These findings from our study is similar to findings from another study by Rudakova and colleagues, on effect of current AEDs on quality of life of PWE which concluded that patients treated with LTM had higher scores than patients treated with CBM [32]. Studies in the past have evaluated impact of newer drugs on QOL in PWE using various tools such as SF-36, QOLA, QOLIE-89 and QOLIE-31 in variety of clinical settings and have consistently demonstrated an improvement in QOL [4, 29, 32]. On the other hand, unsatisfactory effect of CBM on QOL has been attributed to the unfavourable pharmacokinetic profile and adverse drug reaction [12, 32]. However, it will be difficult to ascribe causality and temporal relationship to sole effect of medications because the cross-sectional nature of the study design. A prospective randomised control trial or longitudinal study will be needed to deduce such causal relationship between LTM/CBM and QOL. It is rather surprising that we are unable to demonstrate any association between seizure type and QOL as it was previously demonstrated though with contrasting finding from different studies, while Herodes and colleagues reported lower scores in patients with generalized tonic–clonic seizures, Thomas and colleagues among Indian and Guekht and colleagues in Russian patients demonstrated that focal seizures had lower QOL scores than those with generalized seizures [33, 34]. A plausible reason for this difference are varied duration of seizures and gender specificity in our cohort. Another predictor of QOL identified in this study was the presence of fast frequency background on EEG. The EEG have proven to be a reliable biomarker that helps in detecting cortical abnormalities associated with cognitive decline, an important component of QOLIE-31 [35] . There was a significant association between hormones and QOLIE-31 total score, but this did not attain any significant level on regression model. We thus suggest further evaluation of this potential therapeutic implication of this finding using a larger sample size in longitudinal study. The interplay between the sex steroid hormonal, epilepsy, and AED is complex. Both interictal and ictal discharges have been proposed as altering the sex steroid hormonal axis at the level of the hypothalamus and the pituitary [26, 27]. There is large variability in reported cases due to differing definitions, however, work by Herzog and colleagues has led to a more uniform acceptance of the definition of CE as a twofold increase in daily seizure frequency during specific phases of the menstrual cycle [25,26,27]. This study reported a CE frequency of 6% with no significant difference in both groups. Progesterone has long been shown in several studies to have anti-seizure activities [5]. WWE are prone to seizure in response to decrease level of progesterone especially during premenstrual period [5]. The main anti-seizure effect of progesterone has been linked to its conversion to all pregnanolone, an intermediate precursor, which has positive modulatory effect on GABA-A receptors [37]. Also, progesterone may modulate signalling cascades of inflammation, apoptosis, neurogenesis, and synaptic plasticity and therefore, progesterone may directly exact disease-modifying effects on epileptogenesis [27, 37, 38]. Thus, enhancing the activity of progesterone or its derivatives might be an unexplored therapeutic end point in reducing seizures and improving QOLIE-31 in WWE. Oestrogen is generally believed to have proconvulsant and epileptogenic property in animals and humans [10]. However, some other studies have shown that the effect of oestrogen on seizure susceptibility is highly variable depending on the factors such as treatment duration, dosage, hormonal status and seizure model [7]. Harden and colleagues reported a positive correlation between seizure and the oestrogen–progesterone ratio in WWE picking in the premenstrual and pre-ovulatory period and declining during the mid-luteal phase [5]. The effect of testosterone on seizure is complex and metabolites dependent, while the aromatization metabolite 17beta oestradiol has proconvulsant effect, testosterone on other hand is converted by 5 alpha reductase to 5 alpha dihydrotestosterone and then subsequently to 3 alpha androstenediol which has anticonvulsant property through its GABA modulating effect [40]. Findings from our study was also in keeping with previous findings that depression is a significant predictor of QOLIE and a common co-morbidity among people with epilepsy [29, 31, 41]. In adults, depression and anxiety are the two most frequent mental health-related diagnoses in epilepsy [42, 43]. Depression in people with epilepsy though common sometimes escapes diagnosis even when necessary treatment is needed [44]. In many cases, a combination of anti-epileptic use, psychotherapy and antidepressants are the most effective approach [3]. The manifestation of depression in people with epilepsy has a significant negative toll on seizure control, sexual function, sleep, school or work performance, cognition and consequently on overall QOL. Our study is limited in that we restricted our cohort to females. We were able to identify determinants of QOL in WWE, but causality and relationship could not be ascertained because of the cross-sectional nature of the study especially with regard to medication effect. As such, we propose a prospective longitudinal study to further explore interplay among AEDs, hormones and QOL in WWE.

Conclusion

Overall, we demonstrated higher QOLIE-31 score in the LTM group across all domains except seizure worry. Furthermore, we identified predictors of poorer QOL in WWE from this study include medication effects, educational level, depression, background EEG frequency and duration of epilepsy. While we demonstrated association between progesterone, testosterone, prolactin and LH/FSH ratio with QOL, none of the hormones independently predicts the QOL on linear regression analysis.