Resilience and psychosocial factors in adult with epilepsy: A longitudinal study

Purpose: Resilience is conceptually characterized as a dynamic process encompassing positive adaptation in the context of significant adversity. Our goal was to assess the resilience in people with epilepsy (PWE) and how it impacts longitudinally on psychosocial factors, with a particular focus on the manifestation of stigmatization-related feelings. Methods: We consecutively enrolled 78 adults PWE (42.5 ± 16.2 years old); among them 36 (46.1 %) were seizure-free. All subjects completed at baseline (T0) the Resilience Scale (RS-14) and questionnaires for the assessment of depressive symptoms, anxiety and quality of life: respectively, Beck Depression Inventory-II (BD-II), Generalized Anxiety Disorder-7 (GAD-7) and QOLIE-31 (Q31). All patients were followed up prospectively and re-evaluated after 6 – 22 months (T1; mean: 14 ± 8 months; median 14 months); at follow up they also completed the Stigma Scale of Epilepsy (SSE) for the assessment of the stigma associated with epilepsy. We correlated resilience values with all psychosocial scores at T0 and T1. Factors associated with resilient and vulnerable outcomes were identified. Finally, a multiple stepwise regression analysis was applied to identify predictors for resilience and stigma perception. Results: The results showed for the RS-14 score a significant direct correlation with the Q31 (p < 0.001) and an inverse correlation with the depressive and anxiety symptoms at both times (T0 and T1), as evaluated with BDI-II (p < 0.001) and GAD-7 (p < 0.001). Finally, we found a significant inverse correlation between RS-14 at T0 and the levels of stigmatization assessed with SSE at T1 (p = .015). Using a multiple stepwise regression analysis separately for resilience and stigma perception, depressive symptoms turned out as the best predictors for both variables. Finally, considering longitudinal evaluation we did not observe significant changes in depressive and anxious symptoms, despite a significant reduction in the total number of seizures at follow up. Conclusions: Our study showed that depressive symptoms, anxiety and quality of life were significantly associated with resilience in PwE. Finally, as a novel finding resilience was proved to affect the perception of stigma related to epilepsy more than seizures.


Introduction
Epilepsy is a common neurological disorder characterized by recurrent seizures, that significantly impacts on every-day life of patients for several reasons: unpredictability of seizures, the possible interactions and side effects of anti-seizure medications (ASMs), and the social burden of epilepsy, especially stigma [1,2].Seizures occur without warning, and this uncertainty may worsen anxiety and stress.
Consistently, people with epilepsy (PWE) manifest increased prevalence of both depressive and anxious symptoms compared with the general population [3].In addition, it has been well documented that in PWE the depressive symptoms and difficulties in emotion regulations are significantly associated with worse quality of life (QoL) and increased stigma perception [4].In this clinical context, QoL may be influenced by several factors.Seizure frequency and recency were usually considered major determinants of QoL in people with active epilepsy.Yet, recent data clearly evidenced that psychosocial factors (mostly depression, anxiety and social stigma) may have stronger effect than epilepsy-related ones in both patients with and without complete control of seizures [1,2].Moreover, previous research has shown that there is a complex link between seizure control and QoL, with some individuals reporting poor QoL despite seizure freedom [5] and others reporting good QoL despite continuous seizures [6].On this issue, some researchers pointed out that several psychological and social factors may contribute to optimize QOL for people with a long-term condition such as epilepsy [7].In particular, they found that resilience is an essential factor for understanding QoL in epilepsy.
Resilience is defined as a dynamic process that includes positive adaptation in the context of significant adversity [8].The relationship between resilience and epilepsy is poorly understood, and compared to the abundance of investigations throughout the health research spectrum, there are very few studies examining the impact of resilience in epilepsy [7].Some authors reported that social vulnerability, more adverse effects of antiseizure treatment and depression are important factors associated with resilience in epilepsy [9].Furthermore, higher resilience was found to be significantly correlated with the lack of depressive symptoms and improved QoL in PWE [10] (QoL).Finally, we recently demonstrated that resilience is a strong independent predictor of anxiety and has the ability to modulate its severity in PWE [11].However, to our best knowledge no study so far has specifically investigated the effect of resilience on psychological factors and stigma perception in PWE with longitudinal design.Indeed, the epilepsy is a highly stigmatizing condition [12], felt stigma related to epilepsy affects nearly half of PWE across different continents and has shown to have a negative impact on QoL [13], as documented in several countries, in relation to the cultural and economic environment.
The aim of our study was to evaluate the resilience, through a dedicated scale, in a group of adult PWE and its impact on QoL and psychosocial factors, in particular the presence of stigmatization feelings.We hypothesize that patients with higher resilience scores should be able to better cope with difficulties linked to seizures and epilepsy, thus leading to fewer depressive and anxious symptoms and greater social outcome (improved QoL and reduced stigma perception).Understanding these interactions could help us to more effectively target psychosocial education programs that have been proven to significantly improve QoL and foster positive attitudes toward epilepsy.[14].

Study population and experimental design
In our outpatient clinic for epilepsy (University Campus Bio-Medico of Rome) we consecutively enrolled 78 adults PWE (men/female: 26/52, 42.5 ± 16.2 years old) for the purpose of this study.Tables 1 and 2 report the demographic and clinical characteristics of our PWE cohort.All subjects resided in Rome and in the central regions of Italy.After providing written informed consent and receiving access to the online questionnaires using a unique shared link, subjects were requested to complete them.PWE were required to provide digital consent at the start of the survey to go through the questionnaires; if consent was not given, the survey was terminated.Afterwards, the participants were required to respond to a multiple-choice online survey using a private Google platform.Each patient was given an alphanumeric code that was progressive, in accordance with the GDPR 2016/679 privacy rule.One investigator (FN) kept track of the correspondence between each progressive alphanumeric code and patient identity on a separate worksheet after pseudo-anonymizing the data.Given that the questionnaires were pseudo-anonymized, no personal information was explicitly saved online.The local Ethics Committee gave its approval (PAR 05.22 OSS) to the study procedure.All procedures were in accordance with the 1964 Helsinki Declaration and its later amendments, as well as the institutional and/or national research committee's ethical requirements.
Inclusion criteria were: i) patients affected by focal or generalized epilepsy; ii) PWE > 18 years old; iii) PWE already in follow-up at our epilepsy center (University Campus Bio-Medico of Rome).
The exclusion criteria were intellectual disability, inability to use electronic devices, history of neurological disorders other than epilepsy, presence of severe psychiatric and medical comorbidities.
The Quality of Life in Epilepsy Scale -QOLIE-31 (Q31) [20] is employed as a brief assessment of quality of life in epilepsy.It has seven subscales (seizure worry, overall quality of life, emotional well-being, energy/fatigue, cognitive, medication effects, and social function) that address different psychological, emotional, social, and medical issues.Subscale scores are added together and weighted to create the overall score.
The Stigma Scale of epilepsy (SSE) is a 24-item multiple-choice questionnaire grouped in 5 areas.It addresses a wide range of issues related to how epilepsy is seen in many circumstances (attitudes and behavior toward PWE, perception and sentiments toward seizures, social factors related to the diagnosis of epilepsy).Each response was scored with a 4-point scale ranging from 1 (not at all) to 4 (completely).The total score, which ranges from 0 (no stigma) to 100 (highest stigma), indicates the level of stigma that each patient perceives.We previously documented a very good internal consistency of the Italian version of the SSE (Cronbach's α coefficient of 0.880 for PWE) [13].

Statistical analysis
Descriptive statistics are presented as mean ± standard deviation or median ± range, depending on data distribution.Data distribution was assessed through Kolmogorov-Smirnov tests.
Parametric (t-test, univariate ANOVA) and non-parametric test as appropriate were used to evaluate any effect on the resilience (as evaluated by R-14 scores) by several factors (type of seizures, absence of seizures, epilepsy type, epilepsy syndrome, etc.).Moreover, Spearman correlation coefficients were used to correlate RS-14 score with different factors (age, education, duration of epilepsy, seizure frequency, number of ASMs) and scores of clinical scales obtained at T0 (Q31, BDI-II and GAD) and at T1 (SSE, BDI-II and GAD).The seizure rate was considered as the total number of all type of seizures during 4 months before the beginning of the study (baseline period; T0) and at follow-up (T1).We considered also in same periods of time the total number of generalized tonic-clonic (GTC) seizures, both belonging to generalized and focal epilepsies (in last case focal to bilateral tonic-clonic seizures).
To better study the factors associated with resilience, all sample of PWE was divided in 4 groups depending on the presence/absence of seizures (seizure free and non-seizure free patients, respectively identified as SF and NSF) and Q31 scores at T0 (patients with Q31 score higher/lower than median value, named respectively HiQ31 and LoQ31): a. NSF_HiQ31: patients showing higher values of Q31 despite not being seizure free; they may be considered as "resilient" group.b.NSF_LoQ31: not seizure free patients with lower Q31 values.c.SF_HiQ31: seizure free patients showing higher Q31 values.d.SF_LoQ31: patients showing lower Q31 values despite being seizure free; they may be considered as "vulnerable" group.
One-way ANOVA was applied to investigate variance of resilience (RS-14 score) and other psychosocial variables of interest (BDI-II, GAD-7, SSE) across these 4 different groups.
Furthermore, we performed a linear regression analysis to describe significant relationships between our dependent variables (RS-14 ad SSE) and other demographic and psychometric variables.Then, we employed a multiple stepwise regression analysis to identify the best predictors for resilience and stigma perception related to epilepsy.
Finally, at follow-up four categories of clinical outcome were considered: i) seizure freedom; ii) reduction in seizures' frequency > %; iii) reduction in seizures' frequency < 50 %; iv) worsening or no changes of seizures' frequency.Thus, we searched for possible modifications in seizure's frequency and psychological variables (BDI-II and GAD-7) between T0 and T1.
All statistical analyses were considered significant if the p-value was < 0.05.When necessary, the Bonferroni correction was used to address the alpha inflation due to multiple comparisons.IBM SPSS Statistics version 28.0 (SPSS, Chicago, IL) was used for statistical analyses.

Demographic and clinical features of patients
The PWE's demographic characteristics are shown in Table 1.The average age of the PWE sample was 42.5 ± 16.2 years old, their mean education level was 12.8 ± 3.4 years.Thirty-five (44.9 %) PWE were employed, 15 (19.2 %) were student, 19 (24.3 %) were unemployed and 9 (11.6 %) were retired.Considering civil status, 24 out of 78 (30.8 %) were married and other 18 (23.1 %) had cohabitant or stabile partner, 31 (39.7 %) were single, 2 (2.6 %) divorced and the remaining 3 (3.8%) were widower.Thirty-three (42.3 %) PWE did not drive.Based on the neuroimaging data (brain CT scan and/or MRI), as well as the clinical and EEG features, 35 (44.9 %) out of 78 patients were diagnosed as having focal epilepsy of unknown etiology, 29 (37.2 %) as having structural focal epilepsy and the remaining 14 (17.9 %) as having idiopathic generalized epilepsy (Table 2).Considering focal epilepsies, 47 patients were affected by temporal lobe syndrome and the remaining 17 by extratemporal lobe syndrome.
Seizure frequency was evaluated through clinical diaries that were kept by the patients and their caregivers.During the 4 months before the beginning of the study (baseline period), 36 patients (46.1 %) were seizure-free (SF); the mean seizure frequency for all type of seizures in the remaining 42 patients at T0 was 27.4 ± 55.8, (median: 5.5; range: 1-360) and for GTC seizures 4.2 ± 12.3 (median: 0; range: 0-85).Type, number, dosage, and administration schedule of medications taken by the PWE were recorded: 44 (56.4 %) patients were on a single ASM, (32.1 %) were taking 2 ASMs, 7 patients (8.9 %) 3 ASMs, and only patient (2.6 %) was not taking any ASM.Indeed, the patients were enrolled in our outpatient clinic for epilepsy and most of them were already treated with one or more ASMs (usually antiseizure therapy began at the end of hospitalization).The median number of ASMs used was 1 (range 0-3).Epilepsy mean duration was 15.6 ± 13.9 years (Table 2).
No correlation was observed between RS-14 scores and seizure frequency at both times (T0 and T1) as well as changes of number of ASMs, seizure rate or clinical outcome.In addition, no correlation was documented between resilience and demographic features (age, sex, education, residence, job, civil status, driving) and clinical characteristics related to epilepsy (duration of epilepsy, epilepsy type and syndrome).

Multiple regression analysis
We found that BDI-II [p <.001], GAD-7 [p <.001] and Q31 overall score at T0 [p <.001] evidenced a linear relationship with RS-14 score.Using a multiple stepwise regression analysis, setting RS-14 score as

Discussion
The results of our study clearly evidenced that in PWE resilience is significantly correlated to psychosocial factors: fewer depressive and anxious symptoms (both a T0 and T1) as well as better QoL perception were associated with higher scores of resilience.As a novel finding higher resilience values as evaluated through RS-14 at baseline (T0) correlated with fewer stigmatization feelings at follow-up (T1).
The concept of resilience has been the subject of an intriguing debate over several years [22,23].Ungar [24] proposed that resilience is an interactive process in which the individual acts as a resource seeker and the environment, in its broadest sense, provides resources.Along this line, Van Kessel [25] recently conducted a systematic review on resilience in adults over 50 years old and concluded that the evidence currently available supports both the trait and the environmental theories of resilience.The availability of resources, access to medical care, Fig. 3. One-way ANOVA analysis for resilience (RES-14; p < 001) and other psychosocial features at follow-up (T1): depression (BDI-II; p= <0.001), anxiety (GAD-7; p <.001), and stigmatization feelings (SSE; p =.004).Considering pairwise comparisons, the 'resilient' group (NSF_HiQ31) showed significant fewer depressive feelings (p =.001; **) and anxious symptoms (p <.001; **) and reduced stigma perception (p =.038;*) than more 'vulnerable' group (SF_LoQ31).Fig. 4. Changes of seizure frequency, depressive (BDI-II) and anxious (GAD-7) symptoms between T0 and T1.Despite a significant reduction in the total number of seizures (18,34 ± 50,45 vs 7,90 ± 27,62; p =.034) at follow-up (T1), we did not observe significant modifications in depressive and anxious symptoms.political and societal factors, and the availability of social support are examples of environmental influences.Different theories of resilience have led to significant differences in the operationalization and measurement of resilience across studies, with an increasing number of quantitative scales [26].It is interesting to note that, despite the growing recognition of resilience as a multifactorial construct, literature reviews often highlight the tendency for resilience research to focus on the "individual" rather than their larger social context [7].Although the field of resilience research originated in developmental psychology, it has been extensively applied to health research with the aim to learn how adults cope with chronic disorders like diabetes [27], chronic pain [28], HIV and AIDS [29] and cancer [30].In comparison to the abundance of these studies, those investigating the role of resilience in PWE are spare.In 2011, Taylor et al. [9] found that the most significant predictors for resilient outcome (patients who reported good QoL despite poor seizure control and socioeconomic disadvantage) were absence of depression, fewer adverse treatment effects, and good QoL at baseline (diagnosis of their epilepsy).On the contrary, significant predictors of a vulnerable outcome (patients who reported poor QoL in the context of seizure freedom and socioeconomic advantage) were fair or poor health perception, depression, reduced sense of mastery and more adverse effects of ASMs.Anyway, the authors did not perform a direct evaluation of resilience through a dedicated scale.Later some authors [31] evidenced that resilience was positively correlated with selfcompassion, employment status (being employed was associated with higher resilience), age and negatively with seizure severity.The authors also found that self-compassion predicted higher resilience as well as lower depression and anxiety.Finally, more recently Tedrus et al. [10] showed that higher resilience was associated with seizure control, normal EEG background activity, and monotherapy with ASM, but not with the epileptic syndrome; depressive symptoms correlated negatively with resilience and individuals with greater resilience perceive better QoL.
The data of our study thereby confirm that resilience is strongly associated with psychosocial factors.In particular, the presence of depressive and anxious symptoms negatively correlated with RS-14 scores, suggesting that resilience may be a protective factor against changes of mood and anxiety, helping PWE overcome adversities and challenges facing during life, in particular unpredictability of seizures and social difficulties.Furthermore, depressive symptoms resulted to be the predictive variable best explaining resilience.This finding could indicate that the relationship between depression and resilience is possibly bidirectional: the presence of depressive symptoms may affect resilience and impair PWE's capacity to develop and apply coping strategies, as previous studies suggested [9,31], while low resilience could represent a risk factor in developing depressive symptoms.
According to literature evidence [9,10], improved QoL was associated with greater resilience, and 'resilient trait' may help to explain how patients with poor seizure control and/or disadvantaged socioeconomic status perceive good QoL.Our analysis evidenced that these 'resilient' PWE showed better psychosocial features: less depressive and anxious symptoms, and fewer stigmatization feelings.On the opposite, in the 'vulnerable' patients lower scores of resilience are associated with worse psychological features and higher stigmatization perception.Indeed, as a novel finding in a sample of PWE we demonstrated a significant negative correlation between resilience evaluated at baseline and stigmatization perception related to epilepsy at follow-up.Unfortunately, nowadays the social stigma related to epilepsy is still present in many societies and cultures, both in Western countries as well as in developing ones [32].Many factors [33] represented by demographic, illnessrelated, and psychosocial features may predict stigma in PWE, and these associations were found to be highly culturally specific.The results of present study again confirmed the higher depressive symptoms and worse QoL are the most significant predictor of stigma in PWE, as previously demonstrated [2,4,13,33].For the first time to our knowledge, our study also evidenced that resilience could play a protecting role in this context.People with epilepsy may face discrimination, misunderstanding, or exclusion due to the condition.Developing resilience involves finding ways to cope with epilepsy-related difficulties as well social stigma, such as seeking support from loved ones or joining support groups.Epilepsy can impact emotional well-being, leading to feelings of frustration, depression, or isolation.Resilience entails recognizing and addressing these emotional challenges.Seeking therapy or counselling can be beneficial for individuals with epilepsy, as it provides a safe space to discuss emotions and learn coping strategies.If we accept resilience theory as a multifactorial construct or process rather than only an innate and immutable trait, it is mandatory to support PWE by improving their mood, reducing anxiety or gain a sense of mastery as well as ameliorating social context by building a support network.This effort in turn will enable individuals to develop resilience and be better able to deal with the challenges associated with epilepsy.
In our study demographic features and epilepsy-related features (i.e., seizure frequency, epilepsy type and syndrome, duration of epilepsy, ASMs) did not affect resilience and stigma perception contrary to previous observations [9,10,33].The poor characterization of socioeconomic status and limited dimension of subjects' sample may explain the lack of statistical evidence.On the other hand, according to literature [2,34] seizures' frequency showed a significant direct correlation with depressive symptoms and an inverse correlation with QoL.
Finally, the longitudinal study evidenced that despite a significant reduction in the number of seizures at follow-up, patients did not show significant changes in depressive and anxiety symptoms.This result again suggests that these psychological factors in PWE are more influenced by sociodemographic [35] and emotional [4] features than epilepsy-related ones.Resilience seems to play a crucial protective role against depressive symptoms and anxiety.

Limitations of the study
As previously stated, the small dimension of subjects' sample from a single institution may represent a limitation of the study.Anyway, the sample owns the typical features that are essentially representative of general population affected by epilepsy and has been followed-up longitudinally, while most studies on this topic only provide crosssectional evaluations.The duration of follow-up is quite short, and we cannot exclude that it influenced or determined the lack of significant changes in depressive and anxious symptoms.
Finally, the absence of a detailed socioeconomic status characterization prevents us from highlighting features that could significantly affect psychosocial factors, as previously elucidated [9,33].

Conclusions
This study provides compelling evidence of the multifaceted role of resilience in the lives of PWE.Depressive symptoms, anxiety and quality of life were significantly associated with resilience, which was found to affect the perception of stigma related to epilepsy more than seizures.The revelation of resilience as a protective factor against stigma is a novel contribution to this field.This suggests that fostering resilience in individuals with epilepsy may not only enhance their individual wellbeing but also contribute to the reduction of societal prejudices and stigmatization.Based on these findings, psychoeducation interventions and support programs could develop targeted strategies for enhancing resilience in PWE, aiming to create a more supportive and understanding societal environment.Notably, the lack of correlation between resilience/stigma and seizure frequency or clinical outcomes highlights the distinct influence of psychosocial factors on the well-being of individuals with epilepsy.It emphasizes the importance of addressing not only the medical aspects of epilepsy by improving seizure control, optimizing antiseizure treatment and reducing the adverse effects of ASMs, but also social and psychological dimensions to comprehensively ameliorate the quality of life of PWE.Namely, depressive symptoms should be early identified and treated comprehensively, as they can significantly influence and predict both resilience and feelings of stigmatization.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Table 1
Demographic features of People with Epilepsy (PWE).

Table 2
Clinical features of people with epilepsy (PWE).