Association Between Frailty and Antiseizure Medication Tolerability in Older Adults With Epilepsy

Abstract

Background and Objective

Frailty is an important aspect of biological aging, referring to the increased vulnerability of individuals with frailty to physical and psychological stressors. While older adults with epilepsy are an important and distinct clinical group, there are no data on frailty in this population. We hypothesize that frailty will correlate with the seizure frequency and especially the tolerability of antiseizure medications (ASMs) in older adults with epilepsy.

Methods

We recruited individuals aged 60 years or older with active epilepsy from 4 Canadian hospital centers. We reported the seizure frequency in the 3 months preceding the interview, while ASM tolerability was quantified using the Liverpool Adverse Events Profile (LAEP). We applied 3 measures of frailty: grip strength as a measure of physical frailty, 1 self-reported score (Edmonton frail score [EFS]), and 1 scale completed by a healthcare professional (clinical frailty scale [CFS]). We also administered standardized questionnaires measuring levels of anxiety, depression, functional disability, and quality of life and obtained relevant clinical and demographic data.

Results

Forty-three women and 43 men aged 60–93 years were recruited, 87% of whom had focal epilepsy, with an average frequency of 3.4 seizures per month. Multiple linear regression and zero-inflated negative binomial regression models showed that EFS and CFS scores were associated with decreased ASM tolerability, each point increase leading to 1.83 (95% CI: 0.67–4.30) and 2.49 (95% CI: 1.27–2.39) point increases on the LAEP scale, respectively. Neither the EFS and CFS scores nor grip strength were significantly associated with seizure frequency. The EFS was moderately correlated with depression, anxiety, quality of life, and functional disability, demonstrating the best construct validity among the 3 tested measures of frailty.

Discussion

The EFS was significantly, both statistically and clinically, associated with ASM tolerability. It also showed multiple advantages in performance while assessing for frailty in older adults with epilepsy, when compared with the 2 other measures of frailty that we tested. Future studies must focus on what role the EFS during epilepsy diagnosis may play in ASM selection among older adults with epilepsy.

Epilepsy is a neurologic disorder characterized by excessive, hypersynchronous, and abnormal neuronal activity. This disorder frequently develops in later life, with a higher incidence and prevalence in older adults when compared with that in younger individuals.¹ The lifetime prevalence of active epilepsy can be as high as 16 per 1,000 older people in high-income countries,² with a cumulative incidence of 61 new cases per 1,000 individuals per year.³ The risk of premature mortality among these individuals remains twice as high as that of the general elderly population, often caused directly or indirectly by seizures, side effects of antiseizure medications (ASMs), or underlying medical conditions.¹ As global populations age and life expectancies increase, the burden of epilepsy among older adults will continue to rise.

Despite the importance of this demographic group, little research has been conducted on older adults with epilepsy. They are a clinically distinct group because the types and etiologies of epilepsy and the pharmacodynamics and pharmacokinetics of ASMs are different in older adults when compared with those in younger people.⁴ There is also little literature on the efficacy and tolerability of newer generation ASMs in older adults with epilepsy. This is a major health issue, particularly in individuals at risk of drug interactions and afflicted with comorbidities.⁴ Clinical trials often exclude older or frail participants for a variety of reasons, including the presence of comorbidities, polypharmacy, and altered metabolism. Their management is based on our knowledge of epilepsy in young adults and is not necessarily adapted to the physiologic changes and comorbidities brought on by age and frailty.

Frailty increases in prevalence with age and is generally characterized by the loss of function and autonomy, cognitive decline, and increased vulnerability to physical and psychological insults.⁵ Frailty carries an increased risk of chronic diseases such as cardiovascular disease and other negative health outcomes such as falls and fractures, delirium, hospitalization, admission to continuing care, and mortality.⁶ While many authors consider only physical factors in their definition of frailty, there is general consensus that frailty is also strongly related to other psychosocial and psychiatric factors, including depression, anxiety, quality of life, and social isolation.⁷ Investigators have developed a number of different clinical scales to measure frailty severity in an individual, including the Edmonton frail scale (EFS), the clinical frailty scale (CFS), and the physical performance–based Fried phenotype (which includes grip strength). Despite a high incidence rate of frailty among older adults, counting 43.4 new cases per 1,000 person-years,⁸ there are no studies examining the relationship between frailty and various aspects of epilepsy in older adults, including the efficacy and tolerability of ASMs.

The primary objective of this study was to measure and characterize the association between frailty and seizure frequency and ASM tolerability in older adults with epilepsy. Our second objective was to compare the performance of 3 different clinical measures of frailty (EFS, CFS, and grip strength) and to compare their construct validity, relative to functional disability, quality of life, anxiety, and depression. We hypothesize that increasing frailty will be associated with decreasing functional disability, quality of life, anxiety, and depression.

Methods

Participant Recruitment

Patients aged 60 years and older, with a recent or long-standing diagnosis of epilepsy and receiving at least 1 ASM were consecutively invited to participate in this study following their routine appointment at the ambulatory epilepsy clinics at 4 Canadian epilepsy centers (CHUM Hospital in Montreal; CHUS Hospital in Sherbrooke; Foothills Medical Centre in Calgary; and Royal University Hospital in Saskatoon) from September 1, 2019, to January 13, 2021. From October 2020, following modifications required by the COVID-19 pandemic, all participant encounters occurred either at their home (after an initial conversation through telephone to confirm their interest in participating in the study) or were performed using a videoconferencing software (e.g., Zoom). Individuals with significant cognitive impairment or intellectual disability, as determined by the treating physician, were excluded.

Questionnaires

Frailty was assessed using the EFS, a self-reported questionnaire (possible total score range 0–17)⁹; the CFS, completed by a clinician based on the participant's physical condition and overall demeanor (possible total score range 1–9)¹⁰; and grip strength, an objective measure of physical strength,¹¹ recorded in kilograms using a Jamar Smart Digital Hand Dynamometer.¹² These frailty measures were chosen because of their simplicity of administration and their performance in previous studies.

The CFS was completed by the same interviewer across centers in Montreal, Calgary, and Sherbrooke (the latter 2 performed through videoconferencing) while a second interviewer conducted interviews in Saskatoon. The CFS was assessed as soon as consent was obtained and before measuring the grip strength and administering the EFS and other questionnaires. The evaluator was therefore blinded to the participants' responses to the questionnaires and grip strength to avoid a CFS evaluation that would be biased by a participant's EFS or grip strength performance. Five previously validated questionnaires were then completed by participants during the interview, measuring depression Neurological Disorders Depression Inventory for Epilepsy [NDDI-E]),¹³ anxiety (Generalized Anxiety Disorder 7-item scale [GAD-7]),¹⁴ quality of life (Quality of Life in Epilepsy 10-item scale [QOLIE-10]),¹⁵ adverse effects of ASMs (Liverpool Adverse Events Profile [LAEP]),¹⁶ and degree of functional disability (Activities of Daily Living [OARS-ADL]).^17,18 Among these scales and questionnaires, previously validated French versions of the GAD-7,¹⁹ NDDI-E,²⁰ and QOLIE-10²¹ were administered to participants at the Montreal and Sherbrooke sites. The characteristics of the scales and questionnaires used are summarized in Table 1. Seizure history and personal data, including current medications and somatic and psychiatric comorbidities, were collected from the patients' medical file. The overall burden of comorbidity was measured using the epilepsy-specific comorbidity index (ESI).²²

Table 1.

Characteristics of the Scales and Questionnaires Used

We considered the seizure frequency reported in the 3 months preceding the interview as an imperfect measure related to the concept of ASM efficacy. ASM tolerability was based on the LAEP score, a 19-item scale with potential scores ranging from 19 to 76 and which measures the number and severity of adverse events associated with ASM use (further details are summarized in Table 1).

Analyses

Participants were dichotomized into “frail” and “robust” categories based on their EFS, CFS, or grip strength scores. Grip strength below the 20th percentile of a group or population is widely used as an indicator of frailty.¹¹ In this study, the cutoff was accordingly set at 15.4 kg in women and 27.6 kg in men. Individuals with a CFS score equal or higher than 4 or an EFS equal or higher than 8 were placed in the “frail” category (Table 1). These cutoffs were used in a previous study and are based on the ordinal categories of the 2 scales.²³

We built multiple linear and zero-inflated negative binomial regression (ZINBR) models to characterize the cross-sectional associations between tolerability and seizure frequency, respectively, and frailty (as independent continuous variables in each model). Multiple linear model assumptions (i.e., linear relationship between dependent and independent variables, homoscedasticity, independence between observations, and normal distribution of the residuals) were confirmed by plotting residuals. ZINBR models were built to account for overrepresentation of zeros and overdispersion when modeling seizure frequency.²⁴ The remaining 3 assumptions (i.e., independence, linearity, and Poisson response) were confirmed. One individual reported having 450 focal seizures in the 3 months preceding the interview and had to be excluded from the ZINBR models.

We also assessed the agreement between the EFS, CFS, and grip strength in their ability to detect frailty using Cohen kappa coefficients (κ) (95% CI). Agreement coefficients were interpreted as follows: values ≤0 as indicating no agreement, 0.01–0.20 as none to slight, 0.21–0.40 as fair, 0.41–0.60 as moderate, 0.61–0.80 as substantial, and 0.81–1.00 as almost perfect agreement.²⁵

We also assessed the construct validity of the scales. Construct validity is the extent to which a test or questionnaire measures what it purports to measure, when the construct cannot be directly observed.²⁶ It consists namely of 2 subtypes: convergent validity and discriminant validity. Convergent validity refers to the fact that the construct strongly corresponds to a similar measure, whereas discriminant validity is established when there is no relationship between the construct and a dissimilar measure.²⁶ To study convergent construct validity, we estimated Spearman rank correlation coefficients between each measure of frailty and other clinical variables deemed relevant, namely depression, anxiety, quality of life, and functional disability, represented by the NDDI-E, GAD-7, QOLIE-10, and OARS-ADL scores. Functional disability is one of the components of the frailty syndrome.⁵ Several studies have shown that there is a reciprocal relationship between depression and anxiety and frailty, with each potentially contributing to the other.^27,28 There is a strong and well-documented association between lower quality of life and frailty.²⁹ Correlation coefficients were interpreted as follows: ±0.0–0.09 as indicating no correlation, ±0.1–0.3 as weak, ±0.4–0.6 as moderate, ±0.7–0.9 as strong, and ±1.00 as perfect correlation.³⁰

Statistical analyses were performed using R software,³¹ including the MASS,³² stats,³¹ and pscl³³ packages. The significance level was set at p < 0.05.

Standard Protocol Approvals, Registrations, and Patient Consents

Our study was approved by the research ethics boards of each participating institution. We obtained written and informed consent from each study participant.

Data Availability

Anonymized data may be made available by request from the corresponding author.

Results

Participants

Eighty-six participants ranging from 60 to 93 years of age (mean = 68.9; SD = 6.8) were recruited from the 4 epilepsy centers. Of the 43 women and 43 men, 87% had focal epilepsy, 8% had generalized epilepsy, and 5% had epilepsy of unknown origin. Forty-eight participants reported having a history of nonconvulsive seizures (56%), 17 (20%) a history of convulsive seizures, and 21 (24%) a history of both convulsive and nonconvulsive seizures. A mean of 3.8 (SD = 2.7) comorbidities and 3.8 (SD = 3.0) daily medications other than ASMs were reported per participant. The most frequent comorbidities were dyslipidemia (36%), hypertension (36%), benign or malignant neoplasia (27%), anxiety (19%), and hypothyroidism (17%). Only 7% of individuals had had a stroke in their lifetime, and 7% had been diagnosed with dementia.

We identified 53 robust and 33 (38%) frail individuals based on their EFS score. Of the 69 participants in whom grip strength was measured (before COVID-19), 15 (22%) were deemed frail, with 7 women whose maximum grip strength was less than 17.4 kg and 8 men whose grip strength was less than 27.6 kg. Thirteen (15%) individuals reached the frailty threshold based on their CFS score (Table 2).

Table 2.

Participant Characteristics and Questionnaire Scores According to Frailty Status

Individuals deemed frail according to their EFS had significantly more comorbidities than their robust peers (mean difference [MD] = 1.5; 95% CI: 0.36–2.69; Table 2). This difference, however, was not reflected in their ESI comorbidity score (MD = 0.1; 95% CI: −1.0 to 1.1) (Table 2). A significant difference was detected in their mean score on the other scales, namely in their levels of anxiety (GAD-7; MD = 5.0; 95% CI: 3.2–6.9), depression (NDDI-E; MD = 2.8; 95% CI: 1.2–4.5), quality of life (QOLIE-10; MD = 6.1; 95% CI: 3.3–8.8), and functional disability (ADL; MD = 3.0; 95% CI: −4.1 to −1.8) (Table 2), between frail and robust individuals.

The mean CFS score was 2.3 (SD = 1.4) in women and 2.2 (SD = 1.2) in men (Figure 1B). Individuals with frailty were significantly older than robust individuals (MD = 5.4 years; 95% CI: 1.5–9.3) (Table 2) and demonstrated greater functional disability (MD = 5.0; 95% CI: 3.6–6.4) and depression levels (MD = 2.4; 95% CI: 0.10–4.7) (Table 2). They also reported a significantly lower quality of life than robust adults (MD = 4.4; 95% CI: 0.33–8.40) (Table 2).

Figure 1. Distribution of the 3 Frailty Measures Scores, Stratified by Sex.

(A) Edmonton frail scale (EFS) scores, (B) clinical frailty scale (CFS) scores, and (C) maximum grip strength (GS). The heavy dashed line represents the threshold between robustness (EFS ≤ 5; CFS ≤ 3; GS ≤ 15.4 kg in women) and frailty (EFS > 5; CFS > 3; GS > 15.4 kg in women), whereas the lighter dotted line represents the frailty threshold (GS ≤ 27.7 kg) in men.

While women had an average hand grip strength of 20.0 kg (SD = 5.7), it reached 35.2 kg (SD = 9.6) among men (Figure 1C). As with the CFS, when their status was dichotomized according to grip strength, frail individuals were older than their robust peers (MD = 5.35; 95% CI: 1.5–9.2) and fewer months had elapsed since their last seizure (Table 2). Similar to the EFS and CFS, a significant difference was also observed in functional disability between the groups (MD = 2.9; 95% CI: −4.6 to −1.1) (Table 2).

Association Between Frailty, Seizure Frequency, and Tolerability of ASMs

The tolerability of ASMs, defined by a decreasing LAEP score, was inversely associated with frailty as measured by the EFS and the CFS (Table 3), after adjusting for age, sex, and comorbidity burden. For each point increase in the EFS and the CFS, the LAEP increased on average 1.83 points (95% CI: 1.27–2.39) and 2.49 points (95% CI: 0.67–4.30), respectively. Tolerability was not significantly associated with grip strength in men or in women.

Table 3.

Multiple Linear Regression and Zero-Inflated Negative Binomial Regression Models, Adjusted for Sex, Age, and Comorbidity Burden (ESI Score)

As summarized in Table 3, the seizure frequency was not significantly associated with frailty as defined by the EFS or CFS, nor grip strength in both sexes, after adjusting for age, sex, and comorbidity burden.

Association Between the Frailty Scales and Different ASM Tolerability Domains

The mean neurologic adverse event scores were significantly higher among frail vs robust individuals according to the EFS (MD = 3.48 [95% CI: 1.77–5.19]) (Figure 2A). Intergroup differences were equally significant for cognitive (MD = 2.63 [95% CI: 1.18–4.07]), physical (MD = 1.77 [95% CI: 0.49–3.05]), and psychiatric adverse events (MD = 2.42 [95% CI: 1.33–3.50]). No significant difference was detected between groups divided by CFS (Figure 2B) or hand grip strength (Figure 2C). Similar results were found when adverse events were classified according to the statistically derived categories identified by multifactorial analysis.³⁴ These results are illustrated in eFigure 1, links.lww.com/WNL/C545.

Figure 2. Mean Score Given to LAEP Items, Grouped by Adverse Event Category Reported by Frail, and Robust Individuals According to Their EFS Sscore (A), CFS Score (B), or Grip Strength (C).

Cognitive events (score range: 4–16) include difficulty in concentrating, sleepiness, memory problems, and disturbed sleep; neurological events (score range: 6–24) include unsteadiness, tiredness, headaches, blurry vision, dizziness, and shaky hands; physical events (score range: 5–20) include hair loss, problems with skin, upset stomach, trouble with mouth and gums, and weight gain; and psychiatric events (score range: 4–16) include restlessness, aggressivity, nervousness and/or aggressivity, and depression. Standard error bars and 95% CIs around mean differences (MDs) are shown. p values were obtained by performing a 2-tailed t test. CFS = Clinical frailty scale; LAEP = Liverpool Adverse Events Profile.

We further examined the score given to each item of the LAEP items by participants deemed robust or frail according to their EFS score. As opposed to their robust peers, (Figure 3B), individuals with frailty frequently reported experiencing adverse events “always or often” and rarely answered “never” encountering events (Figure 3A). Adverse events also differed across groups. Tiredness, unsteadiness, and memory problems were often reported in both groups. On the contrary, agitation, feelings of aggression and/or nervousness, headaches, depression, and hand tremors were much more common among individuals with frailty. No significant differences were detected between individuals deemed frail or robust according to their CFS (Figure 3, C and D) or hand grip strength (Figure 3, E and F).

Figure 3. Frequency (Percentage) of Each Adverse Event Reported in the LAEP by Frail or Robust Participants.

Individuals were dichotomized into frail (A) or robust (B) participants according to their EFS score; frail (C) or robust (D) participants according to their CFS score; and frail (E) or robust (F) participants according to their grip strength. CFS = Clinical frailty scale; LAEP = Liverpool Adverse Events Profile; EFS = Edmonton frail scale.

Construct Validity and Interscale Reliability

The results of the construct validity of the frailty scales are summarized in Figure 4. EFS scores were moderately correlated with depression (the Spearman Rho = 0.4 [95% CI: 0.23–0.59]), anxiety (the Spearman Rho = 0.5 [95% CI: 0.32–0.66]), quality of life (the Spearman Rho = 0.5 [95% CI: 0.40–0.68]), and functional disability (the Spearman Rho = −0.6 [95% CI: −0.76 to −0.49]). With the exception of functional status with which it was moderately correlated (the Spearman Rho = −0.6 [95% CI: −0.77 to −0.46]), CFS scores were weakly correlated with these other traits. The correlation between grip strength and other traits were generally not statistically significant. The only significant correlation was among women where grip strength was weakly correlated with functional disability (the Spearman Rho = 0.3 [95% CI: 0.01–0.74]). All p values were inferior to <0.05.

Figure 4. Spearman Correlation Coefficients, Between Frailty Measured Using 3 Different Methods and Other Epilepsy-Related Traits Along With 95% CIs.

A correlation was considered null when the CI crossed zero, regardless of the coefficient.

We found a moderate correlation between EFS and CFS scores (the Spearman Rho = 0.5 [95% CI: 0.37–0.69]) (Figure 4). Men's grip strength was weakly correlated with CFS scores (the Spearman Rho = −0.3 [95% CI: −0.65 to −0.03]) while women's was moderately correlated (the Spearman Rho = −0.6 [95% CI: −0.81 to −0.26]). The correlations between grip strength (for men and women) and EFS scores, on the contrary, were not statistically significant.

Grip strength showed fair agreement with the CFS (κ = 0.34 [95% CI: 0.07–0.61]) when used to categorize individuals as frail or robust. The EFS showed slight agreement with grip strength (κ = 0.15 [95% CI: −0.06 to 0.37]) and fair agreement with the CFS (κ = 0.33 [95% CI: 0.15–0.53]).

Discussion

Epilepsy at a later age brings special considerations, including choosing an ASM that is efficacious and as importantly well tolerated. A central aspect of advancing age is the associated increase in the prevalence of frailty, a syndrome characterized by physical and cognitive vulnerability to stressors, often combined with loss of functional independence and associated with negative health outcomes. Although it is increasingly studied as a risk factor for the development of cardiovascular disease and a predictor of mortality and healthcare utilization,^35,36 frailty remains understudied in neurology. Our study presents an investigation of the role of frailty in the pharmacologic treatment of older individuals with epilepsy.

We did not identify an association between frailty using any of the 3 measures and seizure frequency, after adjusting for age, sex, and comorbidity. One prior intervention study reported a significant reduction in seizures in 14 women with refractory epilepsy enrolled in a 15-week aerobics program.³⁷ The beneficial effect of the physical exercise did not last after the end of the program. Consistent with our findings, however, exercise had no impact on seizure frequency in 2 similar studies with larger sample sizes.^38,39

Our study, on the contrary, demonstrates strong evidence that frailty in older adults with epilepsy correlates with worse ASM tolerability. We found that for each 1-point increase in the EFS, the LAEP was increased by 1.80 points. With possible EFS scores between 0 and 17, a maximally severe EFS score would be associated with a more than 30-point increase in the LAEP score, which would be above the toxicity threshold (generally defined as an increase in 26 points from the baseline score of 19).⁴⁰ At the level of individual LAEP items, a 2-point increase results in the LAEP score changing from 1 (never) to 3 (sometimes), which may be sufficient to warrant a change in ASM. The mechanism by which frailty exerts such an influence on drug tolerability is uncertain. Studies have shown that chronic inflammation, minor injuries, illnesses, and polypharmacy may affect pharmacokinetics, leading to increased drug toxicity.⁴¹ Drug metabolism is further altered by changes in body composition, reduced physiologic reserves, and reduced renal and hepatic function in advancing age.⁴² All these age-related changes characterize frailty. Authors have suggested a shared genetic predisposition between frailty and drug tolerability,⁴³ and the consequences of frailty-associated multimorbidity and related polypharmacy.⁴⁴

Because of the distinct character and multiple definitions of frailty, we sought to demonstrate and compare the validity of 3 frailty measures. Of the 3 measures, the EFS presented multiple advantages over the others. When dichotomized, the EFS identified the highest proportion of frail individuals, when compared with the CFS and grip strength (38% versus 22% and 15%, respectively), and was the only frailty measure that demonstrated significant differences in the individual domains of the LAEP. The EFS had the greatest convergent construct validity, being the only frailty measure to be correlated with all other measures of participant well-being (i.e., anxiety, depression, QOL, and functional disability). When used as a continuous measure, the EFS was correlated with ASM tolerability as measured by the LAEP, after adjusting for age, sex, and the overall comorbidity burden (measured by the ESI). The EFS has been shown to have high sensitivity and specificity for detecting frailty (with the Fried frailty phenotype and the related 5-item FRAIL scale as the gold standard).^45,46 Prior studies have highlighted that an important advantage of the EFS is its multidimensionality because it takes into account several spheres of frailty including cognition, mood, overall physical performance, nutrition, and social status.⁴⁷ This is in stark contrast to the CFS and grip strength, which focus exclusively on physical frailty.

Caution should be exercised when comparing frailty scales. While we found only a weak correlation between grip strength and the other 2 measures, the EFS and CFS were only moderately correlated. This is in contrast to what has been previously found in a cohort of hospitalized patients aged 50 years and older where the same scales were highly correlated and showed significant interscale agreement.⁴⁸ Because each scale measures different domains of frailty using different markers, it certainly should not be assumed that they are all interchangeable. Notably, a longitudinal study comparing 35 scales among 5,000 older adult communities demonstrated a significant interscale variability in their performance.⁴⁷ Our study further illustrates that the scale chosen should depend on the aim of the study and population of interest.

A major strength of this study lies in the use and comparison of 3 clinical measures of frailty. The prevalence of frailty in the same population varies greatly depending on the scale used and on the spheres of frailty it relies upon.⁴⁷ Assessing 3 scales rather than one provides a clearer picture of the association between frailty and the tolerability of ASM. The study is further strengthened by the fact that it was conducted in 4 Canadian centers across 3 provinces, increasing the generalizability of our findings.

Our study is limited, however, by a selection bias that was difficult to control for during recruitment. The 4 recruitment sites are predominantly academic centers and generally follow people with more severe epilepsy. We used a convenience sample, and we do not have complete data on why some people who were approached to participate finally refused. Individuals whose epilepsy is well controlled, likely experiencing fewer side effects from their ASMs, are less likely to be followed in specialist epilepsy clinics and less likely to accept to participate in a study such as ours, as are older, potentially more frail individuals who live in assisted-living residences. Added to this is the fact that people with major cognitive deficits, who would have potentially been considered “very frail,” were excluded from our study because of the impracticality of their completing some of the study questionnaires. Considering that we included people with long-standing epilepsy, there is also a risk of depletion of susceptible bias, where those who remained on their current ASM are those who generally tolerated it. This bias, on the contrary, would likely have decreased the strength of any association between frailty and ASM tolerability, rather than increased it. Our study is further limited by its cross-sectional design. Our results do not allow us to establish a causal link, namely to distinguish whether frailty decreases the tolerability of ASMs in an individual or, conversely, whether frailty could be a consequence of reduced tolerability. We nevertheless believe that the reverse is most likely, that frailty leads to reduced ASM tolerability, given the nature of the domains measured by frailty (e.g., the clock drawing exercise, weight loss, incontinence, grip strength, and the physical aspects measured by the CFS). Last, grip strength could not be measured in every participant. The lack of observed association between grip strength, seizure frequency, and tolerability could be partly driven by the smaller number of individuals. We were unable to impute these data because this would have required our assuming that they were missing completely at random.⁴⁹ We could not exclude the possibility that some participants may have declined a home visit because of frailty in the context of the COVID-19 pandemic.

The EFS was significantly, both statistically and clinically, associated with ASM tolerability. It was found to have multiple advantages in performance while assessing for frailty in older adults with epilepsy, when compared with the 2 other measures of frailty that we tested. Future studies should investigate the potential role the EFS during epilepsy diagnosis may play in ASM selection among older adults with epilepsy. The EFS may be used to allow future trials to target individuals most at risk of poor ASM tolerability.

Glossary

ASMs

antiseizure medications

ADL

activities of daily living

CFS

clinical frailty scale

EFS

Edmonton frail score

ESI

epilepsy-specific comorbidity index

GAD-7

generalized anxiety disorder 7-item scale

LAEP

Liverpool Adverse Events Profile

MD

mean difference

NDDI-E

neurologic disorders depression inventory for epilepsy

QOLIE-10

quality of life in epilepsy 10-item scale

ZINBR

zero-inflated negative binomial regression

Appendix. Authors

Study Funding

This study was made possible partly through a grant from the Canadian Frailty Network (CAT2017-19; with matching funds from Eisai Canada and the Fondation du CHUM).

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/N for full disclosures.