Introduction

Becker and Duchenne muscular dystrophies (BMD and DMD) are two neuromuscular diseases caused by several mutations, including point mutations, deletions, or duplications, in the dystrophin gene [1, 2], affecting one in 18,000–30,000 male births [3] and one in 3500–9000 male births [4], respectively. In BMD, these mutations do not affect the reading frame, allowing for partially altered but functional dystrophin expression. However, in DMD, the reading frame is affected, with an absolute or almost absolute absence of dystrophin. Alterations of dystrophin lead to sarcolemmal instability, muscle degeneration, and replacement by fibrotic tissue, with progressive muscle weakness and reduced ambulation ability. BMD progression is usually slow, with ambulatory ability in adulthood and a normal life expectancy [2, 5]. Conversely, and despite treatment optimization, the emergence of new treatments, and phenotype variability, DMD continues to have a poor prognosis, with loss of ambulation in adolescence and life expectancy into the third or fourth decade of life [5,6,7,8,9].

In addition to muscle, dystrophin is also expressed in the brain as full-length isoforms (Dp427c and Dp427p) as well as short isoforms (Dp140 and Dp71). These isoforms are expressed in the hippocampus, the Purkinje cells of the cerebellum, and the amygdala, with important roles in GABAergic transmission (Dp427), fetal development (Dp140), glutaminergic transmission and blood–brain barrier formation (Dp71). Depending on the mutation site, one or more isoforms will be affected. Thus, an association of BMD and DMD has been observed with several neuropsychiatric and neurological disorders, including epilepsy [10,11,12,13,14]. In the case of epilepsy, some mechanisms have been proposed that could explain the increased risk associated with BMD/DMD. The absence of Dp427 produces an aberrant clustering of GABAA receptors, with an increase in extrasynaptic GABAA receptors. All this leads to a dysfunctional GABA inhibitory pathway, increasing the spread of seizures. Moreover, Dp427 has been shown to be upregulated in temporal lobe epilepsy in people without dystrophy, indicating a possible compensatory mechanism. In BMD/DMD, this mechanism could be lost, lowering the seizure threshold and increasing the risk of epilepsy. The absence of Dp71 could also contribute to neuronal hyperexcitability through altered potassium and water homeostasis due to alterations in Kir4.1 and aquaporin-4 clustering in astrocytic/astroglial terminal feet, leading to increased potassium levels. Finally, the absence of Dp71 causes important alterations in the integrity of the blood–brain barrier, exposing the brain to substances that cause hyperexcitability (e.g., albumin) [15, 16].

Although epilepsy is known to be common in dystrophinopathies, the overall prevalence in populations with BMD/DMD is unknown. It is also possible that the genotype in BMD/DMD (i.e., which isoforms are affected) influences the prevalence of epilepsy, as occurs in other neuropsychiatric and neurodevelopmental disorders in these patients [14]. Therefore, this systematic review and meta-analysis aims to estimate the prevalence of epilepsy in populations with BMD or DMD and to determine whether there is a genotype–prevalence association considering the mutation site.

Methods

We conducted this systematic review and meta-analysis according to the Meta-analyses of Observational Studies in Epidemiology (MOOSE) guidelines [17] and the Cochrane Collaboration Handbook [18]. Moreover, the study was registered in PROSPERO (Registration Number: CRD42021284753).

Search strategy

The Medline, Scopus, Web of Science, and Cochrane Library databases were systematically searched for relevant studies published from inception to January 2022. Grey literature, including the Theseo, OpenGrey, Google Scholar, and Networked Digital Library of Theses and Dissertations databases, was searched. The references of the included studies were also reviewed. Search terms included prevalence, incidence, survey, questionnaire, frequency, epidemiology, ratio, rate, dystroph*, “duchenne muscular dystrophy”, “becker muscular dystrophy”, duchenne, dmd, bmd, neurological disorder*, neurological disease*, epilepsy, seizure, with appropriate Boolean operators. Study authors were contacted if necessary. The complete search strategy is included in Supplementary Appendix S1. The literature search was conducted independently by two reviewers (CP-M and IC-R), and disagreements were resolved by consensus or by a third reviewer (VM-V).

Inclusion/exclusion criteria

The inclusion criteria were as follows: (1) participants: study populations with BMD or DMD (no age restrictions); (2) design: observational studies, including cross-sectional, retrospective, prospective, or case series studies; and (3) outcomes: (i) prevalence of epilepsy (main outcome), and (ii) genotype–prevalence association (secondary outcome). Mutations downstream of exon 45 were considered Dp140−, while mutations downstream of exon 63 were considered Dp71−. There were no language restrictions.

Exclusion criteria were as follows: (1) participants: studies including participants with other dystrophies, studies unable to determine the prevalence of epilepsy in patients with BMD or DMD; (2) design: studies with an insufficient sample to obtain a proportion (e.g., single case studies); (3) outcome: cohorts of only patients with epilepsy, studies reporting only febrile seizures.

Data extraction

An ad hoc table was performed with the data extracted from the included studies. It included (1) reference (authors and year of publication); (2) country; (3) design (cross-sectional, prospective, retrospective, case series); (4) population (disease, sample, age); (5) outcome (prevalence of epilepsy, genotype–epilepsy association).

Risk of bias assessment

We assessed the risks of bias of the included studies with the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies of the United States National Institutes of Health National Heart, Lung, and Blood Institute [19]. This tool consists of 14 domains evaluating different aspects of study design and statistical methods and studies are scored as good, fair or poor. Considering the number of domains with risks of bias, the overall bias was good, fair, or poor.

Risk of bias assessment was conducted independently by two reviewers (CP-M and IC-R), and disagreements were resolved by consensus or with a third reviewer (VM-V).

Data synthesis

A narrative synthesis and ad hoc table of each study was performed for the main outcome (i.e., prevalence of epilepsy) and for the secondary outcome (genotype–prevalence association). The prevalence and its 95% confidence interval are shown, with 0.00 indicating 0% prevalence and 1.00 indicating 100% prevalence. The genotype–prevalence association was expressed, when possible, as a prevalence ratio and 95% confidence interval. When not possible, the percentage of patients with Dp140 or Dp71 present or absent in the population of participants with epilepsy was described. Occasionally, the site of the mutation was also considered.

A random effects meta-analysis [20, 21] was performed to determine the prevalence of epilepsy according to population subgroup (i.e., BMD, DMD, BMD + DMD). The overall prevalence of epilepsy in all participants with these dystrophinopathies was also estimated. Heterogeneity (I2) was evaluated, and it was considered not important if it was < 30%, moderate if it was 30–50%, substantial if it was 50–75%, and considerable if it was > 75% [18, 22]. The p value of the heterogeneity was also considered. Finally, publication bias was assessed visually using the funnel plot [23]. A sensitivity analysis with study-by-study exclusion was conducted to determine the possible influence of studies with estimates that distort the real estimate.

All statistical analyses were performed with the statistical program STATA SE, version 15 (StataCorp, College Station, TX, US). Moreover, the statistical package metaprop, available for STATA, was used to estimate prevalence.

Results

Of the 4154 studies identified, eight met the inclusion/exclusion criteria (Fig. 1) [24,25,26,27,28,29,30,31] and were included in the systematic review (Table 1) and meta-analysis, including three studies on BMD, five on DMD, and two on BMD + DMD. Twelve studies were excluded with reasons (Supplementary Table S1).

Fig. 1
figure 1

PRSIMA flowchart of study selection

Full size image
Table 1 Baseline characteristics of the participants in the included studies
Full size table

There were two studies of participants from Italy [29, 29], two from Japan [26, 31], two from the United Kingdom [24, 28], two from the United States [25, 28], and one each from Australia [28], Belgium [28], China [30], Ireland [28], Iran [27], and the Netherlands [28]. The systematic review and meta-analysis included 1536 participants, of whom 157 had BMD, 872 had DMD, and 507 had either BMD or DMD. The sample size of the studies ranged from 28 to 307 participants. However, only two studies reported the mean age of the participants, which was lower in DMD participants than in BMD participants. Eight studies determined the prevalence of epilepsy, four reported the genotype of patients with epilepsy, and of these, only two reported the genotype of participants with and without epilepsy (a necessary condition for calculating the prevalence ratio). Finally, in three studies data were extracted from medical records, and in five studies from interviews or questionnaires. In half of the studies the type of epilepsy of the patients and whether or not they were well controlled were available. The method of data collection and the type of epilepsy diagnosed in each study are included in Supplementary Table S2.

Systematic review

Table 2 shows the findings of the prevalence of epilepsy in this population, and Table 3 and Fig. 2 show the genotype–prevalence association.

Table 2 Prevalence of epileptic disorders in each study
Full size table
Table 3 Association of epileptic disorders with the genotype of the participants
Full size table
Fig. 2
figure 2

Forest plot of the prevalence ratio and 95% confidence interval by genotype-outcome of each study

Full size image

In BMD participants, the prevalence of epilepsy ranged from 0.07 to 0.08, while in DMD participants, it ranged from 0.02 to 0.13. In the two studies that did not separate BMD and DMD participants, the prevalence ranged from 0.02 to 0.06. There was no statistically significant difference in prevalence between BMD, DMD, and BMD + DMD groups. Regarding the genotype–prevalence association, in BMD, a PR = 0.48 (0.03, 8.71) and PR = 0.46 (0.03, 7.44) were observed for Dp140+ vs. Dp140− and for Dp71+ vs. Dp71−, respectively [26]. Conversely, in DMD, a PR = 0.50 (0.11, 2.15) and a PR = 1.07 (0.07, 16.46) were observed for upstream mutations exon 31 vs. 31–62 and for Dp71+ vs. Dp71− [29]. Finally, in a study with a relatively large sample of participants with epilepsy with a known genotype [28], there were more participants with Dp140− than with Dp140+, although the prevalence ratio could not be estimated.

Risk of bias assessment

According to the Quality Assessment Tool for Observational Cohorts and Cross-Sectional Studies from Study Quality Assessment Tools, the overall bias was scored as good in all studies. By domain, a risk of bias was only detected in 1 out of 8 studies (12.5%) regarding the number of participants who eventually participated in the study. Three domains were not eligible for assessment, based on sample size justification, whether exposure was assessed more than once over time, and whether the outcome assessor was blinded. No significant risk of bias was observed in the remaining domains. The risk of bias assessment is described in Supplementary Table S3.

Meta-analysis

The pooled prevalence showed the following estimates. In BMD, 0.07 (0.03, 0.11); in DMD, 0.05 (0.02, 0.08); and in BMD + DMD, 0.03 (0.02, 0.05). The overall prevalence of epilepsy in BMD/DMD dystrophies was 0.05 (0.03, 0.07). There were no statistically significant differences between the groups or between the groups and the overall prevalence (Fig. 3).

Fig. 3
figure 3

Meta-analysis of prevalence of epilepsy in Becker muscular dystrophy, Duchenne muscular dystrophy, and Becker muscular dystrophy/Duchenne muscular dystrophy

Full size image

Heterogeneity was substantial in DMD and in overall prevalence, with I2 = 73.79% (p = 0.00) and I2 = 60.37% (p = 0.01), respectively (Fig. 3). Heterogeneity in BMD and BMD + DMD could not be calculated. However, in BMD, a high homogeneity was observed between the results obtained, while in BMD + DMD, a moderate inconsistency could be observed. Finally, the funnel plot showed no publication bias (Supplementary Figure S1). Sensitivity analyses with study-by-study exclusion did not show a statistically significant effect.

Discussion

Main findings

This systematic review and meta-analysis provide an estimate of the prevalence of epilepsy in the population with BMD and/or DMD. The prevalence was 7% in BMD participants and 5% in DMD participants, with an overall prevalence of 5% among participants with Becker/Duchenne dystrophinopathies. Since the inconsistency of the individual results was not especially high, it was confirmed that there is a particularly high prevalence of this disorder in this population, higher than that observed in the general population. Finally, although a deleterious effect cannot be ruled out when Dp140 is absent in DMD, the data are inconclusive, and further research is needed.

Interpretation

Our findings showed a prevalence of epilepsy of 7% in BMD, 5% in DMD, and 3% in BMD + DMD participants. These results were higher than the prevalence in general cohorts, which is estimated at 0.5–1.0% in children and adolescents and 0.5–1.6% in the total population [32,33,34,35,36,37]. Moreover, some authors [26, 28] suggest that the prevalence could be even higher due to the designs of their studies, which included interviews without questions for parents [26], or because some patients who reported epilepsy that could not be verified with clinical data were excluded from the prevalence estimates [28]. However, the data from these studies were consistent with the estimate from our meta-analysis. Otherwise, some variability in the results was observed, which could be due to small sample size (e.g., the study with the highest prevalence had a relatively small sample size) [27], selection bias, recall bias, or study design (i.e., data collection).

Regarding the genotype–epilepsy prevalence association, the available data are very limited, and therefore, no clear conclusions can be drawn. On the basis of the two studies with available prevalence ratios, a harmful trend was observed for deleterious Dp140 in BMD [26] and for mutations between exons 31–62 in DMD [29]. Interestingly, a mechanism by which the alterations of Dp140 might increase the risk of epilepsy has not yet been proposed since Dp140 mainly influences fetal development [38,39,40]. Conversely, no association with epilepsy has been found for Dp71 [26, 29], in which there might be a justifying mechanism, as previously proposed [15, 16]. With the present results, the most plausible hypothesis is that Dp427 is the most relevant isoform in the development of epileptic disorders. However, to determine the roles and possible influences of Dp140 and Dp71, more research is needed.

Our findings have some direct and indirect implications for the care of these patients and for research into these diseases. First, because of the high prevalence of epilepsy in BMD/DMD, a high use of anticonvulsant treatments is expected, which together with glucocorticoids may further increase the risk of bone demineralization, requiring special monitoring [41, 42]. Second, there is an association between epilepsy and neuropsychiatric disorders, including autism spectrum disorder, attention deficit hyperactivity disorder, and emotional disorders, and this is an aspect that should be considered in the evaluation of these patients [28, 43,44,45,46,47]. Third, it is suggested that some types of epilepsy (i.e., absences) may be underdiagnosed. Some studies [28, 29] showed that 29–39% of epileptic episodes are “absences”. Interestingly, in one study [29], two of the four patients who were diagnosed as absences did so after EEG following parent-reported staring episodes, which they misinterpreted as a sign of poor attention. Conversely, other studies did not report patients with absences, either because they did not include them or because they did not differentiate the type of epilepsy. Therefore, epilepsy should be actively sought by the neurologist or pediatrician treating patients with BMD/DMD. Fourth, the included studies that reported information showed partially or totally refractory epilepsy in 14–38% of patients, which poses a challenge for the management of these patients [24, 28, 29]. Fifth, the future development of treatments that cross the blood–brain barrier and that restore dystrophin expression at the brain level could [48], theoretically, improve the control of epilepsy in DMD. Fifth, although dystrophin is produced in BMD, the prevalence of epilepsy in BMD is similar to that in DMD, which might indicate that it is not only the presence of full-length dystrophin that is essential to avoid developing this disorder but that the correct length and structure is also required.

Limitations

Some limitations of this study should be considered. First, the limited number of studies and participants for BMD could limit the validity of the results for this group of patients. Moreover, only four studies considered epilepsy as the main outcome, and only two studies included genotype–prevalence association. This especially limited the interpretation of the results of prevalence ratios. Second, there was substantial heterogeneity in the results of DMD studies. Although the results of the sensitivity analyses were satisfactory, the estimates of the meta-analysis should be considered with caution. Third, due to the selection criteria and nonresponses in observational studies, selection or recall bias cannot be ruled out, which could slightly affect the individual estimates obtained. Fourth, to observe and correctly interpret the associations of interest, a large national database with diagnosis-based codes may be a better method than a cross-sectional cohort of 300 patients at most, or retrospective series. Fifth, some seizure disorders (i.e., absences) may be underdiagnosed due to confusion of some signs with other neuropsychiatric disorders.

Conclusions

BMD and DMD are disorders strongly associated with a higher prevalence of epilepsy compared to the general population, with a prevalence of 5–7%, although some authors suggest that it could be somewhat higher. However, genotype was not clearly associated with epilepsy risk, although further studies are needed because of the detrimental trend observed for patients with Dp140− and because of the notional risk of Dp71−. The possible relationship between epilepsy and other neuropsychiatric disorders, such as autism spectrum disorder, attention deficit hyperactivity disorder, and emotional disorders, which occur in patients with epilepsy but without BMD/DMD, awaits further research. Finally, the use of anticonvulsants in this population could affect bone quality, requiring strict medical control.