Disease burden of myotonic dystrophy type 1

To the best of our knowledge, this study represents the first comprehensive examination of the disease burden of DM1 in the UK as recorded though the INQoL. Our results show that the vast majority of affected patients suffer from some degree of muscle weakness, muscle locking, pain, fatigue, and difficulties with swallowing, and that these symptoms collectively have a non-trivial impact on their ability to perform activities of daily living. Indeed, in our sample, 23% of patients indicated that their leisure activities were considerably, very much, or extremely affected by DM1, 47% did not work due to their illness, 33% reported requiring a considerable amount, very much, or an extreme amount of assistance, and 52% stated that their illness caused difficulties in their relationship with their partner/spouse. Our data also show that DM1 is often associated with impairment in psychological well-being, including anxiety, depressive feelings, frustration, and low confidence/self-esteem. Taken together, these findings underscore the detrimental impact of DM1 on affected patients and highlight the substantial unmet medical need in this disease population.

In line with expectations and previous research [8‐11], our INQoL results show that muscle weakness and fatigue constitute two of the most common manifestations of DM1, affecting the vast majority of patients. In addition, these symptoms were also found to be associated with the most difficulties to the patients. We identified no significant differences concerning the crude prevalence of muscle weakness by sex, but did find, in contrast to previous research [12] that a significantly larger proportion of women (48% vs. 18%) experienced a considerable amount, a lot, or an extreme amount of fatigue.

Considering emotional well-being in DM1, most patients in our sample reported feeling anxious/worried, depressed, and/or frustrated to some degree, and many also indicated low confidence/self-esteem. Data indicating impaired mental health in DM1 have been previously reported [8‐10, 13‐16]. Interestingly, in the present study, we found evidence suggesting that women with DM1 were much more likely to report depressive feelings than men, a result also noted in previous research [8]. Yet, as discussed by Peric et al. [15], it is not known if these psychological traits constitute symptoms or manifestations of DM1 (i.e. as part of the underlying pathological process), or if they are developed as a consequence of living with the disease, or a combination of both. Nonetheless, given the prevalence observed in our study, mental health in patients with DM1 warrants further exploration.

Previous studies [8‐10, 17] have estimated mean INQoL muscle weakness scores at between 49 and 70, muscle locking 38 and 63, pain 25 and 42, fatigue 42 and 60, activities 35 and 55, independence 30 and 47, social relationships 10 and 23, emotions 24 and 44, body image 6 and 62, and total score 19 and 52. Our estimates for the total sample all lie within these ranges, with the exception of social relationships [26], indicating a relatively higher burden on this aspect of life in our cohort. However, given the substantial variability in estimates (also between patients within samples), and considering non-trivial differences with respect to demographic and clinical characteristics of patients across studies, it is likely not meaningful to further compare our results. This is also related to the psychometric properties of the INQoL, where the Likert scoring algorithm would be expected to fail to adhere to the criteria of fundamental measurement, which is a requirement for invariant comparison [18, 19]. Put differently, although it is clear that a lower/higher score indicates a lower/higher disease burden, it is not clear how a specific score, or differences in scores at various points across the scale continuum should be interpreted. Indeed, the high association between DM1 symptoms and amount and importance of difficulties noted as part of this study suggest that combining Likert scores for these items may not be appropriate due to non-trivial item dependency. The ordinal nature of the INQoL is also likely to affect the possibility to detect and interpret changes in total scores over time. For example, despite worsening Muscular Impairment Rating Scale (MIRS) outcomes, Peric et al. [17] found that total INQoL scores, as well as several subscale scores, improved over the course of 6 years. Moreover, in a recent multicentre, single-blind, randomised trial, Okkersen et al. [20] identified no significant changes in total INQoL scores at 10 months from baseline between patients with DM1 receiving cognitive behavioural therapy plus standard care and optional graded exercise and those receiving standard of care alone, despite significant changes in the main endpoint (i.e. the DM1 activity and participation scale for clinical use [DM1-Activ^C]) and other secondary endpoints. Additionally, Hamilton et al. [21] recently reported some inverse correlations between objective measures of disease severity and patient self-reported outcomes, suggesting disease-dependent lack of insight can sometimes mask physical symptoms.

The data presented as part of this study indicate that male patients may experience greater impairment in their relationships life compared to their female counterparts. Specifically, a significantly larger proportion of men (70% vs. 40%) reported of difficulties with their friends due to DM1, and men with the disease were much less likely to be married or in a relationship (36% vs. 70%). Evidence of a relatively larger impact on the social life of male patients has been reported for DM1 [22], as well as other chronic, disabling diseases, such as multiple sclerosis [23], and warrants further investigation.

Previous research [24‐26] have demonstrated an association between CTG repeat length and age at onset in DM1. Such age at onset associations have been further improved by estimating the progenitor CTG allele length to overcome the confounding effects of age-dependent somatic expansion [7]. Previous attempts to correlate CTG length to progressive phenotypes have typically yielded poorer associations. Indeed, Antonini et al. [13] and Rakocevic-Stojanovic et al. [10] found no association between quality of life (as measured through the SF-36 and INQoL, respectively) and CTG repeat length. Here, we have estimated the progenitor CTG allele length to overcome the confounding effects of age-dependent somatic expansion and included an age-CTG interaction to account for the progressive nature of the symptoms and the inherent sampling biases found within DM1 families (due to the profound anticipation observed in DM1, older sampled individuals inherited smaller alleles, and vice versa). These analyses have yielded good correlations for overall INQoL score (R² = 0.19) and for a number of the sub-domains: muscle weakness (R² = 0.21), fatigue (R² = 0.09), muscle locking (R² = 0.09), activities (R² = 0.18), independence (R² = 0.23), relationships (R² = 0.10) and emotions (R² = 0.12). These data highlight the utility of careful genetic and statistical analyses that addresses the sampling biases inherent within DM1 populations. Outcome measures that are predicted by underlying genetic factors more likely to reflect primary disease processes and thus are more likely to respond positively to effective therapeutic interventions. Conversely, the lack of association with genotype for the pain, muscle locking, drooping, double vision, and body image scores suggests that these domains are not sensitive reporters of the disease process. Genotype to phenotype correlations such as these may therefore be of utility in selecting the most informative outcome measures for patient management and clinical trials.

Previous studies have also identified unexpected distributions of INQoL scores across measures of disease severity and age. For example, Peric et al. [17, 27] found that worsening of muscular weakness from mild to severe was associated with a lower disease burden (i.e. higher quality of life). In addition to the sampling biases discussed above, potential reasons for these counterintuitive findings include lack of disease awareness, in particular in patients with less severe disease [8], coping mechanism [28] (which has been identified for other rare, neuromuscular diseases, such as Duchenne muscular dystrophy [29]), and the properties of the scale, as discussed above.

Our study is subject to a few limitations. First, similar to most research of patients with rare diseases, our results are limited by a relatively small sample, which affects the precision and potentially also generalizability of derived estimates. This would be expected to be of particular relevance considering the heterogeneous presentation of DM1. Additionally, given the underlying variability in outcomes of interest and the magnitude of expected differences, conducting research in small patient populations may also have a negative impact on possibilities to perform meaningful evaluation of results across subgroups. For this reason, outcomes of our testing of differences by, for example, sex should be interpreted with some caution. Second, our self-reported data may be subjected to bias due to, e.g. incorrect reporting. Third, due to the observational nature of our data, we were unable to draw conclusions regarding causality. Concerning interpretation of results, it is also worth pointing out that we were unable to adjust our estimates for patients’ intelligence quotient (a variable that may have an impact on how patients perceive, and are able to report of, their disease burden), and that none of the patients in our sample received cognitive behavioural or exercise therapy (which recently have proven beneficial for the capacity for activity and social participation in DM1 [20]).