Comparisons of intellectual capacities between mild and classic adult-onset phenotypes of myotonic dystrophy type 1 (DM1)

Myotonic dystrophy type 1 (DM1) is an autosomal dominant genetic disorder with high but incomplete penetrance and the commonest adult-onset form of muscular dystrophy. As a muscle disease, DM1 is characterized by an inability to relax voluntary muscle contractions (myotonia) and by progressive distal to proximal muscle weakness. Associated findings can include involvements in almost every systems such as the cardiac, respiratory, gastrointestinal, endocrine, ocular, and central nervous system (CNS) [1]. DM1 results from the expansion of an unstable trinucleotide cytosine-thymine-guanine (CTG) repeat mutation located in the 3′ untranslated region of a gene (19q13.3) encoding a putative protein kinase (DMPK) [2]. When transcribed into CUG-containing RNA, mutant transcripts aggregate as nuclear foci that sequester RNA-binding proteins, including members of the muscleblind (MBNL) family, resulting in a spliceopathy of downstream effector genes [3]. CTG repeats in DM1 patients can range from 50 to several thousands, with a tendency toward increased repeats with successive generations (anticipation). The length of the CTG repeats is partly correlated to the severity of the disease and the age at onset of symptoms [4]. Considering various European populations, the worldwide prevalence of DM1 could be estimated at 5–20 per 100 000 [1]. In 2010, the highest prevalence reaches 158 per 100 000 in the Saguenay-Lac-Saint-Jean (SLSJ) region located in the eastern part of the province of Quebec (Canada) [5].

Involvement of the CNS occurs in the large majority of patients with DM1 [1],[6],[7], particularly when symptoms appear early in life. This can range from a condition of intellectual disability (a characteristic often associated with congenital DM1 in which symptoms are manifest from birth), to behavioural changes (e.g. reduced initiative, inactivity, apathetic temperament) [1]. Higher cognitive function disabilities are variably impaired [7]–[13] but a trend toward reduced frontal lobe performances has often been reported, along with many significant brain changes (e.g. brain atrophy, cell loss, ventricular enlargement, diffuse white matter lesions as well as significant cerebral blood flow reduction in frontotemporal lobe regions) [1],[14]. Excessive daytime sleepiness is a prominent feature of DM1 and is most often considered as independent from respiratory dysfunction or nocturnal sleep disruption [15],[16]. Moreover, psychopathological disturbances such as avoidant, dependent and paranoid personality traits are frequent in adult form of DM1 [17],[18].

Neuropsychological studies have shown evidences for a generally lower intelligence level in the DM1 population as compared to normal control subjects [13],[14],[19]–[21], with no clear evidence of a progressive intellectual decline over time [7],[8],[20],[22],[23]. The highest prevalence of intellectual impairment has been reported in studies that have included both patients with adult and congenital onset of DM1, even if it is commonly accepted that this latter form is usually characterized by significantly lower intelligence quotients (IQs) [8]. Moreover, classic IQ measures, as obtained by the Wechsler Adult Intelligence Scale-Revised (WAIS-R) [24],[25] may potentially vary with the examinee’s physical abilities. As the degenerative and progressive muscles affection is an important feature of muscular dystrophies, one may argue that lower WAIS-R scores could partly relate to muscular weakness, muscular atrophy or myotonia [25].

The objectives of the present study are thus to describe and compare the intellectual abilities of a large sample of DM1 patients with mild and classic adult-onset phenotypes, to estimate the validity of the WAIS-R for the evaluation of intelligence in DM1 patients with muscular weakness, and to appraise the relationship of IQ to CTG repeat length, age at onset of symptoms, and disease duration. Since the striking increase of individuals with a mild phenotype of DM1 in relationship with the availability of genetic counselling and predictive testing [5], it is relevant to better characterize these milder patients as they will represent a very common phenotype of DM1 population in a near future.

This study evaluated the intellectual abilities of a large sample of DM1 patients with mild and classic adult-onset forms. Reduced intelligence associated with DM1 has been mentioned in many studies including the first descriptions of the disease [36]. Intellectual impairment of different degrees has been described in previous reports but some of them can be criticized for as being biased relatively to the inclusion of congenital DM1 patients or small sample sizes [8]–[10],[12],[37]. The present study offers more reliable estimates of intellectual functioning as we used a large and uniform cohort of DM1 patients with mild and classic adult-onset phenotypes, with the exclusion of any congenital or childhood onset patients. Our results first demonstrate that intelligence level of mild and classic adult-onset DM1 patients is characterized by a low general intellectual functioning in both verbal and nonverbal abilities. For our total sample (n = 188), the mean WAIS-R FSIQ was 82.6, corresponding to low average IQ according to Wechsler [24]. By comparison, a study including 17 classic adult-onset DM1 patients gave similar results on a full WAIS administration, with a FSIQ of 85.3 (VIQ = 87.2; PIQ = 84.1) [19]. Moreover, a recent report [21] on a sample of 121 DM1 patients revealed a mean IQ of 85.0 on an abbreviated WAIS-III, which is also similar to our results. It should nonetheless be mentioned that this latter study included 7.9% of juvenile phenotype. Our findings also contrasted with other previous studies of classic DM1 patients reporting a mean total FSIQ of 94.2 (n = 37) [12], 95.0 (n = 47) [38] or 96.7 (n = 23) [39], which corresponds to average intelligence levels. In addition, one study reported a median FSIQ of 91 on 14 classic adult-onset DM1 patients [40], while another observed a mean FSIQ of 98.8 from a SPM calculation on 50 DM1 patients [41]. Our study does not validate these findings. Regarding the FSIQ, standard deviations ranged from +1.27 to −2.60 SD, and 64.4% (n = 121) had a FSIQ significantly lower (>1 SD) than the normative sample mean. Moreover, almost 5% of the patients obtained an IQ in the intellectual disability range, which is more than twice the percent of the WAIS-R standardization sample. Only one patient had an IQ in the high average range of intellectual functioning.

Studies that measured IQ in mild DM1 patients have generally not detected any defect in global intelligence. For example, a group of 36 late-onset DM1 patients showed a mean SPM IQ of 107.4 [11], while 13 mild form patients obtained a mean IQ of 98.9 (range 83–116) on a two-subtest abbreviated WAIS-R [13]. In the present study, 59.4% of our mild DM1 patients (n = 37) were below the average range of intellectual functioning, namely low average (n = 16, 43.3%), borderline (n = 5, 13.5%), and intellectual disability (n = 1, 2.7%). One could argue that these divergent findings regarding mild phenotype could be explained by differences in levels of fatigue or excessive daytime sleepiness (EDS). To clarify this issue, one of our previous study on the same cohort of patients provide interesting data concerning clinical correlates associated with presence or absence of fatigue and/or EDS [15]. All DM1 participants were categorized into four distinct groups according to their excessive fatigue and EDS level. Post hoc comparisons showed that FSIQ varies neither with excessive fatigue nor EDS status. Another partial explanation may come from the group of 29 patients who did not complete all subtests of the WAIS-R short form. Their FSIQ scores were based on a lower composite reliability estimates than the one of the seven-subtest short form (.91 vs .97 compared to the entire WAIS-R), which led to a significant lower FSIQ. Despite this previous assumption, we believe that this could have a negligible impact on the results, considering the presence of only 7 mild DM1 patients in that group, and an equal proportion of mild and classic adult-onset phenotypes in both groups.

The definition of a mild phenotype of DM1 also varies from one study to another. The original terminology by Koch et al. [42] classified as “mild” patients those with an age at onset over 40 years old. The definition used in the present study took advantage of the deoxyribonucleic acid (DNA) analysis results and is a combination of a small expansion of CTG repeats, the absence of muscular impairment or a late age at onset. From a clinical perspective, the combined use of neuromuscular and genetic criteria allow a more sensitive characterization of these milder patients. Using a definition of mild cases based only on a small expansion of less than 200 CTG repeats, we also observed many patients with FSIQ scores below the average range of intellectual functioning as shown in Figure 2. In all, our results strongly suggest that intellectual impairment is an extremely common and important feature in DM1, not only among the classic adult-onset patients but also among the least severe forms of DM1, with low IQ scores compared to general reference population.

Our DM1 patients also obtained significant low scores on every WAIS-R subtests, with better scores on Digit Span, Arithmetic, and Block Design for the mild group. Moreover, we found a significant relationship between FSIQ and (CTG)n. This can partly explain the fact that the mild group of DM1 patients obtained higher IQ indices. In fact, patients with more CTG repeats have a lower IQ than those with smaller expansions (r _s  = −0.41, p < 0.001). This is consistent with previous studies who noted an association between higher CTG repeat size and some cognitive impairment in classic adult-onset DM1 patients [19],[21],[38]. Such significant negative relationships with FSIQ have also been described in severe and mild congenital DM1 children [43] and childhood [43],[44] forms of DM1. In combination with these reports, our study on classic adult-onset and mild DM1 portrays the natural course of this complex multisystemic neuromuscular disorder throughout a clinical continuum in its CNS manifestations. We believe that these results are further evidence of an existing overlap between the different phenotypes in DM1. From this point of view, higher IQs and better scores on some subtests in favour of the mild group of patients could rather reflect a later onset of the disease than a complete distinct cognitive phenotype. We also found that FSIQ was not linked to the age at onset of symptoms. For patients affected by DM1, the symptoms reported at onset are usually myotonia, distal weakness or both [45]. In the present study, the absence of such correlation suggests that there is no close relationship between the muscular involvement and the CNS impairment as measured by the FSIQ. This suggestion of two separate processes in DM1 evolution has also been recently described in a study that analysed cerebrospinal fluid levels of three biomarkers in patients with DM1, namely total tau (T-tau), phosphorylated tau (P-tau) and 42-amino-acid form of ß-amyloid (Aß ₄₂). Although the Aß ₄₂/P-tau ratio was decreased in classic adult-onset DM1 patients, along with a tendency of increased T-tau and P-tau levels, the authors did not find any correlation between these biomarkers and disease duration or severity of muscular involvement [46]. DM1 is a progressive disease affecting the muscles, but the question if CNS dysfunction in DM1 is also a progressive condition remains by cons uncertain. It was suggested that progressive CNS dysfunction could potentially account for cognitive decline rather than a true neurodegenerative process per se [47]. The present results showed that FSIQ scores significantly decreased with the increase of the disease duration. This finding may support the hypothesis of a certain global decline in cognitive abilities in DM1 over time but needs to be confirmed by longitudinal designs. A follow-up study (mean follow-up: 7.3 ± 2.7 years) on 20 classic adult-onset DM1 patients showed a specific and progressive impairment in attentional capacities, without extension to additional areas of cognition [22]. Over a four-year period, another group of 34 DM1 patients (including 2 congenital forms and 32 classical forms with juvenile or classic adult-onset) showed a significant deterioration in linguistic functions, with a tendency towards decline in executive abilities, confirming a predominant involvement of cognitive functions subserved by fronto-temporal areas of the brain [23].

It is well acknowledged that educational attainment is highly correlated to psychometric measures of intelligence [24]. In this context, as IQs between mild and classic adult-onset DM1 patients are quite similar, it is not surprising to observe that years of education are the same for both groups. One of our previous study also demonstrated a relationship between increased (CTG)n and higher risk of material and social deprivation [27]. Thus, classic adult-onset DM1 patients are more likely to show higher unemployment, lower family income, and higher reliance on social assistance compared with mild DM1 patients. One could argue that this discrepancy of the social impact between mild and classic adult-onset DM1 patients is probably explained by the lower motor skills of the latter. Since classic adult-onset DM1 patients obtained the lowest IQs, with a high frequency in the intellectual disability range, it is reasonable to believe that intellectual impairment is an important contributing factor in the socioeconomic deprivation portrayed in patients with DM1 [27]. From our clinical experience, additional factors such as daytime sleepiness, fatigue, personality traits, apathy or other cognitive deficits could also contribute to the vulnerability of these patients in experiencing socioeconomic inequalities.

Intellectual evaluation, as for any neuropsychological investigation, depends in some way on motor abilities, which may be impaired in DM1. We used SPM as a motor-independent test of intelligence to assess muscular impairment as a possible confounding factor in the establishment of WAIS-R IQ indices. We found a strong correlation between total WAIS-R and SPM FSIQ estimation, both for the entire sample and within each subgroup of DM1 patients. Moreover, our results showed a discrepancy in favour of a significant higher mean PIQ score compared to VIQ, added to a more frequent PIQ > VIQ pattern. These results altogether suggest that the WAIS-R is a valid instrument to evaluate global intelligence in mild and classic adult-onset DM1 patients, on the one hand, and that muscular weakness in DM1, particularly of the upper limbs, relates not to intellectual impairment, on the other. Moreover, our results are in agreement with a previous study showing that both verbal and nonverbal WAIS subtests were consistently lower in a group of classic adult-onset DM1 patients compared to normal and neurological controls such as spinal muscle atrophy, suggesting that cognitive impairment is not a consequence of motor impairment [19].

This study has some limitations that should be addressed in future researches. First, assessment of adaptive skills was not included. In addition to better characterize our DM1 patients with very poor IQ scores, a complete assessment of the adaptive behavior would have been useful in determining how well these individuals respond to daily demands from the environment. Also, our study does not include unaffected family members of the households of the patient or case controls. Despite that our results are compared to population-based norms, the inclusion of a family member unaffected by DM1 could have clarified if low IQ scores are linked to the disease or if they are a consequence of any socioeconomic deprivation previously reported in DM1 population [27]. In our study, IQ scores were obtained using the French adaptation of the WAIS-R [30], although this version is not validated in a French Canadian population. A French European standardization sample exists but we decided to use US norms since we consider that these data are more likely to represent the intellectual levels of North-American adults. A study who compared the French and American WAIS-R standardization samples showed that the two are very similar, except for two variables. Education level is slightly lower for the French subjects, while the French sample included an age group (75–79 years) that is missing from the American sample, which leads to give a greater weight to the variable Age [48]. Finally, whenever possible, future studies in DM1 could correlate their neuropsychological data to neuroimaging or functional investigations such as PET (positron emission tomography) or SPECT (single-photon emission computed tomography).

The impact of DM1 on brain and cognition is well described. It is generally acknowledged that intelligence is lower in this population as compared to normal control subjects, and that intellectual disability is characteristic of the congenital and childhood onset forms of the disease. Our findings showed that intellectual impairment is an extremely common and important feature of DM1, and could also be observed in individuals mildly affected by DM1, with low IQ scores compared to the general reference population. Since the introduction of predictive DNA testing, the proportion of patients with a mild phenotype is an increasing phenomenon in DM1 [5], and health care providers involved in the follow-up of these patients should be aware of their intellectual capacities and should adapt their interventions accordingly.