Keywords

Adults, Spinal muscular atrophy, Morbidity, Clinical epidemiology

Introduction

Spinal muscular atrophy (SMA) is a rare neuromuscular condition characterized by a group of hereditary, progressive motor neuron destructive diseases [1]. SMA encompasses varied severities, which is often clinically classified based on the timing of onset of motor dysfunction. For example, SMA type I is the most common form (approximately half of SMA patients) and hypotonia, delayed motor function milestones, and other motor neuron related issues (e.g., feeding, respiration) present within the first 6-months of life. [2] SMA type II (~30% of SMA patients) and type III (~10% of SMA patients) are less severe forms, present as never acquiring the ability to walk (type II) or losing the ability to walk in childhood (type III), and many survive into adulthood [3]. SMA type IV is the least common (~5% of SMA patients) and the onset of functional decline begins in the adult years [4,5].

What links all individuals with SMA is some degree of motor dysfunction in the lifespan, which can increase the risk for a variety of morbidities across physiological systems, especially with aging. Moreover, the disease severity of SMA correlates with the functional protein levels of the survival motor neuron [6,7]. This protein can be expressed in tissue outside of the spinal cord, suggesting the possibility for a direct pathological role of SMA on non-neuromuscular physiologic systems [8]. To date, there is little research examining the onset of morbidities across the adult lifespan for individuals with SMA [9]. This is mainly because the majority of SMA-related research has focused on elucidating the molecular mechanisms of SMA and developing groundbreaking disease-modifying therapies. Beginning in December 2016, several therapies for SMA have been approved by the U.S. Food and Drug Administration and have shown efficacy in reducing mortality and the need for mechanical ventilation and improving motor function [5,10-16]. However, there may still be deficits in motor abilities, especially if therapy is initiated after the disease has become symptomatic. Thus, identifying morbidity profiles across physiologic systems as individuals with SMA age into and throughout their adult years is necessary to assist in prioritizing clinical care to manage the person with SMA as opposed to the physiologic system SMA primary impacts.

The primary objective of this study was to identify the risk of morbidities across several physiologic systems for adults with SMA as compared to the general population of adults without SMA. This study used real-world clinical data in the “pre-treatment era” prior to the first approved disease-modifying therapy to establish a “baseline” of morbidity profiles for adults with SMA. It was hypothesized that adults with SMA would have a higher risk of all examined morbidities, and that the risk of many morbidities would be elevated earlier in the adult lifespan compared to adults without SMA.

Methods

Data sources

This was a retrospective cohort study that accessed administrative claims from the Medicare fee-for-service (20% random sample) database Part A and B (hospital and medical insurance) and Optum’s de-identified Clinformatics ^® Data Mart Database. This study did not access Medicare’s Part C or D (Medicare Advantage Plan and prescription drug coverage). Patient-level claims from 01/01/2008-12/22/2016 were ascertained to examine outcomes prior to the first U.S. Food and Drug Administration approved disease-modifying therapy for SMA. Medicare is a federal program that provides medical insurance to individuals ≥ 65-years-old with or without disabilities and with certain disabilities if younger than 65-years-old, including SMA. Optum is a national database of individuals with commercial medical insurance or with a Medicare Advantage Plan. The cohort with SMA was obtained from the Medicare and Optum database. As Optum provides greater representation of adults < 65-years-old compared to Medicare and reasonable representation of adults ≥ 65-years-old, the cohort without SMA was obtained from the Optum database to provide a representative background population across the adult lifespan.

Administrative claims are used for billing reimbursement of healthcare services. Researchers can identify medical conditions, via clinician-based diagnoses, by searching patient-level claims for unique codes attached to the claim. All medical conditions examined in this study were identified by International Classification of Diseases, Ninth or Tenth Revision, Clinical Modification (ICD-9/ICD-10) codes, which are listed in Supplementary Table S1. The University’s Institutional Review Board approved this study as non-regulated as the data are de-identified prior to administering to researchers.

Cohort selection

A flow chart of inclusion/exclusion steps is presented in Figure 1. Adults ≥ 18-years-old with SMA were identified in the Medicare database as ≥ 1 inpatient claim for SMA or ≥ 2 outpatient claims for SMA, where each outpatient claim was on a separate day within 12-months of one another. This is consistent with recommendations by the Chronic Conditions Data Warehouse. For Optum, diagnoses were not separated by inpatient or outpatient services due to differences in the underlying data structure. Therefore, the algorithm of ≥ 2 claims for SMA described above was used regardless of where the claim originated.

To be included in the analysis, adults had ≥ 2-years of continuous or mostly continuous enrollment in a health plan between 01/01/2008-12/22/2016. It is not uncommon for insurance enrollment periods to have a short break for reasons unrelated to the outcomes and conclusions drawn from a study like this. Therefore, this study allowed for up to 3 breaks in enrollment, where each break was < 30 days and separated by ≥ 6-months of continuous enrollment. This approach helps to balance detection and selection biases to draw meaningful conclusions regarding morbidity prevalence across a 2-year period. The majority of the cohort with SMA in this study had an unbroken 2-year enrollment (93.7%), while 5.8% had 1 break, 0.5% had 2 breaks, and 0% had 3 breaks in enrollment.

The start and end of the 2-year period was centered around the first SMA claim date as much as possible. This method was employed as the diagnostic codes for SMA do not distinguish between the types of SMA. Other research teams have used the age at the first diagnostic claim for SMA, but this was in pediatrics [17] and the claims data available in this study do not include information from birth. Although, we anticipate that the majority of the cohort will have pediatric-onset SMA (types I-III) as opposed to adult-onset SMA (type IV), given that the latter contributes to ~5% of the population with SMA [5]. For those with > 1-year of enrollment prior to and after the first SMA claim date, the start of the 2-year period (to examine morbidity prevalence) was 1-year prior to the first SMA claim date and the end date was 2-years after the start date. For those with < 1-year of enrollment prior to their first SMA claim date, the first enrollment date was the start date and the end date was 2-year after. For those with < 1-year of enrollment after the first SMA claim date, the end date was the last available date and the start date was 2-years prior.

Adults with 0 claims for SMA were considered for inclusion in the cohort without SMA, and their start and end dates were randomly assigned within 01/01/2008-12/22/2016. To enhance interpretation of crude prevalence estimates, adults without SMA were matched (without replacement) to each adult with SMA at a 1:200 ratio (case:control) for age (± 2 years), sex, U.S. region of residence, race (as white vs. non-white due to data limitations), and study entry year.

Morbidities

Selection of morbidities was guided by established and validated comorbidity indices using claims, including the Charlson [18,19], Elixhauser [20], and Whitney [21,22] Comorbidity Index. The Whitney Comorbidity Index was developed for adults with cerebral palsy and is more relevant to capture morbidity and multi-morbidity profiles for populations with neurological conditions than other commonly used indices. Collectively, there were 30 morbidities examined across physiological systems, such as cardiovascular, metabolic, respiratory, musculoskeletal, gastrointestinal, liver, kidney, and cancer, as well as mental health disorders. A multi-morbidity score was calculated as the sum of the 30 morbidities, which provides some form of evidence of medical complexity in terms of the number of physiologic systems impacted by morbidities.

For consistency across databases, the presence of each morbidity was defined by ≥ 2 claims (inpatient or outpatient, any position), where each claim was on a separate day within the 2-year period. The sensitivity and specificity of the examined morbidities ranges from fair to excellent. It is assumed that the ability to detect each morbidity does not differ by cohort as there is no evidence to suggest differential detection. Therefore, the relative effect estimate of each morbidity comparing adults with vs. without SMA is assumed to be more representative of the true relative risk.

Characteristics

Age determined based on the study start date, sex, race, and U.S. region of residence were retrieved. To further characterize the Medicare cohort with SMA, the original reason for Medicare entitlement was obtained. Given the sample size, age was categorized as 18-39, 40-54, 55-64, 65-74, and ≥ 75-years-old to capture various stages of the adult lifespan.

Statistical analysis

Baseline descriptive characteristics were described for the cohorts. The prevalence of each morbidity was described by sex for adults ≥ 18-years-old. The prevalence of each morbidity was also described by age group combining females and males. There were too few outcome cases to examine prevalence estimates by sex and age group simultaneously.

Logistic regression estimated the odds ratio (OR) with 95% confidence intervals (CI) of each morbidity comparing adults (females and males combined) with vs. without SMA after adjusting for age (continuous) and sex. Effect modification by age and sex was tested for each comorbidity by including the main effects of cohort (SMA vs. non-SMA), age, and sex and their 2-way interactions. If there was evidence of effect modification (interaction term, P < 0.05) then additional analyses were performed that stratified by the variable. A 3-way interaction (cohort ^* age ^* sex) was not considered due to sample size limitations.

To compare the multi-morbidity score between cohorts, a generalized linear model with zero-inflated Poisson distribution was developed after adjusting for age (continuous) and sex. This approach utilizes a two-part model that is useful for dealing with over dispersion due to excess zero values common to multi-morbidity scores [23]. The first part of the model estimates the OR of having a zero vs. non-zero (i.e., ≥ 1) value, where an OR < 1.00 would indicate that the cohort with vs. without SMA is less likely to have 0 morbidities, and thus more likely to have ≥ 1 morbidity. The second part of the model estimates the incidence rate ratio (IRR) as the count of morbidities conditional on non-zero values; i.e., comparing SMA vs. without SMA that have ≥ 1 morbidity. Effect modification by age and sex was examined as described above.

Any variable where estimates had n ≤ 11 cases were suppressed as part of the Data Use Agreement for both databases to preserve patient de-identification. Analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

Role of the funding source

This work was not supported by any funding source.

Results

Baseline descriptive characteristics of adults with SMA (n = 2,427) and matched adults without SMA (n = 484,528) is presented in Table 1. The majority (99.5%) of the cohort with SMA had suitable matches at a ratio of 1: 200, while the remaining 0.5% (n = 12) had suitable matches ranging from 25 to 197. These 12 adults with SMA and their corresponding matches were included in the analysis and no weighting was applied.

Morbidities

The sex-stratified prevalence of each of the 30 morbidities is presented in Table 2. The age group-stratified prevalence (females and males combined) of each morbidity is presented in Figure 2 and Figure 3. Adults with SMA had a significantly elevated age- and sex-adjusted OR for each morbidity compared to adults without SMA, which ranged from 1.61 (hypothyroidism) to 24.3 (intellectual disabilities) (Table 2).

There was strong evidence of effect modification by age for 24 morbidities ( P for interaction, all ≤ 0.006). In general, the OR comparing adults with vs. without SMA was the highest for the youngest age group (18-39 year-olds) and declined with older age groups. However, the OR of all of these morbidities in the oldest age group (i.e., ≥ 75-year-olds) was still significantly elevated, ranging from 1.30 (cancer, 95% CI = 1.06-1.59) to 5.96 (neurogenic bowel/bladder, 95% CI = 3.54-10.0) (Supplementary Table S2).

There was strong evidence of effect modification by sex for 13 morbidities ( P for interaction, < 0.001 to 0.038). The OR for these morbidities was significantly elevated for females and males with vs. without SMA, but the strength of the association differed by sex. In general, the OR for females (with vs. without SMA) was higher than males (with vs. without SMA) for the cardiovascular diseases and sleep disorders, while the OR for males (with vs. without SMA) was higher than females (with vs. without SMA) for bone fragility, dementia, mental health disorders, and nutrition/gastrointestinal disorders (Table 3).

Multi-morbidity

The median (interquartile range) number of morbidities was 5 (2, 8) for females with SMA, 2 (0, 4) for females without SMA, 4 (2, 8) for males with SMA, and 1 (0, 3) for males without SMA. The number of morbidities in the following analysis were examined from 0 to ≥ 15, as the number beyond 15 morbidities became exceedingly small.

There was evidence of effect modification by sex in the zero-inflation model ( P for interaction, 0.018), but not in the count model ( P for interaction, 0.530). There was evidence of effect modification by age for females and males in the zero-inflated model ( P for interaction, 0.015 and 0.010, respectively) and count model ( P for interaction, both < 0.001). The results are presented in Table 4 by sex for the full cohort and after stratifying by age group. Females and males with vs. without SMA had a lower OR of zero-values in the zero-inflation model, indicating a higher proportion with ≥ 1 morbidity. The strength of the association declined with older age. Females and males with vs. without SMA had a higher IRR in the count model, indicating that among those with ≥ 1 morbidity, adults with SMA had on average a higher number of morbidities per person. The strength of the association declined with older age, but still remained significantly elevated across all age groups.

Discussion

The findings from this study suggest that adults with vs. without SMA had a higher risk of morbidities across several physiologic systems and multi-morbidity that were disproportionately occurring at younger ages. This is important as many of the examined morbidities are typically associated with advanced aging (e.g., ≥ 65-years-old) and may be overlooked clinically for young and middle-aged adults with SMA, thus missing out on critical opportunities for early detection and prevention. These findings highlight the need for clinical awareness of a high-early burden of morbidities as children with SMA age into and throughout their adult years.

The heterogeneity of physiologic systems impacted by morbidities observed in this study is consistent with prior studies. For example, it is known that bone fragility, including osteoporosis and fractures, and respiratory complications are more common among individuals with vs. without SMA [24-26]. A recent study by Lipkin, et al. [9] reported a higher risk of morbidities across cardiovascular, metabolic, skeletal, gastrointestinal, and reproductive systems among individuals ≤ 65-years-old with vs. without SMA.

In the current study, the most common morbidities that were present in at least 25% of females with SMA were hypertension, chronic pulmonary disease, bone fragility, osteoarthritis, anemia, cardiac arrhythmias, fluid/electrolyte disorders, diabetes, and hypothyroidism. For males with SMA, it was hypertension, cardiac arrhythmias, diabetes, fluid/electrolyte disorders, chronic pulmonary disease, anemia, and osteoarthritis. The adjusted OR suggests that the morbidities in at least 25% of adults with SMA were 61% (hypothyroidism) to 7.8-fold (fluid/electrolyte disorders) higher compared to matched adults without SMA. This highlights the magnitude of the problem given large relative differences based on morbidities that have a high absolute risk in adults with SMA (≥ 25%). It is noteworthy to mention the elevated prevalence and adjusted OR of other debilitating morbidities, including renal disease, dementia, and mental health disorders.

Claims data provides evidence about the presence of a morbidity. Claims data does not inform on the severity of the morbidity and this study did not examine if and how each morbidity was being managed through clinical interventions or its impact on quality of life. Thus, this study provides a numerical approximation regarding the potential extent of (multi-)morbidity risk for adults with SMA, but not related qualitative aspects. Further, due to the range of sensitivity/specificity of identifying morbidities using claims, the absolute risk of morbidities observed in this study may vary with the true absolute risk observed in the greater population. In the absence of evidence, an assumption was made that there is negligible differential detection of morbidities by claims across cohorts. On one hand, individuals with SMA are more likely to use healthcare services [27] which increases the chances for a diagnosis. On the other hand, premature morbidity onset can be overlooked clinically for many conditions that do not cause pain or are visually obvious. Taken together, speculation follows that the extent of differential detection bias by cohort is unlikely to be large enough to overturn the broad conclusions drawn regarding a generally higher and earlier risk of morbidities across physiologic systems for adults with vs. without SMA.

A strength of this study was the reasonably large sample of adults, considering that SMA is a rare condition and little research has been conducted for this adult population. Sequestering a sufficient volume of health-related data to provide clinically meaningful information from rare clinical populations is a grand challenge, and little is known about morbidity outcomes as people with SMA transition into and throughout adulthood. Another strength was the number of morbidities investigated. The comprehensive assessment of clinically diagnosed morbidities across physiologic systems and mental health disorders among adults with SMA may prompt for improved screening strategies and identification of at-risk individuals.

There are limitations to this study. First, claims data does not reliability differentiate by SMA type, there is no genetic information to phenotype SMA, and there was no reliably available proxy information for adults for this study, such as age when SMA-related symptoms first emerged. The survival rate differs as a function of the type of SMA [25]. Thus, the elevated prevalence of morbidities in the younger adult cohort may reflect a more medically complex sector of the SMA population (e.g., proportionally more SMA type I-III). However, even in the older age groups, which would presumably be relatively less medically complex patients with SMA (e.g., proportionally more SMA type IV), the morbidity risk was still substantially elevated. Future research is needed to distinguish age-related morbidity profiles by SMA type to further assist clinical care in identification of at-risk patients.

Second, the validity of using claims to identify people with SMA is unknown. Prior validation studies have found that identification of people with motor neuron disease to have a sensitivity of ~80-93% and a specificity ≥ 99.0%, but the codes used encompassed rather than isolated SMA [28]. The algorithm used in this study is consistent with the Chronic Conditions Data Warehouse for neuromuscular and muscular dystrophy populations and balances sensitivity and specificity. Although, other studies have used slightly different algorithms to identify SMA, such as ≥ 1 inpatient claim or ≥ 2 outpatient claims within 30 days of one another, or ≥ 1 inpatient claim or ≥ 3 outpatient claims [9,29]. Nevertheless, selection bias is possible, such that this study included patients with SMA who received more clinical care, which could overestimate the extent of morbidity risk for the greater adult population with SMA.

Third, generalizability of study findings for the greater adult population with SMA is not exactly known. While studies of prevalence and incidence for this rare population have significant challenges [30], reported prevalence estimates range from 1 to ~7 per 100,000 people; although, these studies were among those < 20-years-old [30-34]. In the current study, there were 654 adults with SMA identified from the Optum database and 12,300,242 adults without SMA from the Optum database that met all other eligibility criteria prior to matching. The prevalence of SMA in this Optum cohort was 5.3 per 100,000 adults, which is within the range previously reported [30-34]. A prevalence estimate from the Medicare database was not possible due to differences in enrollment not allowing for a full age spectrum background population for those without SMA. Further, to be eligible for analysis, this study required 2-years of continuous enrollment. Given the elevated mortality rate, the cohort with SMA may have been disproportionately healthier than the cohort without SMA compared to their own background populations, which may have led to a slight underestimation of the relative risk.

In conclusion, adults with SMA are vulnerable to an array of morbidities and multi-morbidity across several physiologic systems and mental health disorders. Morbidity onset may occur earlier in the adult lifespan for those with vs. without SMA. Increasing clinical awareness of the high-early risk of morbidities, improving screening strategies, and developing appropriate referral systems may help to reduce the burden of morbidities for this vulnerable population.

Acknowledgements

All persons that have contributed to the work significantly are listed. This study was not supported by funding. All authors report no conflict of interest.

Supplementary File

Supplementary Tables

References

1. Sugarman EA, Nagan N, Zhu H, Akmaev VR, Zhou Z, et al. (2012) Pan-ethnic carrier screening and prenatal diagnosis for spinal muscular atrophy: Clinical laboratory analysis of >72,400 specimens. Eur J Hum Genet 20: 27-32.
2. D'Amico A, Mercuri E, Tiziano FD, Bertini E (2011) Spinal muscular atrophy. Orphanet J Rare Dis 6: 71.
3. Zerres K, Rudnik-Schoneborn S, Forrest E, Lusakowska A, Borkowska J, et al. (1997) A collaborative study on the natural history of childhood and juvenile onset proximal spinal muscular atrophy (type II and III SMA): 569 patients. J Neurol Sci 146: 67-72.
4. Kolb SJ, Kissel JT (2015) Spinal Muscular Atrophy. Neurol Clin 2015 33: 831-846.
5. Nicolau S, Waldrop MA, Connolly AM, Mendell JR (2021) Spinal Muscular Atrophy. Semin Pediatr Neurol 37.
6. Coovert DD, Le TT, McAndrew PE, Strasswimmer J, Crawford TO, et al. (1997) The survival motor neuron protein in spinal muscular atrophy. Hum Mol Genet 6: 1205-1214.
7. Lefebvre S, Burlet P, Liu Q, Bertrandy S, Clermont O, et al. (1997) Correlation between severity and SMN protein level in spinal muscular atrophy. Nat Genet 16: 265-269.
8. Groen EJN, Perenthaler E, Courtney NL, Jordan CY, Shorrock HK, et al. (2018) Temporal and tissue-specific variability of SMN protein levels in mouse models of spinal muscular atrophy. Hum Mol Genet 27: 2851-2862.
9. Lipnick SL, Agniel DM, Aggarwal R, Makhortova NR, Finlayson SG, et al. (2019) Systemic nature of spinal muscular atrophy revealed by studying insurance claims. PLoS One 14: e0213680.
10. Darras BT, Chiriboga CA, Iannaccone ST, Swoboda KJ, Montes J, et al. (2019) Nusinersen in later-onset spinal muscular atrophy: Long-term results from the phase 1/2 studies. Neurology 92: e2492-e2506.
11. De Vivo DC, Bertini E, Swoboda KJ, Hwu WL, Crawford TO, et al. (2019) Nusinersen initiated in infants during the presymptomatic stage of spinal muscular atrophy: Interim efficacy and safety results from the Phase 2 NURTURE study. Neuromuscul Disord 29: 842-856.
12. Finkel RS, Chiriboga CA, Vajsar J, Day JW, Montes J, et al. (2016) Treatment of infantile-onset spinal muscular atrophy with nusinersen: A phase 2, open-label, dose-escalation study. Lancet 388: 3017-3026.
13. Finkel RS, Mercuri E, Darras BT, Connolly AM, Kuntz NL, et al. (2017) Nusinersen versus sham control in infantile-onset spinal muscular atrophy. N Engl J Med 377: 1723-1732.
14. Mercuri E, Darras BT, Chiriboga CA, Day JW, Campbell C, et al. (2018) Nusinersen versus sham control in later-onset spinal muscular atrophy. N Engl J Med 378: 625-635.
15. Al-Zaidy S, Pickard AS, Kotha K, Alfano LN, Lowes L, et al. (2019) Health outcomes in spinal muscular atrophy type 1 following AVXS-101 gene replacement therapy. Pediatr Pulmonol 54: 179-185.
16. Mendell JR, Al-Zaidy S, Shell R, Arnold WD, Rodino-Klapac LR, et al. (2017) Single-dose gene-replacement therapy for spinal muscular atrophy. N Engl J Med 377: 1713-1722.
17. Belter L, Cruz R, Kulas S, McGinnis E, Dabbous O, et al. (2020) Economic burden of spinal muscular atrophy: An analysis of claims data. J Mark Access Health Policy 8: 1843277.
18. Charlson ME, Pompei P, Ales KL, MacKenzie CR (1987) A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 40: 373-383.
19. Quan H, Li B, Couris CM, Fushimi K, Graham P, et al. (2011) Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol 173: 676-682.
20. Elixhauser A, Steiner C, Harris DR, Coffey RM (1998) Comorbidity measures for use with administrative data. Med Care 36: 8-27.
21. Whitney DG, Basu T (2021) Whitney Comorbidity Index to monitor health status for adults with cerebral palsy: Validation and thresholds to assist clinical decision making. Dev Med Child Neurol 63: 853-859.
22. Whitney DG, Kamdar NS (2021) Development of a new comorbidity index for adults with cerebral palsy and comparative assessment with common comorbidity indices. Dev Med Child Neurol 63: 313-319.
23. Lambert D (1992) Zero-inflated poisson regression, with an application to defects in manufacturing. Technometrics 34: 1-14.
24. Vestergaard P, Glerup H, Steffensen BF, Rejnmark L, Rahbek J, et al. (2001) Fracture risk in patients with muscular dystrophy and spinal muscular atrophy. J Rehabil Med 33: 150-155.
25. Wijngaarde CA, Stam M, Otto LAM, van Eijk RPA, Cuppen I, et al. (2020) Population-based analysis of survival in spinal muscular atrophy. Neurology 94: e1634-e1644.
26. Wijngaarde CA, Veldhoen ES, van Eijk RPA, Stam M, Otto LAM, et al. (2020) Natural history of lung function in spinal muscular atrophy. Orphanet J Rare Dis 15: 88.
27. Armstrong EP, Malone DC, Yeh WS, Dahl GJ, Lee RL, et al. (2016) The economic burden of spinal muscular atrophy. J Med Econ 19: 822-826.
28. St Germaine-Smith C, Metcalfe A, Pringsheim T, Roberts JI, Beck CA, et al. (2012) Recommendations for optimal ICD codes to study neurologic conditions: A systematic review. Neurology 79: 1049-1055.
29. Chen G, Sharif B, Gerber B, Farris MS, Cowling T, et al. (2021) Epidemiology, healthcare resource utilization and healthcare costs for spinal muscular atrophy in Alberta, Canada. J Med Econ 24: 51-59.
30. Verhaart IEC, Robertson A, Wilson IJ, Aartsma-Rus A, Cameron S, et al. (2017) Prevalence, incidence and carrier frequency of 5q-linked spinal muscular atrophy - a literature review. Orphanet J Rare Dis 12: 124.
31. Arkblad E, Tulinius M, Kroksmark AK, Henricsson M, Darin N (2009) A population-based study of genotypic and phenotypic variability in children with spinal muscular atrophy. Acta Paediatr 98: 865-872.
32. Darin N, Tulinius M (2000) Neuromuscular disorders in childhood: A descriptive epidemiological study from western Sweden. Neuromuscul Disord 10: 1-9.
33. Merlini L, Stagni SB, Marri E, Granata C (1992) Epidemiology of Neuromuscular disorders in the under-20 population in Bologna Province, Italy. Neuromuscul Disord 2: 197-200.
34. Tangsrud SE, Halvorsen S (1988) Child neuromuscular disease in southern Norway. Prevalence, age and distribution of diagnosis with special reference to "Non-Duchenne muscular dystrophy". Clin Genet 34: 145-152.

---