Respiratory insufficiency (restrictive lung disease, impaired cough) is a potential complication of FSHD [1]. Evaluation of respiratory function among those with FSHD typically relies on measures of forced vital capacity (FVC), maximal inspiratory pressure (MIP), maximal expiratory pressure (MEP), and polysomnograms/sleep studies. Most studies reporting on these measures have been small clinical samples conducted internationally [2–7]. Findings have been mixed with some studies reporting impaired respiratory function, [6–9] but others reporting normal respiratory function among most study participants [2,4,5,10]. Decreased respiratory function has been consistently noted among people with more severe FSHD symptoms or longer disease duration [2,4]. Only two studies were identified that examined early onset (<18 years of age) FSHD, with both reporting impaired respiratory function in this patient population [11,12]. A recent single-site US study reported nearly 50% of people tested showed reduced FVC and recommended continued study of respiratory complications in FSHD using population-based samples to improve generalizability [13].

There are few studies on respiratory insufficiency in individuals with FSHD in the US, and the sampling for several of the studies included people with risk factors known to increase risk of respiratory insufficiency [3–5,7,10,13,14]. Further characterization of respiratory dysfunction in this population early in disease confirmation will give additional insights into the monitoring and management of respiratory function in these individuals, in particular baseline assessments of respiratory function [1]. As such, we used data from the US population-based Muscular Dystrophy Surveillance, Tracking, and Research Network (MD STARnet) to describe patterns and results of respiratory monitoring of people with FSHD.

Overall, 173 people with FSHD were included in the MD STARnet, of which 170 people met the inclusion criteria. Most were confirmed by genetic testing, diagnosed with FSHD1, and classified as adult-onset (Table 1). The sample was largely non-Hispanic white and evenly distributed between males and females. Most had known family history of FSHD, remained ambulant, and had no documentation of scoliosis or lordosis during follow-up. Nearly one-third could be classified as ever obese. Average ages of first and last health encounters captured by MD STARnet were 39 (SD=21.0) and 42 (SD=21.3) years, respectively; median follow-up time was 4 years (minimum = 1 year, maximum = 9 years; IQR = 2-6 years). Demographic characteristics tended to be similar between those with early- and adult-onset FSHD; proportions for presence of risk factors for respiratory insufficiency tended to be higher for early-onset (Table 2).

One in five people had FVC testing; other respiratory tests were less frequent (Table 3). The first FVC testing occurred within an average of 1.9 (SD = 1.8) years of their first health encounter. The median FVC at initial documented testing was 77% of predicted; median MIP and MEP were 58.0 cmH₂O and −50.0 cmH₂O, respectively. At least mildly impaired respiratory function (<80% FVC) was noted for a small percentage of people overall but for the majority of those tested. No individual had a recorded FVC <1.0L.

Higher proportions of respiratory testing were observed among those with at least one risk factor for respiratory compromise (Table 4) and for each factor individually (Table 5) than people without. Conversely, the proportion of people with any abnormal respiratory function test was generally as high or higher among people without any risk factors compared to people with at least one risk factor (Table 4). Proportions of people who had respiratory testing were higher among those with four or more neurology/neuromuscular visits (FVC=32.4%, MIP/MEP=16.9%, polysomnogram=19.7%) compared to people with fewer than four visits (FVC=11.1%, MIP/MEP=4.0%, polysomnogram=10.1%) (data not shown).

Using a US population-based, multisite sample, we found 20% of people with FSHD had documented respiratory testing during 2008-2016. Although we had a small number of people with known risk factors for respiratory insufficiency and found more frequent evaluations among these people, the proportion of people tested was lower than would be expected given longstanding recognition of the importance of pulmonary monitoring for at-risk patients. Our study supports the current recommendations of baseline respiratory testing in FSHD given our finding of impaired respiratory function among people without known risk factors for respiratory insufficiency.

Previous studies of respiratory function of people with FSHD have reported variable rates of abnormal test results ranging from less than 10% to near 50% [2,4–13]. The variability of reported findings is likely due in part to disease stage of the underlying population, with more severely affected patient populations demonstrating greater respiratory insufficiency [2–6,9–12]. Unlike most published series, our sample is restricted to people with confirmation of an FSHD diagnosis during the surveillance period, possibly skewing to a younger population and milder stage of disease. Because our sample is retrospective and a minority of patients underwent testing, we cannot estimate the frequency of respiratory compromise. However, we note that among those tested, we observed at least mild respiratory insufficiency in more than one-half.

Interestingly, we observed similar frequencies of at least mildly abnormal FVCs among tested people with and without risk factors. It is possible that other clinical factors prompted testing in those without the risk factors we considered here. Alternatively, mild respiratory compromise might be common and unaffected by the studied risk factors. A 5-year follow up study of respiratory function in 92 people with FSHD showed a mean baseline FVC of 79% predicted. Rapid decline in FVC was seen in a minority and associated with stage of disease (weakness) and spinal deformity within 5 years [6]. Although only a small proportion of our sample had MIP/MEP testing, approximately one-half were abnormal. This suggests that impaired airway clearance and increased aspiration risk might be more common than recognized in FSHD and merits screening questions, clinical assessment, and selected formal testing. We note that peak cough flow, another common measure of airway clearance, was too infrequent (n=8) to analyze separately and MD STARnet did not collect data on insufflator/exsufflator use or symptoms (e.g., frequent or prolonged respiratory infections) that may indicate impaired airway clearance in people with FSHD.

FSHD is usually a relatively slowly progressing disorder. Some with the disease might have limited routine healthcare, particularly as it relates to FSHD. We observed that those who had more frequent neurology/neuromuscular provider visits during the surveillance period were more likely to have respiratory function testing recorded than those with less frequent visits. This difference suggests that those who have more specialized healthcare visits may be more likely to have care that allows monitoring of respiratory symptoms or risk factors that may prompt respiratory testing. Alternatively, those with more advanced disease or those experiencing symptoms related to respiratory dysfunction might be more likely to seek care and thus be tested. Finally, it could be that those with fewer documented visits, or less centralized care, are receiving care at clinic sources not accessible to the MD STARnet. Available MD STARnet data do not allow us to distinguish these possibilities.

Study limitations include reliance on available medical records and small numbers limiting statistical comparisons. We did not collect indications (e.g., shortness of breath, symptoms of nocturnal hypoventilation [morning headaches, feeling sleepy in the daytime, not feeling refreshed in the morning, dizziness]) for respiratory testing among those tested. It is possible that respiratory symptoms reported in clinic, but not included as part of our risk profile, prompted some evaluations. This may have biased initial test results towards evidence of respiratory insufficiency. In addition, routine FVC testing can be physically challenging and inaccurate due to disease-related facial weakness; thus, some people might have attempted testing, but valid results could not be determined. MD STARnet does not collect severity of signs and symptoms (e.g, facial weakness) of FSHD. Thus, we were unable to identify these situations, and such people would have been considered untested if the attempt was not recorded in the medical record and abstracted [3,6]. In addition, MD STARnet does not collect data on the method used for pulmonary testing (mask vs mouthpiece, for example) that might affect the accuracy of the test result. Testing quality was not documented for all respiratory testing.

Our study found that few people with newly diagnosed FSHD have routine respiratory function screening but a substantial proportion of those tested showed evidence of respiratory insufficiency. Additionally, although people with risk factors for respiratory complications were more likely to be tested, respiratory insufficiency was identified among people without these risk factors at similar or even higher proportions than found among those at risk. Combined with a recent MD STARnet cross-sectional study that found respiratory complications were a contributing factor to death in over 50% of deceased individuals with FSHD, [19] these findings suggest respiratory insufficiency may be a frequent complication of FSHD regardless of perceived risk level. In addition to recommended baseline testing, our findings suggest regular respiratory monitoring of people with FSHD should be considered, even in the absence of classic signs and symptoms of respiratory insufficiency [1,4]. Future studies are needed to establish the interval at which monitoring should occur.