The bi-directional relationship between sleep and inflammation in muscular dystrophies: A narrative review

Muscular dystrophies vary in presentation and severity, but are associated with profound disability in many people. Although characterised by muscle weakness and wasting, there is also a very high prevalence of sleep problems and disorders which have significant impacts on quality of life in these individuals. There are no curative therapies for muscular dystrophies, with the only options for patients being supportive therapies to aid with symptoms. Therefore, there is an urgent need for new therapeutic targets and a greater understanding of pathogenesis. Inflammation and altered immunity are factors which have prominent roles in some muscular dystrophies and emerging roles in others such as type 1 myotonic dystrophy, signifying a link to pathogenesis. Interestingly, there is also a strong link between inflammation/immunity and sleep. In this review, we will explore this link in the context of muscular dystrophies and how it may influence potential therapeutic targets and interventions.


Introduction
Sleep is defined as a reversible state of disengagement and unresponsiveness to the environment (Carskadon and Dement, 2005). Sleep aids in immunity, conserves energy, and there is evidence of altering neural connectivity and plasticity (Krueger et al., 2016). Sleep disorders are relatively common in the general population but have been associated with multisystem dysfunction, reduced quality of life and reduced ability to carry out daily tasks. Sleep disorders are particularly prevalent in those with muscular dystrophies, which are a clinically heterogenous group of disorders caused by genetic mutations which affect muscle function. They affect approximately 1/3000 to 1/8000 individuals worldwide (Ashizawa et al., 2018) and there are many different types, which vary in muscles affected, signs and symptoms, severity, and age of presentation.
Duchenne muscular dystrophy (DMD) is the most prevalent, affecting approximately 10 in 100,000 males (Duan et al., 2021). It is caused by mutations in the gene coding for the protein dystrophin which is vital to the integrity of muscle fibre cell membrane. Thosewith DMD usually use a wheelchair by age 12, with death typically occurring in third to fourth decade (Duan et al., 2021). Becker muscular dystrophy (BMD) is a less severe form of DMD, caused by mutations which allow partially functional dystrophin protein (Duan et al., 2021). Myotonic dystrophy (DM) is the most common form of muscular dystrophy in adults, affecting about 8 in 100,000 people (Thornton, 2014). It involves myotonia (delayed relaxation of muscle after contraction) and muscle weakness along with non-muscular symptoms such as cataracts, cardiac issues, and sleep and fatigue problems (Thornton, 2014). There are two types, type 1 (DM1) caused by a CTG repeat expansion in the DMPK gene, and type 2 (DM2) caused by a CCTG repeat expansion in the ZNF9 gene, with DM2 generally being a milder form. Both of these mutations lead to repetitive RNA which has a toxic effect in cells and tissues, likely contributing to the diversity of symptoms seen in this disorder (Thornton, 2014). Adult onset DM1 usually presents in late teens to 30s, but is highly variable. Premature death in adulthood frequently occurs due to pulmonary or cardiac complications. DM2 age of onset is similar, but no effect on life expectancy has been noted. There is also a congenital form of DM1 which is very severe.
Facioscapulohumeral muscular dystrophy (FSHD) is another relatively common muscular dystrophy, affecting about 4 in 100,000 individuals (Richards et al., 2012). It mainly affects the muscles of the voluntary muscles of the face, scapula, and upper arm and often affects one side of the body more than the other. It is caused by overexpression of the DUX4 gene which is toxic to muscle (Lemmers et al., 2010;Wallace et al., 2011). Life expectancy is usually unaffected but there are impacts on quality of life. Other less common muscular dystrophies include limb-girdle (LGMD), oculopharyngeal, and Emery-Dreifuss, which have slightly different presentation, but share the common characteristic of progressive muscle weakness.
Sleep disorders and fatigue are common complaints among those with a muscular dystrophy. Sleep disordered breathing (SDB) is one cause that is highly prevalent across neuromuscular diseases (~ 79 %), mainly manifesting as obstructive sleep apnea (OSA) and occasionally central sleep apnea (CSA) (Aboussouan, 2015). In normal physiology, there is a reduction in ventilatory drive in sleep, but those with muscular dystrophies are at risk of a greater than normal reduction due to weakness of the muscles involved in respiration (Aboussouan, 2015). OSA is usually present in the first decade in DMD, and hypoventilation is evident thereafter. Those with DMD are also at increased risk of CSA, hypopnea and hypoxemia (Nozoe et al., 2015). The proportion of those with DM1 who had evidence of sleep apnea varies from 28 % to 93 %, and from 38 % to 64 % in DM2 (Subramony et al., 2020). There is also increased incidence in FSHD (Della Marca et al., 2009).
There are also other sleep problems evident in those with muscular dystrophy aside from SDB. Excessive daytime sleepiness is a very common complaint in those with DM1, occurring in 25-82 % of cases (Subramony et al., 2020). This daytime sleepiness has been seen to be independent from sleep apnea, as sleep apnea is detected clinically in those without EDS and vice versa (Kiyan et al., 2010;van der Meché et al., 1994). This indicates a central nervous system (CNS) cause, and this pattern is also seen in DMD (Hoque, 2016) and FSHD (Della Marca et al., 2009) although EDS is less prevalent. Other sleep problems reported in DM1 include increased periodic limb movements, time in bed, decreased sleep quality, and changes in sleep architecture (Subramony et al., 2020;Romigi et al., 2011;Yu et al., 2011). In addition, those with DMD had decreased sleep efficiency, increased REM sleep latency and a reduced percentage of REM sleep (Nozoe et al., 2015). There is evidence of impaired subjective sleep quality in those with FSHD which correlates with disease severity (Della Marca et al., 2007), and musculoskeletal pain may be playing a role in sleep disturbance (Della Marca et al., 2013).
SBD has massive clinical implications in muscular dystrophies, as it is associated with and contributes to morbidity and mortality of muscular dystrophies. Not only this, but sleep disturbance is associated with a decrease in quality of life in those with muscular dystrophies and is rated highly in importance of symptoms to patients (Crescimanno et al., 2019;Heatwole et al., 2006). Nocturnal non-invasive ventilation (NIV) is an effective treatment for SDB, increasing survival and quality of life in a range of neuromuscular disorders (Guilleminault et al., 1998). It also improved sleep quality, decreased daytime sleepiness, and slowed decline of pulmonary function (Gomez-Merino and Bach, 2002;Simonds et al., 1998). However, there are challenges to the implementation of NIV. In cohorts which use NIV for SDB control without a muscular dystrophy, a low rate of compliance is well documented and results from factors such as the lifestyle changes required to use the machine during the night and side effects including arousals and irritation (Yetkin et al., 2008;Shapiro and Shapiro, 2010). This is reflected in the DM1 cohort and may minimise the potential benefit (Boussaïd et al., 2016). Stimulants, mainly modafinil, have been used to counter daytime sleepiness in DM1 (Ashizawa et al., 2018). It is used in an open-label fashion but has seen some success in trials with DM (Annane et al., 2006;MacDonald et al., 2002;Talbot et al., 2003). It has yet to be trialled in other muscular dystrophies, and will be discussed further later in the review.
It is evident that sleep disorders are an important issue in muscular dystrophies and the current treatments are insufficient. In this narrative review, we aim to discuss the evidence for the influence of inflammation in the pathogenesis of muscular dystrophies, and how this influence could be exploited to provide new and effective treatments to treat sleep and other relevant symptoms. Although sleep differences in muscular dystrophies are well established in the literature, to our knowledge there are no current reviews investigating how inflammation may influence sleep, and vice versa, in this cohort. This review will provide an overview of a potential bi-directional relationship between sleep and inflammation in muscular dystrophies, to form a basis for further research into the relationship to improve treatment options. Publications were found and reviewed through Google Scholar and PubMed searches of the key words 'Muscular dystrophy and inflammation' and 'Muscular dystrophy and sleep'. Further publications were reviewed from bibliographies in articles found through these searches, conducted between January and March 2022.

Influence of sleep on inflammation
It has been long known that sleep, inflammation, and immunity are closely intertwined. A prime example of this is that insomnia, a very common affliction in the general population, increases risk of inflammatory disorders and infectious diseases (Irwin, 2019;Ohayon, 2002). Conversely, those with inflammatory disorders and infectious diseases are more likely to have sleep problems (Irwin, 2019). The disease risks in chronic inflammation are much the same as those seen with chronic sleep disturbance, such as depression, Alzheimer disease and cancer (Irwin, 2019). In support of this link, sleep disturbance has been seen to alter multiple markers of inflammation. The pattern of secretion of inflammatory cytokines is influenced by both sleep and circadian rhythms (Lange et al., 2010). There is peak production of proinflammatory cytokines (interleukin-2 (IL-2), IL-6, IL-12, tumour necrosis factor-α (TNFα), interferon-γ (IFNγ)) during sleep at night, while some anti-inflammatory cytokines such as IL-10 peak during the day (Lange et al., 2010). The different sleep stages have also been found to influence nocturnal changes in inflammatory cytokine levels (Irwin, 2019). Immune cells also show distinct diurnal rhythms. Leukocytes are influenced by cortisol which peaks at night and adrenaline which peaks during the day, and sleep consolidates and stabilises these rhythms. Nocturnal sleep also suppresses lymphocytes, monocytes and natural-killer cells.
Sleep deprivation shifts the maximum of inflammatory cytokines from the night to the day, causing an over secretion during the day which encourages a proinflammatory state (Lange et al., 2010). Proinflammatory IL-6 and TNFα are consistently increased in chronic sleep loss and T cell function and natural killer (NK) cell activity are decreased (Lange et al., 2010). In a large meta-analysis of over 50,000 individuals across 72 studies, sleep disturbance was associated with higher levels of C-reactive protein (CRP) and IL-6, both proinflammatory markers (Irwin et al., 2016). Short sleep duration (< 7 h per night) when sleep was assessed continuously was associated with increased CRP but not IL-6, but long sleep duration (> 8 h) when compared with normal reference (Lemmers et al., 2010;Wallace et al., 2011) hours) showed increases in both markers. In a longitudinal study, sleep disturbance at baseline was a significant predictor of inflammation (as measured by circulating IL-6 and CRP) five years later (Cho et al., 2015). Circulating levels of inflammatory markers have also been found to be elevated in clinical populations which have a high level of sleep disturbance. In a meta-analysis of levels of inflammatory markers in OSA, CRP, TNFα, IL-6, IL-8, ICAM, selectins and VCAM were all found to be increased (Nadeem et al., 2013). Alcoholic patients had decreased IL-6 production and decreased NK cell activity, along with altered patterns of IL-6 secretion (Redwine et al., 2003). Rotating shift workers have elevated total leucocyte count and CRP levels (Puttonen et al., 2011;Sookoian et al., 2007).

The influence of inflammation on sleep
This relationship works both ways, as altering inflammation can also influence sleep. When insomnia is successfully treated clinically with specialised cognitive behavioural therapy (CBT) or Tai Chi, there are corresponding decreases in inflammatory signatures, which are comparable to the effects of improved diet and exercise (Irwin, 2019;Irwin et al., 2015). Pharmacological blockade of TNF in depressed patients with increased inflammation normalised sleep parameters such as wake after sleep onset compared to those with low inflammation (Weinberger et al., 2015). Similar results were seen for those with alcohol dependence (Irwin et al., 2009). Anti-TNFα therapy also significantly improved sleep efficiency in rheumatoid arthritis, although this may be attributed to improvement in other symptoms such as pain, which allowed participants to sleep better (Karatas et al., 2017;Taylor-Gjevre et al., 2011). Similar results were seen in the treatment of psoriasis (Strober et al., 2012;Thaçi et al., 2014). More convincingly, a meta-analysis on the effect of continuous positive airway pressure (CPAP) treatment on inflammatory markers in OSA found that CRP, IL-6, IL-8 and TNFα were decreased by CPAP treatment, indicating a normalisation of inflammatory markers by treating the underlying sleep problem (Xie et al., 2013). The effects of pharmacological TNFα blockade in those with OSA found that there was a significant decrease in sleepiness as measured by the multiple sleep latency test compared with placebo, which was three times higher than that seen with CPAP (Vgontzas et al., 2004). The number of apneas per hour were also significantly decreased, along with IL-6 levels.

Sleep and immunity
Along with sleep having an influence on inflammation, there is also a link between sleep and immunity, the body's response to infection. Immune defence mechanisms are encouraged during sleep, including the accumulation of antigen presenting cells in lymphoid tissue, activation of T cells, increased production of IL-12, and a shift towards Th1 immune responses (Lange et al., 2010). This is reversed in chronic and experimental sleep disturbance, where there is a shift towards Th2 activity, decreased IL-2 expression, and increased IL-10 expression (Lange et al., 2010). Therefore, disturbed sleep can increase susceptibility to infection. This is illustrated by decreased immunity in response to sleep deprivation. It was found that those who got an average of less than 7 h of sleep were almost three times more likely to develop the common cold when inoculated with rhinovirus than participants who got eight or more hours sleep (Cohen et al., 2009). It has also been documented that sleep duration of less than 5 h increases pneumonia, common cold, and respiratory infection susceptibility (Irwin, 2019). In addition, vaccine responses may be blunted as a consequence of sleep loss. In those receiving hepatitis A vaccination, participants who had normal sleep the night after vaccination produced nearly twice as much antibody after 4 weeks compared to those who were kept awake the night after vaccination (Lange et al., 2003). In summary, these studies indicate the need for sleep for an adequately functional immune system.
The link between the immune system and sleep is also evident in the alteration of sleep architecture in response to viral infections. This can be changes in sleep duration, continuity, and duration of sleep stages (Irwin, 2019). It is hypothesised that inflammatory cytokines produced in response to viral infections is responsible for these effects (Irwin, 2019). This has been illustrated in mice deficient in inflammatory receptors, who have altered sleep responses to infection. Bacterial infections also alter sleep architecture, but in a slightly different way by increasing and/or decreasing non-REM sleep duration (Davis et al., 2015;Kapás et al., 2008). Different patterns are seen depending on the bacteria. Again, this emphasises the ability for immunity to influence sleep.

Inflammation/immunity in muscular dystrophies
There is evidence of inflammation and altered immunity both being associated with and playing a role in pathogenesis in multiple forms of muscular dystrophy. In DMD, altered immune function has long been known to play a prominent role in disease, whereas in other disorders such as myotonic dystrophy, the association is less clear and still emerging. This section will detail the links between inflammation or immunity and muscular dystrophy in the literature.

Myotonic dystrophy
Altered immune system function was first identified in DM as reduced levels of circulating immunoglobulin G (IgG) in those with DM compared to controls, which was attributed to hypercatabolism of IgG (Wochner et al., 1966). There was a correlation between reduced IgG levels and CTG repeat length, indicating a relationship to disease severity. Deficiency in ability to produce antibody in response to immunological challenge was also noted (Grove et al., 1973). Increased levels of circulating cytokines IL-6, TNF-α, and IL-1β have been found in those with DM (Johansson et al., 2000;Mammarella et al., 2002). All three of these markers are proinflammatory cytokines, with roles in both the acute response to infection as part of the innate immune response but can also support chronic inflammation (Amaral et al., 2020;Gabay, 2006;Pahwa et al., 2018). In the congenital form of DM, muscle samples were found to have an upregulation of the IL-6 signalling pathway which correlated with muscle immaturity (Nakamori et al., 2017). The IL-6 pathway is also involved in myocyte maturation and atrophy, as mice that overexpress IL-6 show severe muscular atrophy which is reverse by IL-6 blockade (Tsujinaka et al., 1996). In addition to the IL-6 pathway, there was also significant upregulation in processes involved with inflammation as a whole. A similar enrichment for upregulation of genes associated with the innate immune response, particularly interferon signalling, was found in DM1 and DM2 cataract samples (Rhodes et al., 2012).
In a study of laboratory test results in 126 DM1 patients revealed that enzymes and hormones with immunological influence, such as luteinising hormone (LH), follicle-stimulating hormone (FSH), creatine kinase and lactate dehydrogenase, were regularly abnormal in DM1 patients (Heatwole et al., 2006). Total white blood cell counts in DM1 were also low compared to the normal population count levels (22 % of tests). There is also evidence of dysregulation of enzymes involved in maintaining oxidative homeostasis and the total antioxidant system (Toscano et al., 2005), potentially contributing to oxidative stress which influences and activates inflammation (Chatterjee, 2016). Gene enrichment analysis in an inducible DM1 model in glial cells discovered inflammation and immune response as major cellular deregulated processes, indicating impaired immune function of glia in DM1 (Azotla--Vilchis et al., 2021). This may have an influence on CNS pathology, as in Mbnl2 knockout mice there is an overexpression of proinflammatory microglia in the medial prefrontal cortex (Ramon-Duaso et al., 2018). There has been further characterisation of impaired immune function in glia, with exaggerated RNA toxicity in the Bergmann glia of the cerebellum found in both DMSXL mice and humans (Sicot et al., 2017), and glial cells isolated from DMSXL mice have molecular signatures indicative of impaired differentiation (González- Barriga et al., 2021). In a recent study of the blood transcriptome associated with DM1 patients in two independent cohorts, DM1 severity was related to alterations in innate and adaptive immunity which were indicative of muscle wasting in multiple methods of pathway analysis (Nieuwenhuis et al., 2022). This indicates that DM1 severity may be associated with innate and adaptive immunity processes. Similarly, another study using OPTI-MISTIC trial data investigated mRNA expression in blood from patients before and after CBT intervention (van Cruchten et al., 2022). Again, enrichment analysis linked CTG repeat-induced disease and treatment response to immunological pathways such as cytokine and chemokine signalling.

Duchenne muscular dystrophy
The immune system and inflammatory processes have a prominent role in the muscle pathology of DMD. There is strong activation of much of the innate immune system prior to the onset of clinical symptoms (Rosenberg et al., 2015). This likely results from the membrane instability associated with a loss of dystrophin, and the consequential leak of cytoplasmic material into the extracellular space (Rosenberg et al., 2015). This leads to the infiltration of DMD muscle by a wide range of immune cells, including CD4+ and CD8+ T cells, macrophages, neutrophils, and eosinophils (Evans et al., 2009). Here, they mediate the sequential degradation and fibrotic regeneration of muscle. In support of this statement, depletion of CD4+ T cells reduces muscle damage by 75 % and CD8+ T cells reduces it 61 % in the mdx mouse model of DMD (Spencer et al., 2001;Wehling-Henricks et al., 2008). Similar reductions in muscle necrosis are also seen with the depletion of neutrophils (Hodgetts et al., 2006) and the absence or inhibition of inflammatory mediators (Radley et al., 2008;Villalta et al., 2011). Furthermore, lymphocyte depletion reduces levels of fibrosis and transforming growth factor-β (TGFβ), characteristic of the remodelling phase in DMD muscle (Farini et al., 2007). Mast cells also likely play a role in pathogenesis, as they are abundant and degranulated surrounding DMD muscle (Gorospe et al., 1994a(Gorospe et al., , 1994b. Mast cell degranulation releases proteases which can stimulate lysis of nearby cells, and also releases cytokines which further contribute to the proinflammatory environment. Blocking mast cell degranulation reduced muscle damage in the DMD mouse model by 59 %  and DMD individuals also have 10 times more mast cells than non-DMD individuals (Gorospe et al., 1994b). Regulatory T cells (Tregs) are yet another immune cell type seen to have a role, as depletion again increased muscle damage and inflammation in the mdx mouse model, showing increased activation of M1 macrophages (Villalta et al., 2014). Treatment with the Treg stimulating cytokine IL-2 decreased muscle injury, through increasing IL-10 concentrations to stimulate M2 macrophages.
There are two types of macrophage class in muscle, M1; and M2. M1 macrophages are stimulated by proinflammatory cytokines such as TNFα and IFNγ and contribute to muscle lysis by increasing production of nitric oxide (Evans et al., 2009). M2 macrophages, which are stimulated by cytokines produced by Th2 cells such as IL-4 and IL-10, promote tissue repair by inhibiting inflammation and M1 cytotoxicity and inducing satellite cell proliferation (Evans et al., 2009). There are a greater number of M1 macrophages in early DMD, likely contributing to muscle damage and the proinflammatory environment, therefore macrophage class switching is a therapeutic target. M2 macrophages are abundant in the regeneration phase of DMD.
In addition, cytokines and chemokines play a role in the destruction of muscle tissue in DMD, with both gene expression and protein levels of many molecules increased. There is widespread activation of many proinflammatory pathways such as NFκb and TGFβ, which also play a role in fibrosis and tissue remodelling. There is elevated serum TNFα, but not elevated expression of TNFα and IL-1β in muscle tissue in DMD (Evans et al., 2009). Despite this, TNFα blockade in a DMD mouse model prevented damage to and preserved function of muscle (Hodgetts et al., 2006;Radley et al., 2008), indicating a role for this proinflammatory cytokine in muscle pathogenesis.
Major histocompatibility complex (MHC) class I proteins are not normally expressed on skeletal muscle fibres, but there is a 3 fold increase in their presentation in those with DMD possibly as a result of the presence of proinflammatory cytokines in the muscle (McDouall et al., 1989). This further recruits T and B cells to the muscle which contribute to a chronic proinflammatory state. There may also be an auto-immune component to DMD. Dystrophin reactive T cells have been identified in DMD patients (Flanigan et al., 2013). This may be due to dystrophin being expressed in the thymus in healthy individuals, but lacking in DMD individuals, leading to a retention and release of dystrophin reactive T cells.
This compilation of inflammatory and immune processes in DMD is not exhaustive, further details are reviewed in (Rosenberg et al., 2015;Evans et al., 2009). Due to the inflammatory nature of DMD pathogenesis, treatment with glucocorticoids is the first line pharmacotherapy prescribed. Glucocorticoid treatment in DMD improves muscle strength and function, reduces risk of loss of mobility and progression, and reduces overall risk of death (Manzur et al., 2008;McDonald et al., 2018). Other potential anti-inflammatory treatments will be discussed in Section 4.2.

Facioscapulohumeral muscular dystrophy
There is also evidence of inflammatory processes playing a role in the pathology of FSHD. The causal gene, DUX4, codes for a transcription factor has been found to be involved in immune system regulation. Gene ontology analysis in myoblasts transduced to express mutant DUX4 found that downregulated genes were enriched in immune pathways, including around 350 genes mostly involved in innate immunity (Geng et al., 2012). These genes were unchanged compared to non-transduced cells, but increased compared to cells transduced with a control lentivirus or unmutated DUX4 lentivirus. This indicates that there was a deficiency in innate immune response to the lentivirus inoculation in cells expressing mutant DUX4 compared to controls, therefore the immune response was being suppressed by the FSHD causative mutation. β-defensin 3 (DEFB103) was seen to be a key regulator of this process, as its expression was significantly induced and it is known to inhibit the transcription of proinflammatory genes (Semple et al., 2011), in this case inhibiting the induction of the immune response to the lentivirus. β-defensin 3 (DEFB103) was also found to be expressed in FSHD cultured muscle biopsies, but not in healthy muscle, and has a potential role in myogenic differentiation. This indicates a potential role for immune mechanisms in FSHD pathogenesis (Geng et al., 2012).
Moreover, there is evidence of altered immune system function in vivo. Circulating IL-6 levels were found to be twice as high in a large cohort of FSHD participants compared to control, and levels also correlated with disease severity measures (Gros et al., 2021). In addition to systemic inflammatory markers there is evidence of inflammation in the muscle, as all FSHD patients had higher levels of mononuclear cell infiltrates in muscle compared to controls and the number of necrotic muscle fibres correlated with number of inflammatory cells (Arahata et al., 1995). Necrotic muscle cells were not invaded by T cells; therefore the immune pathology does appear to be different to DMD. Inflammatory infiltrates containing mainly CD8+ T cells were found in FSHD muscle identified as inflamed using T2-weighted short tau inversion recovery magnetic resonance imaging sequences (Frisullo et al., 2011). FSHD patients with evidence of muscle inflammation had higher levels of active circulating CD12+ T cells compared to control individuals. Peripheral blood mononuclear cells (PBMCs) isolated from FSHD patients spontaneously produced higher levels of IFNγ, TNFα, IL-12, IL-6 and IL-10 compared to PBMCs from normal individuals. There was a correlation between the percentage of circulating active CD8+ T cells and proportion of muscles showing evidence of inflammation, indicating active circulating immune cells may influence muscle inflammation and therefore FSHD progression. Investigating the replacement of muscle with fat in FSHD using the same MRI technique found that the replacement of muscle by fat in FSHD patients occurred over twice as fast in those with evidence of muscle inflammation and this was correlated with the severity signal of the inflammation, supporting the theory that inflammation contributes to muscle damage in FSHD (Dahlqvist et al., 2019). Finally, oxidative stress is higher in FSHD muscles compared to control, and this was correlated with muscle function as measured by 2 min walk distance and endurance limit time (Turki et al., 2012).
Steroids have not been seen to be effective in FSHD. In a trial of prednisone in FSHD patients, there were no significant changes in strength or muscle mass after 12 weeks (Tawil et al., 1997), therefore there is no evidence yet that inflammation in FSHD can be targeted to improve clinical course.

Limb girdle muscular dystrophy
There are multiple types of LGMD, and they vary in evidence for immune system involvement. In type 2B LGMD, muscle inflammation so common that it can often be confused with polymyelitis due to its inflammatory nature and acute presentation. This form is caused by dysferlin deficiency, which like dystrophin (lacking in DMD) has a role in muscle fibre plasma membrane integrity (Bansal et al., 2003;Selcen et al., 2001). It is also normally expressed in monocytes, therefore the mutant version in disease may cause their dysfunction. Peripheral blood monocytes isolated from type 2B LGMD patients and the corresponding mouse model showed increased phagocytic activity compared to control (Nagaraju et al., 2008). There was evidence of active mononuclear cells in both mouse and human muscle, and another study characterised these inflammatory infiltrates in necrotic muscle of those with dysferlin deficiency as consisting of CD4+ T cells, macrophages, CD8+ T cells (Gallardo et al., 2001). In LGMD type 2I, there is also evidence of patients having a DMD-like phenotype with muscle biopsy showing inflammatory infiltrates and responding well to prednisolone treatment (Darin et al., 2007). Lastly, the Sgcb-null mouse model of LGMD type 2E showed inflammation and macrophage infiltration of affected muscles, which was in some cases greater than that seen in the mdx model of DMD (Gibertini et al., 2014). This model also develops more severe muscular dystrophy than the mdx mouse, again emphasising a potential role of the immune system in phenotype severity.

Emery-dreifuss muscular dystrophy
Mutations in LMNA, which cause Emery-Dreifuss muscular dystrophy and some forms of LGMD, were found in 11 of 20 (55 %) patients with inflammatory myopathy onset at 2 years old or younger (Komaki et al., 2011). Steroid treatment showed some improvement in four of eight patients given it. In addition, heart tissue from a mouse model of Emery-Dreifuss muscular dystrophy was shown to be enriched for gene expression of proteins involved in inflammation including the MAPK signalling pathway, but it did not show inflammatory changes upon histological examination (Muchir et al., 2007). Tenascin-C, a marker which may reflect severity of inflammation in myocarditis, was found to be increased in Emery-Dreifuss muscular dystrophy patients (Niebroj-Dobosz et al., 2011).
To summarise, there is evidence of immune system alteration and/or inflammation in all of the main muscular dystrophies. An overview of these alterations is presented in Fig. 1. Despite this, there is a clear benefit to targeting the immune system therapeutically in only DMD, due to lack of evidence in other MDs. As there are multiple associations with immunity or inflammation and disease severity, this role should potentially be further investigated, especially in the context of sleep.

Link to other chronic conditions
Inflammation has long been known to be a risk factor for cardiovascular disease, as local vascular and systemic inflammatory cascades are heavily involved in atherothrombosis. There is a large body of evidence consistently showing that increased markers of inflammation such as IL-6 and TNFα are associated with increased cardiovascular risk (Mendall et al., 1997) and circulating CRP has long been known to be predictive of coronary events (Willerson and Ridker, 2004). Treating inflammation has been seen to mitigate this risk (Willerson and Ridker, 2004;Ajala and Everett, 2020). Cardiovascular disease is one of the greatest causes of early death among those with muscular dystrophies. Approximately 90 % of patients with DMD have cardiomyopathy and heart failure account for 20 % of deaths, and this increases to 50 % in BMD (Finsterer and Stöllberger, 2003). In DM1, approximately 30 % of deaths are due to cardiac complications (Mathieu et al., 1999). As these individuals with muscular dystrophy are already at increased risk of heart problems, it is undesirable to add to this through increased systemic inflammation as a consequence of sleep problems. Treating inflammation, as seen in the studies above, may reduce cardiovascular risk, but this has yet to be validated in muscular dystrophy populations. Treating sleep problems may be a more accessible way to reduce this risk, while increasing quality of life.
Similarly, there are increased rates of insulin resistance and susceptibility to diabetes seen in those with DM1. This is likely as a result of splicing defects due to the causative mutation influencing the insulin receptor transcript (Savkur et al., 2001) with insulin-signalling having been highlighted as a contributor to other symptoms in DM1 (Nieuwenhuis et al., 2019). Insulin resistance is also seen in DMD/BMD (Rodríguez-Cruz et al., 2015) and this risk is compounded by glucocorticoid treatment, which brings risks of insulin resistance and type 2 diabetes itself (Andrews and Walker, 1999). Chronic sleep loss is a risk factor for the development and exacerbation of insulin resistance, with potential mechanisms including increased release of hormones such as adrenaline and cortisol, increased sympathetic nervous system activity, and increased risk of weight gain (Van Cauter, 2011). Another potential cause is inflammation. There is a solid association between chronic proinflammatory states and insulin resistance (De Luca and Olefsky, 2008). TNFα, IL-6 and CRP are all commonly elevated in insulin resistance and diabetes, along with many other immune pathways (De Luca and Olefsky, 2008), while studies in population cohorts demonstrate that insomnia leads to increase risk of type II diabetes (Liu et al., 2022). Those with SDB and OSA show increased insulin resistance, and there is an effect independent of obesity and body mass index (BMI) (Ip et al., 2002;Punjabi et al., 2004). This leads us to consider that sleep and inflammation may be adding to diabetes risk in those with muscular dystrophy. In a study of DM patients found a relationship between serum TNF receptor 2 levels and insulin action, providing a potential link to inflammation in this cohort (Fernández-Real et al., 1999). As with cardiovascular risk, the mitigation of the risk of diabetes in those with muscular dystrophy is very important, in order to avoid adding to their disease burden. Adequate sleep and inflammation control may be an avenue to achieve this. Fig. 1. Alterations in inflammatory and immune markers present across muscular dystrophies. Some of these alterations are central to pathogenesis in dystrophies such as Duchenne muscular dystrophy, while others have an emerging role that has yet to be fully characterised. Not all of these inflammatory alterations have been detected in all muscular dystrophies, but are representative of alterations seen across different muscular dystrophies. See text for further detail on specific changes in each disorder. Created using Biorender.com.

Anti-inflammatory treatments in muscular dystrophies
As mentioned previously, there are immunosuppressant glucocorticoids in use as the mainstay of treatment for DMD. However, their antiinflammatory action may be counteracted to an extent by their tendency to cause obesity as a side effect. Obesity can directly worsen SDB, as the degree of SDB has a linear correlation with BMI (Sawnani et al., 2015). In addition, glucocorticoids reduce bone mineral density in those with DMD, further increasing fracture frequency (Bianchi et al., 2003;Crabtree et al., 2018). They also have the potential to cause sleep disturbance and disorders by disrupting the HPA axis with chronic use (Huscher et al., 2009). There is little research into the specific effects of glucocorticoid treatment on sleep in DMD and this is likely difficult to undertake now as they are the standard of care providing clear benefit. However, there may be potential to trial other agents targeting the immune system both for DMD and other muscular dystrophies which have both inflammatory and sleep involvement. There has been some interest in using glucocorticoids in the treatment of other muscular dystrophies such as LGMD, where it has seen to be safe but there is little evidence of efficacy thus far (Zelikovich et al., 2022).
Other anti-inflammatory agents that have been trialled included the anti-oxidant idebenone which was seen to reduce the loss of respiratory function in those with DMD who were not using glucocorticoids (Buyse et al., 2015). The immunomodulary hormone adiponectin has been found to decrease muscle damage in mdx mice by downregulating the NLRP3 inflammasome, which was further verified in human myotubes (Boursereau et al., 2018). The use of non-steroidal anti-inflammatory drugs (COX inhibitors), including naproxcinod, showed mixed results in the mdx mouse model (Miglietta et al., 2015;Serra et al., 2012). More specific anti-inflammatory agents have seen more success. Infliximab (Remicade®) and Etanercept (Enbrel®), clinically used TNFα blockers have been seen to reduce inflammation (Infliximab) and protect muscle strength (Etanercept) in mdx mice (Grounds and Torrisi, 2004;Pierno et al., 2007). Both have been used effectively to treat psoriasis and arthritis in children and therefore have an established safety profile. In addition, as mentioned earlier there has been improved sleep in cohorts using these medications (Karatas et al., 2017;Taylor-Gjevre et al., 2011), therefore it would be interesting to see the effect of these drugs on sleep and fatigue in DMD and other muscular dystrophies.
Metformin is being considered for use in both DMD and DM. Metformin is widely used in the treatment of diabetes, but its mechanism of action has long been under investigation without many concrete conclusions. However, there are definite effects on the immune system. Metformin suppresses cytokine expression, along with impairing monocyte differentiation and proinflammatory cytokine secretion, and influencing immune pathways such as NF-kB signalling (Rena et al., 2017). In healthy individuals, it was seen to alter neutrophil to lymphocyte ratio which is a marker of inflammation (Cameron et al., 2016). In a trial with DM1 patients, metformin increased performance in the 6-minute walk test indicating an improvement in muscle function (Bassez et al., 2018). In DMD, metformin used in combination with L-arginine reduced oxidative stress in muscles, along with improving clinical scores and two-minute walk distance (Hafner et al., 2016). However, only 5 people were included in this trial and more mixed results were seen in a larger trial (Hafner et al., 2019). There are further positive effects seen in the corresponding mice models (Dong et al., 2021) with a decrease in inflammatory infiltrates in muscle (Mantuano et al., 2018). Some positive effects of metformin on sleep have been noted in those with diabetes and polycystic ovary syndrome (PCOS) (El-Sharkawy et al., 2014;Kajbaf et al., 2014), but it remains to be seen if this effect is maintained in muscular dystrophies.
Modafinil is a stimulant used to treat excessive daytime sleepiness in narcolepsy, shift-work sleep disorder and in those with OSA who still have EDS with the use of CPAP (Kumar, 2008). It is used open label in DM, and trials have shown evidence of improvement in sleep measures in this cohort (Kumar, 2008). Patients have also reported very positive effects (Hilton-Jones et al., 2012). As with metformin, its mechanism of action has not been fully characterised despite extensive research, including monoaminergic effects. Relative to inflammation, there is evidence of modafinil decreasing neuroinflammation in multiple animal models. This action occurs through decreasing recruitment and activation of monocytes and T cells and cytokine production (Zager, 2020). There are also effects on systemic inflammation, with modafinil attenuating the behavioural response to LPS-induced inflammation in mice (Zager et al., 2018). Modafinil prevented an increased in hippocampal TNF and IL-1β induced by sleep deprivation in rats, along with anxiety-like behaviour (Wadhwa et al., 2018). Peripheral proinflammatory cytokine levels are also decreased in response to modafinil in animals (Han et al., 2018;Yousefi-Manesh et al., 2019). Potential mechanisms for this anti-inflammatory action include increased extracellular dopamine, action on dopamine transporter (DAT) expressed by immune cells, and improved blood-brain-barrier functionality (Zager, 2020). Mechanisms of influence on the peripheral immune system remain less clear. Neuroinflammation has yet to be characterised in muscular dystrophies, but due to the clear link to sleep disruption, it may be beneficial characterise this further, especially in studies where modafinil is prescribed in muscular dystrophies for the treatment of daytime sleepiness.

Exercise as therapy targeting inflammation and sleep
Exercise is widely regarded as a non-pharmacological treatment for sleep disorders. Taking OSA as an example, physical activity improves excessive daytime sleepiness, sleep efficiency, number of awakenings and quality of life (Norman et al., 2000). Physical activity increases total sleep time, delays REM onset, and reduces REM sleep (Driver and Taylor, 2000). Exercise is also associated with lower rates of inflammation, as those who are physically active have lower markers of systemic inflammation, such as IL-6, TNFα, and CRP, and this is often independent of BMI (Beavers et al., 2010). It is hypothesised that this may occur through the promotion of SWS in response to increased body temperature, but also through mood stabilising abilities (Uchida et al., 2012). However, due to the close relationship between sleep and inflammation, it is reasonable to suggest that effects on chronic inflammation may also be playing a role. As mentioned previously, when insomnia was successfully treated with specialised CBT or Tai Chi (a form of mindfulness with movement) there was a normalisation of inflammatory signatures including reduced monocyte production of proinflammatory cytokines and gene expression . This indicates there is a normalisation of the proinflammatory phenotype in response to resolution of the sleep disorder.
There is evidence of exercise improving sleep and other symptoms in muscular dystrophies. A systematic review showed aerobic and strength training were effective therapies in those with neuromuscular disorders (Cup et al., 2007). In a large trial of 255 DM1 patients, CBT in combination with graded exercise therapy has been seen to lead to an improvement in ability to carry out daily and social activities, in addition to an improvement in fatigue and daytime sleepiness (Okkersen et al., 2018). Aerobic exercise has also been seen to increase muscle surface area in those with DM1 (Mikhail et al., 2022;Ørngreen et al., 2005). Analysis of blood mRNA levels linked this improvement to alteration in immunological gene signatures towards normal levels (van Cruchten et al., 2022). An increase in sleep quality was seen in FSHD participants who received CBT to increase exercise (Voet et al., 2014). Exercise also has the added benefit of reducing cardiovascular and insulin resistance risk, which as discussed previously can be an issue in those with muscular dystrophies (Henriksen, 2002). This points towards exercise therapy for the multiple benefits that have been seen, including the sleep and inflammation effects.
However, people living with a muscular dystrophy can often be limited in their capacity to exercise, depending on the specific disorder and the stage. Therefore, this may be an option for those with less severe disease symptoms. Along with the previously discussed limitations to NIV treatment of sleep disorders, this leads us to believe that a combination of pharmacological and exercise therapies to treat muscular dystrophies would be desirable to provide benefit to the greatest number of patients.
A summary of the potential benefits of targeting the immune system in muscular dystrophies is given in Fig. 2.

Conclusion
Sleep problems are common across muscular dystrophies, and include sleep disordered breathing, excessive daytime sleepiness, and decreased sleep efficiency and quality. Together they represent a symptom set which has a large quality of life impact in addition to the associated muscle wasting. Relative to these problems, altered inflammation and immunity are also common factors across muscular dystrophies. Frequent observations across all types include increased levels of circulating general proinflammatory markers such as TNFα and IL-6 which support chronic inflammation and enrichment for gene expression associated with immunity/inflammation in blood and tissues. Muscular dystrophies such as DMD, FSHD and types of LGMD have immune cell infiltration in damaged muscle and thus there is a more evident role in pathogenesis. Sleep disturbance drives inflammation and immune system dysfunction, therefore the sleep problems seen in muscular dystrophies may be adding to this immune related pathology seen across disorders. These factors also increase the risk of comorbidities and other chronic diseases such as depression, Alzheimer disease and cardiovascular disease. Consequently, it is desirable to reduce both the burden of sleep problems and inflammation in this cohort. Some success has been seen with drugs such as modafinil and metformin which have anti-inflammatory action, but we have yet to see the use of more targeted, specific anti-inflammatory agents. Lifestyle interventions such as increased exercise also appear to have the potential to normalise sleep and inflammation, but they may be limited to those with less severe disease. Further research may better characterise the ability of these interventions to have meaningful impacts on sleep and inflammation, and should be considered going forward due to the lack of disease modifying interventions in muscular dystrophies.

Data Availability
No data was used for the research described in the article. Fig. 2. The relationship between muscular dystrophies, inflammation, and sleep, and how it can be targeted therapeutically along with the potential benefits of this approach. There are a wide range of mechanisms through which sleep and inflammation can be altered, including pharmacotherapies in the form of anti-inflammatory agents (e.g. anti-TNFα agents), stimulants (e.g. modafinil) and metformin. Exercise can also influence both factors as a lifestyle intervention, and nocturnal non-invasive ventilation can specifically target sleep disordered breathing. Improving sleep and inflammatory markers in those with muscular dystrophies has the potential for a wide range of benefits such as reducing the risk of co-morbid chronic conditions, increasing quality of life and reducing fatigue, along with possibly slowing disease progression. Created using Biorender.com.