Clinical review: intensive care unit acquired weakness

ICUAW is a frequent problem. The reported incidence varies depending on the patient population studied and on the timing of the evaluation. Weakness at awakening was found to be present in 26–65 % of patients who were mechanically ventilated for 5–7 days, respectively [6, 7], and 25 % of these remained weak for at least another 7 days after awakening [8]. Among long-term ventilated patients (≥10 days) ICUAW was diagnosed in up to 67 % [9]. In another study 11 % of patients treated in the ICU for at least 24 hours developed ICUAW [10], and when the ICU stay increased to 7–10 days ICUAW at awakening was present in 24–55 % of patients [10‐12]. In patients suffering from acute respiratory distress syndrome (ARDS), an ICUAW incidence of 60 % has been reported at the time of awakening [13], and at hospital discharge this incidence was still 36 % [14].

The mechanisms underlying ICUAW are complex and involve functional and structural alterations in both the muscles and the nerves.

In CIP, the pathological finding is axonal degeneration [15]. The pathogenesis of such axonal degeneration remains incompletely understood, in part explained by the invasiveness of nerve biopsies. Factors that play a role are microvascular changes in the endoneurium evoked by sepsis [16], which promotes vascular permeability and allows penetration of toxic factors into the nerve ends [17]. The endoneural edema resulting from increased permeability may impair energy delivery to the axon followed by axonal death. Direct toxic effects and mitochondrial dysfunction evoked by hyperglycemia may contribute to this process [18‐20]. Animal experiments also identified channelopathy as a cause of rapidly occurring but reversible neuropathy [21]. Whether such a channelopathy also occurs in ICU patients, as a functional precursor of axonal degeneration that is found later in the course of critical illness, remains unclear [22].

In CIM, several factors are thought to play a role by negatively affecting muscle structure and function, all interacting in a complex manner [22]. A detailed overview of mediators and molecular mechanisms involved was published recently [23]. First, muscle atrophy can occur quite early during critical illness [24]. Muscle atrophy is brought about by increased breakdown and decreased synthesis of muscle protein [24, 25]. Several processes during critical illness may promote such wasting of muscle protein, which preferentially involves myosin [25, 26]. These include inflammation, immobilization, the endocrine stress responses, the rapidly developing nutritional deficit, impaired microcirculation, and denervation [22, 27, 28]. Key proinflammatory mediators involved include tumor necrosis factor alpha, interleukin-1, and interleukin-6 [23]. Recently, another stress-induced cytokine, growth and differentiation factor-15 (GDF-15), a member of the transforming growth factor beta super family, was identified as a mediator of muscle atrophy during critical illness [29]. In addition to muscle atrophy, other factors may contribute to decreased muscle function. Muscle membrane inexcitability induced by sodium channel dysfunction is an early finding in CIM [30] and predicts the development of clinical muscle weakness [31]. An altered intracellular calcium homeostasis, affecting the excitation–contraction coupling, has also been shown to contribute to reduced contractility in animal sepsis models [32, 33]. The muscles of septic patients show signs of bioenergetic failure, comprising oxidative stress, mitochondrial dysfunction, and ATP depletion [34]. Finally, muscle atrophy and weakness are not synonymous [22]. Indeed, muscle quality might be even more important than muscle mass in determining muscle function and the development of ICUAW [35]. Autophagy, which is a cellular housekeeping system that mediates removal of damaged large organelles and protein aggregates, is an important factor in maintaining muscle fiber integrity [36], and deficient autophagy may have a crucial role in the development of ICUAW [35].

Several risk factors for the development of neuromuscular complications in the ICU have been reported. At the time when CIP was first described, sepsis, systemic inflammatory response syndrome (SIRS), and multiple organ failure (MOF) were considered to play a key role. Indeed, in the early case reports, sepsis was frequently present [2, 37]. Hence, subsequent studies specifically focused on patients with sepsis and organ failure and reported a high incidence of neuromuscular complications [38‐40]. ICUAW was therefore considered to be the peripheral component of a general neuromuscular organ failure evoked by sepsis, with septic encephalopathy being the central component. An independent role for sepsis [35], bacteremia [10, 41], SIRS [40, 42], and MOF [8] was subsequently suggested by several prospective studies. Indirectly, a role for sepsis and MOF in the development of neuromuscular complications of critical illness was further supported when other independent risk factors were identified, such as the use of vasopressors [41, 43] or aminoglycosides [10], certain mediators of inflammation [44], and the presence of septic encephalopathy [39]. The risk associated with a higher severity of illness, as reflected by the Acute Physiology and Chronic Health Evaluation score, was demonstrated in several studies [10, 35, 42, 45]. Hyperglycemia was also found to be an independent risk factor for electrophysiological [38] and clinical [10] signs of ICUAW, and increasing insulin dose reduced signs of ICUAW [46]. These data offer the potential to prevent ICUAW by controlling glycemia with insulin (see ‘Prevention and therapy’). Providing early mobilization effectively reduced ICUAW [46], which is in line with the documented risk reduction in clinical or electrophysiological manifestations of ICUAW, associated with several indirect reflections of the duration of immobilization, such as duration of ICU stay [38, 43] and of mechanical ventilation prior to awakening [8], and the time until awakening and clinical assessment of muscle strength [35]. In one study, the duration of bed rest was the only factor consistently associated with weakness 3–24 months post ICU discharge [14]. This underscores the potential importance of treatments aiming to reduce the duration of immobilization in the ICU (see ‘Prevention and therapy’).

After a first case report of pronounced weakness in a patient treated with corticosteroids and neuromuscular blocking agents [47], multiple others followed. Denervated and steroid-treated animals showed muscle changes similar to those observed in critically ill patients [48]. Some prospective studies indeed suggested that corticosteroids [8, 35, 45] and (prolonged) neuromuscular blocking agents [18, 35, 39] may contribute to CIP/CIM or ICUAW. However, other prospective studies, of which some were randomized controlled trials, could not confirm such a role for corticosteroids [10, 13, 14, 18, 39, 40, 42‐44, 46, 49] or for neuromuscular blocking agents [14, 50]. These apparently contradicting findings suggest that the relationship between these drugs and neuromuscular complications is more complex and depends on other factors such as dose, timing, and concomitant glycemic control [18]. Age was also identified as an independent risk factor for ICUAW [35, 46]. This may to some extent reflect the importance of the premorbid physiological muscle reserve, although the actual premorbid functional status is determined by many other factors. The premorbid functional status is often difficult to document due to the unplanned nature of ICU admissions. One report identified parenteral nutrition as a risk factor for CIP [39]. The effect of nutritional strategies as a modifiable factor is further discussed in ‘Prevention and therapy’. Other factors identified only in single studies might be incidental findings (female sex) [8] or related to other risk factors (hyperosmolality [39], hypoalbuminemia [38], renal replacement therapy [41]) and await confirmation.

Only few studies specifically addressed risk factors for respiratory weakness in the ICU. These studies demonstrated that respiratory muscle weakness is associated with infection or sepsis [51‐53], disease severity [53], and peripheral weakness [51]. The phrenic nerve and diaphragm also show similar electrophysiologic abnormalities as the peripheral nerves and muscles [54], so these data suggest that respiratory weakness is indeed part of ICUAW [51, 53]. Additionally, the duration of mechanical ventilation may contribute to weakness [55, 56] and atrophy of the diaphragm [56, 57]. This time-dependent and early development of diaphragmatic atrophy and dysfunction is also labeled “ventilator-induced diaphragmatic dysfunction” (VIDD) [58].

ICUAW is characterized by symmetrical and flaccid weakness of the limbs, which is more pronounced in the proximal muscles than in the distal muscles [8, 59]. Facial and ocular muscles are often spared [60]. Patients with ICUAW therefore typically respond to a painful stimulus with facial grimacing but with no or minimal withdrawal of the limbs. Tendon reflexes are generally reduced, although these can be normal. In the case of coexistent CIP, sensory symptoms may be present, including reduced or absent sensitivity to pain, temperature, and vibration. However, examination of the sensory function is generally difficult in critically ill patients.

When ICUAW is present, the respiratory muscles are often affected [51]. This contributes to delayed weaning from mechanical ventilation, which is often the clinical problem with which these patients present.

Aggressive treatment of sepsis is considered to be a cornerstone in the prevention of ICUAW. Although anti-inflammatory therapies in sepsis in general have been disappointing and specific neuromuscular effects have not been studied, targeting newly discovered inflammatory mediators involved in muscle atrophy, such as GDF-15, may open new perspectives. Further preventive strategies include the reduction or avoidance of other known risk factors but only few randomized trials addressing these are available.

Insulin treatment aimed at normalizing glycemia as compared with tolerating hyperglycemia up to the renal threshold significantly reduced the incidence of electrophysiological signs of CIP/CIM and the need for prolonged mechanical ventilation in long-stay medical [18] and surgical [43] ICU patients. A subsequent multicenter trial [105] found increased mortality in patients treated to strict normoglycemia as compared with those receiving insulin to target somewhat higher blood glucose levels. Consequently, the optimal target for blood glucose remains a matter of debate and ongoing efforts are being made to more safely control glucose in the ICU [106].

Reducing the duration of immobilization is another potentially important target in the prevention of ICUAW. This target can be realized by decreasing the levels of sedation. Approaches aimed at sedating patients to the minimal level needed for comfort and safety clearly have overall beneficial effects [107]. The actual impact of such strategies on ICUAW has not been studied. Reducing the duration of immobilization with early physical therapy adjusted to the medical situation, muscle strength, and level of cooperation is another strategy that was shown to be safe and feasible in the ICU and improved outcome [108]. Additionally, quadriceps strength at hospital discharge improved in long-stay patients receiving passive or active exercise training using a bedside ergometer, as compared with patients receiving standard physiotherapy [109]. This training involved 20 minutes of bedcyling 5 days per week, from the 5th day following ICU admission. Both functional status and health-related quality of life at hospital discharge were also improved [109]. Early mobilization and occupational therapy within 72 hours of initiation of mechanical ventilation in addition to daily sedation interruption improved outcome, including functional status, although the incidence of ICUAW was not reduced [110]. The training consisted of an individually tailored daily program starting with passive range-of-motion exercises in unresponsive patients, progressing to active range-of-motion exercises, bed mobility, sitting upright, transfer training, and eventually walking. A secondary analysis of this trial reported that, after correction of other risk factors, both early mobilization and increasing insulin dose independently prevented ICUAW [46]. Despite clear benefits from early mobilization of patients in the ICU, daily practice appears to vary significantly among centers and clinical settings [111, 112] and many barriers exist, hampering a wide implementation. Identification of subgroups of patients who are most likely to benefit would allow a better allocation of resources [113]. Patients with ICUAW might represent a group of particular interest but data supporting such an approach are currently lacking [114]. Analogous to the peripheral muscles, early mobilization of the diaphragm allowing for spontaneous breathing while avoiding overactivation is considered standard practice to minimize respiratory muscle weakness [56, 115].

A substantial number of ICU patients are not able to participate in active mobilization early. In this population, electrical muscle stimulation (EMS) could theoretically be used to activate and cultivate muscles during this phase. Several studies suggest potential beneficial effects of EMS in critically ill patients. The evidence remains inconclusive, however, because of equivocal results and several methodological issues [116, 117].

Finally, malnutrition was originally considered a major contributor to ICUAW and correction of the nutritional deficit using parenteral feeding appeared evident [2]. This paradigm was recently challenged by Puthacheary et al. [24], who reported that increased protein delivery during the first week in the ICU was associated with more pronounced muscle wasting. In addition, avoiding parenteral nutrition during the first week in the ICU, as compared with early supplementation of deficient enteral feeding, reduced the incidence of weakness and enhanced recovery thereof [35]. This beneficial effect was explained by enhanced autophagic quality control in the muscle fibers, whereas markers of atrophy remained unaffected. These findings suggest that the early catabolic phase during critical illness cannot be averted by artificial nutrition. Certain catabolic pathways, in particular autophagy, may be crucial to maintain muscle quality and function. Hence, targeting autophagy may be worthwhile to investigate further as a future treatment strategy. Also, early “trophic” feeding via the enteral route instead of full enteral feeding during the first week in the ICU did not have detrimental effects on physical function or strength measurements up to 1 year following ARDS [98].