Muscle atrophy in mechanically-ventilated critically ill children

Abstract

Importance: ICU-acquired muscle atrophy occurs commonly and worsens outcomes in adults. The incidence and severity of muscle atrophy in critically ill children are poorly characterized.

Objective: To determine incidence, severity and risk factors for muscle atrophy in critically ill children.

Design, setting and participants: A single-center, prospective cohort study of 34 children receiving invasive mechanical ventilation for ≥48 hours. Patients 1 week- 18 years old with respiratory failure and without preexisting neuromuscular disease or skeletal trauma were recruited from a tertiary Pediatric Intensive Care Unit (PICU) between June 2015 and May 2016. We used serial bedside ultrasound to assess thickness of the diaphragm, biceps brachii/brachialis, quadriceps femoris and tibialis anterior. Serial electrical impedance myography (EIM) was assessed in children >1 year old. Medical records were abstracted from an electronic database.

Exposures: Respiratory failure requiring endotracheal intubation for ≥48 hours.

Main outcome and measures: The primary outcome was percent change in muscle thickness. Secondary outcomes were changes in EIM-derived fat percentage and "quality".

Results: Of 34 enrolled patients, 30 completed ≥2 ultrasound assessments with a median interval of 6 (IQR 6-7) days. Mean age was 5.42 years, with 12 infants <1 year (40%) and 18 children >1 year old (60%). In the entire cohort, diaphragm thickness decreased 11.1% (95%CI, -19.7% to -2.52%) between the first two assessments or 2.2%/day. Quadriceps thickness decreased 8.62% (95%CI, -15.7% to -1.54%) or 1.5%/day. Biceps (-1.71%; 95%CI, -8.15% to 4.73%) and tibialis (0.52%; 95%CI, -5.81% to 3.40%) thicknesses did not change. Among the entire cohort, 47% (14/30) experienced diaphragm atrophy (defined a priori as ≥10% decrease in thickness). Eighty three percent of patients (25/30) experienced atrophy in ≥1 muscle group, and 47% (14/30)-in ≥2 muscle groups. On multivariate linear regression, increasing age and traumatic brain injury (TBI) were associated with greater muscle loss. EIM revealed increased fat percentage and decreased muscle "quality".

Conclusions and relevance: In children receiving invasive mechanical ventilation, diaphragm and other skeletal muscle atrophy is common and rapid. Increasing age and TBI may increase severity of limb muscle atrophy. Prospective studies are required to link muscle atrophy to functional outcomes in critically ill children.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. Flowchart of the study population.

Fig 2. Loss of muscle thickness in critically-ill children.

A. Percent change in muscle thickness between the first two ultrasound assessments in the entire study cohort. Dashed red line indicates a 10% decrease. Gray line at zero is shown for clarity. Box boundaries represent 25th and 75th percentiles, whiskers– 5th and 95th percentiles. B. Percent of patients experiencing muscle atrophy, defined as >10% loss of muscle thickness. C. Frequency distribution of patients by the number of muscle groups showing atrophy.

Fig 3. Change in muscle thickness and correlation with age in critically-ill children.

A-C. Percent change in muscle thickness in the four examined muscle groups in all patients (A), infants < 1 yr of age (B), and children > 1 yr of age (C). Scatter data are shown with mean ± SD. The red line indicates a 10% decrease in thickness. The gray line indicates zero for convenience. *WSRT p < 0.05. D-F. Simple linear regression of percent change in muscle thickness as a function of age in months for the Biceps (D), Tibialis (E) and Quadriceps (F). Red dashed lines indicate 95% CI for the slope.

Fig 4. Effect of age group and TBI on change in muscle thickness in critically-ill children.

A. Change in muscle thickness in infants < 1 yr of age (open) and in children > 1 yr of age (closed). *p = 0.084, **p = 0.025, ***p = 0.002, paired t-tests. To account for multiple comparisons, discovery (q) was determined using the two-stage linear step-up procedure of Benjamini, Krieger and Yekuteli (BKY) with FDR = 10%; *q = 0.061, **q = 0.028, ***q = 0.005. B. Change in muscle thickness in patients without TBI (open) and with TBI (closed). *p = 0.022, **p = 0.038, t-tests. Similar to A, discovery was determined with BKY and FDR 10%; * and **q = 0.043. For both A and B, dashed red line indicates a 10% decrease, and solid gray line–zero. Box boundaries represent 25th and 75th percentiles, whiskers– 5th and 95th percentiles.

Fig 5. Changes in EIM-derived values in critically-ill children.

A. Change in muscle “quality” (arbitrary values). B. Change in fat percentage. Scatter data are shown with mean ± SD. * p < 0.05, ** p < 0.01, paired t-tests.