Table 2.
Age and Sex-Adjusted Associations of Individual Baseline and Critical Illness Variables With Fatigue Symptoms Over 6- and 12-Month Follow-up (N = 732)
| Variable | Mean Difference (95% CI) in Fatiguea [Positive Value = Less Fatigue] |
|---|---|
| Baseline status prior to ICU admission | |
| Age (per 10 y) | –0.7 (–1.5 to 0.1) |
| Male sex | 7.4 (5.1 to 9.8) |
| White race | –2.7 (–5.8 to 0.5) |
| Employment (full- or part-time vs unemployed) | 6.0 (3.6 to 8.5) |
| Diabetes comorbidity | –0.3 (–3.1 to 2.6) |
| Critical illness status | |
| Medical ICU | 0.2 (–2.2 to 2.6) |
| APACHE III score (per 10 points) | 0.5 (0.03 to 1.0) |
| Treatment in ICU stay with | |
| Vasopressor | –1.9 (–4.5 to 0.6) |
| Corticosteroidb | –2.1 (–5.1 to 1.0) |
| Neuromuscular blocker | 1.4 (–2.4 to 5.2) |
| ICU length of stay (per 5 d) | –0.5 (–1.1 to 0.1) |
Each row is a separate regression model evaluating the age- and sex-adjusted association of the variable named in each row with fatigue, using a longitudinal time-averaged random effects regression model. For the variables of age and male sex, the regression model only adjusted for sex and age, respectively. Fatigue was measured by using the transformed score from the validated Functional Assessment of Chronic Illness Therapy scale (range, 0-100), with higher scores representing less fatigue. See Table 1 legend for expansion of abbreviation.
Missing data: race, n = 22; employment, n = 36; APACHE III score, n = 21; ICU length of stay, n = 3. Not all of the parent studies collected data for opioid, corticosteroid, or neuromuscular blocker use. For opioid and neuromuscular blocker, N = 495 with no missing data; and for corticosteroid, N = 634 with missing data, n = 22.
Defined as receiving > 20 mg of methylprednisolone-equivalents on one or more days in the ICU.