The cross-sectional area of the erector spinae muscle is an adverse indicator for patient with acute exacerbation of chronic obstructive pulmonary disease

Characteristics of patients with AECOPD

Figure 2 showed the flow chart of the study. There were 236 AECOPD patient included in the present study. According to the quartile of ESM_CSA, patients were divided into 2 groups: less than the quartile (< 23.65cm², lower group), and more than quartile (≥ 23.65cm²,normal group), The demographic characteristics of the 236 patients who were included in the data analysis are presented in Table 1. There were 59 and 177 patients in the lower group and normal groups respectively. There were more males, current smokers and less comorbidities with coronary heart disease (CHD), diabete in the ESM_CSA normal group(p < 0.05). And there were more subjects with old age and frequent exacerbation in the ESM_CSA lower group. There were nonsignificant associations of ESM_CSA with PaO2, pH, BMI, PaCO2, lung function, comorbidity with hypertension, congestive heart failure, renal dysfunction, exacerbations during preceding year and respiratory failure (P > 0.05).

ESM_CSA and 1-year rehospitalization for AECOPD

A total of 112 patients suffered from 273 hospitalizations for AECOPD within 1 year after discharge from the hospital. The survival plot of the time to first hospitalization for AECOPD according to the erector spinae muscle area on admission is presented in Fig. 3 (p = 0.001 by the log-rank test).

The survival plot of the time to first hospitalization for AECOPD according to the levels of the cross-sectional area of the erector spinae muscle.

Univariate cox regression analysis showed that ESM_CSA, PaCO2, BMI, frequent exacerbation and respiratory failure were associated with the 1-year first hospitalization for AECOPD (Table 2).

Multivariate cox regression analysis showed that only ESM_CSA and BMI were associated with the 1-year first hospitalization for AECOPD (Table 2).

For the rate of hospitalization for AECOPD, in univariate poisson regression analysis, ESM_CSA were associated with the risk of hospitalization for AECOPD (IRR = 0.57, 95% CI 0.45–0.73, P = 0.000). In the multivariate analyses, ESM_CSA remained an independent predictor of a higher risk of hospitalization for AECOPD (IRR 0.56, 95% CI 0.43–0.72) at the 1-year follow-up. The other factors associated with the rate of hospitalization for AECOPD were PaO2, PaCO2, frequent exacerbation in the previous years, respiratory failure, and lung function status (Table 3).

ESM_CSA and 1-year mortality

Seventeen patients died within 1 year after discharged from the hospital, and the 1-year mortality were 15.3% and 4.5% for the ESM_CSA lower group, and normal groups respectively. The survival plot of the time to 1-year death according to the levels of ESM_CSA on admission is presented in Fig. 4 (p = 0.007 by the log-rank test).

The survival plot of the time to death for AECOPD according to the levels of the cross-sectional area of the erector spinae muscle.

In univariate cox regression analyses, age, PaO2, CHD, CHF were associated with 1-year mortality, whereas PaCO2, and BMI were not. However, multivariate cox regression analysis confirmed that age, PaO2, CHF were associated with 1-year mortality (Table 4).

Both univariate (HR = 0.29 95% CI: 0.11–0.75, P = 0.011) and multivariate cox regression analyses (HR = 0.35, 95% CI: 0.12–0.99, P = 0.048) showed that ESM_CSA were associated with 1-year mortality (Table 4).

Sensitivity analysis

We performed sensitivity analysis to evaluate the effect of the ESM_CSA adjust to gender, gender and height on 1-year mortality and rehospitalizations for AECOPD. The sensitivity analysis showed that when ESM_CSA adjust to gender, lower ESM_CSA were a independent risk factor for first rehospitalization and rehospitalization rate within 1 year (Table 5). When ESM_CSA adjust to gender and height, lower ESM_CSA-I were an independent risk factor for first rehospitalization and rehospitalization rate within 1 year (Table 5). At the same time, univariate cox regression analyses showed that the lower ESM_CSA adjust to gender were risk factor for 1-year mortality. However, multivariate cox regression analyses showed ESM_CSA adjust to gender were not risk factor for 1-year mortality (Table 5). Univariate cox regression analyses showed that the lower ESM_CSA-I were risk factor for 1-year mortality. However, multivariate cox regression analyses not confirmed it (Table 5).