Pulmonary recovery from COVID-19 in patients with metabolic diseases: a longitudinal prospective cohort study

Abstract

The severity of coronavirus disease 2019 (COVID-19) is related to the presence of comorbidities including metabolic diseases. We herein present data from the longitudinal prospective CovILD trial, and investigate the recovery from COVID-19 in individuals with dysglycemia and dyslipidemia. A total of 145 COVID-19 patients were prospectively followed and a comprehensive clinical, laboratory and imaging assessment was performed at 60, 100, 180, and 360 days after the onset of COVID-19. The severity of acute COVID-19 and outcome at early post-acute follow-up were significantly related to the presence of dysglycemia and dyslipidemia. Still, at long-term follow-up, metabolic disorders were not associated with an adverse pulmonary outcome, as reflected by a good recovery of structural lung abnormalities in both, patients with and without metabolic diseases. To conclude, dyslipidemia and dysglycemia are associated with a more severe course of acute COVID-19 as well as delayed early recovery but do not impair long-term pulmonary recovery.

Figure 1

Preexisting metabolic comorbidities are related to the course of acute COVID-19. Bars depict the prevalence of obesity (A), diabetes mellitus (B), and dyslipidemia (C) before COVID-19 onset in patients with mild, moderate and severe acute COVID-19. N_total = 145, N_mild = 34, N_moderate = 38, N_severe = 73. P-values are reported according to Chi-Square tests.

Figure 2

Dysglycemia and dyslipidemia at early post-acute follow-up after COVID-19. Patients were prospectively followed after acute COVID-19 and the presence of dysglycemia (A) and dyslipidemia (B) were assessed at 60 days post-COVID-19 follow-up. Bars depict the prevalence of dysglycemia and dyslipidemia in patients with mild, moderate and severe acute COVID-19. N_total = 145, N_dysglycemia = 62, N_dyslipidemia = 32. P-values are reported according to Chi-Square tests.

Figure 3

Adipocytokine concentrations at early post-acute follow-up are related to acute COVID-19 severity. Adiponectin (A) and leptin (B) serum concentrations at 60 days post-COVID-19 follow-up. Patient subgroups are shown according to the acute COVID-19 severity. Error bars depict one standard error. N_mild = 34, N_moderate = 38, N_severe = 73. P-values were determined using the Kruskal–Wallis test.

Figure 4

Serum markers of thrombo-inflammation in patients with dysglycemia and dyslipidemia at post-COVID-19 follow-up. COVID-19 patients were prospectively followed after acute COVID-19 and concentration of IL-6 (A, E), d-dimer (B, F), ferritin (C, G), and adiponectin (D, H) were prospectively assessed at 60, 100, 180, and 360 days after the onset of COVID-19. Subgroups are presented according to the presence of dysglycemia (A-D) or dyslipidemia (E–H), and statistical differences between these subgroups at each time point are indicated according to Mann Whitney-U tests *P < 0.05 and ***P < 0.001. Points depict means, error bars indicate standard error (SE). N_60days = 145, N_100days = 138, N_180days = 119, N_360days = 92.

Figure 5

Pulmonary recovery after COVID-19 in patients with dysglycemia or dyslipidemia. Structural pulmonary impairment was assessed with computed tomography (CT) at four post-COVID-19 follow-ups. The severity of structural lung abnormalities was graded from 0 to 25 points (the higher the more severe) as described in the methods section. Time-dependent structural lung recovery according to the presence/absence of dysglycemia (A) or dyslipidemia (B) are presented. (C) Structural lung recovery time resulting in a 50 percent (τ) reduction of CT lung impairment as compared to the early post-acute follow-up in individuals with dysglycemia, dyslipidemia, and the total cohort are shown, whereas we found no significant differences in lung recovery time between the investigated subgroups. Mann–Whitney-U test (A, B) and Kruskal–Wallis test (C) were performed to analyze statistical differences between subgroups for each time point. Points indicate means, error bars depict standard error (SE). N_60days = 145, N_100days = 138, N_180days = 119, N_360days = 92.

Figure 6

Logistic ordinal modeling of chest CT abnormality severity at the 60-day follow-up as a function of inflammatory and metabolic parameters. Chest CT abnormalities were classified as none (CT severity score [CTSS]: 0), mild (CTSS: 1–5), moderate (CTSS: 6–10) and severe (CTSS: $\ge$ 11). Effects of inflammatory (C-reactive protein [CRP], interleukin 6 [IL6], d-dimer [DDimer], ferritin [FT]), metabolic biomarkers (triglycerides [TG], high-density lipoprotein [HDL], adiponectin [ADIPOQ], leptin [LEP]) and metabolic disorders (obesity, dyslipidemia and dysglycemia) on chest CT abnormality severity at the 60-day follow-up were assessed by ordinal logistic regression. Odds ratio (OR) with 95% confidence intervals (CI) for the explanatory variables in univariable models (A), models adjusted for age class and sex (B) and models adjusted for age class, sex and acute COVID-19 severity (C) are shown in Forest plots. Point colour codes for significance and model estimate sign. Points are labeled with OR and 95% CI values.