Post-COVID-19 syndrome and insulin resistance 20 months after a mild COVID-19

ObjectiveSARS-CoV-2 infection is associated with impaired glucose metabolism. Although the mechanisms are not fully understood, insulin resistance (IR) appears to be a central factor. Patients who had a severe acute phase, but even asymptomatic or with mild COVID-19, have an increased risk of T2DM. After the acute phase, post-COVID-19 syndrome (PCS) also seems to be related to this metabolic disturbance, but there is a paucity of studies. This study aims to evaluate a possible relationship between PCS and IR after mild COVID-19 and, if confirmed, whether there are differences by sex. Subjects and methodsRetrospective observational cohort study including subjects who had mild COVID-19 between April and September 2020 in a community setting. None had been vaccinated against SARS-CoV-2 at inclusion, and previous T2DM and liver disease were exclusion criteria. Patients who met NICE criteria were classified as PCS+. Epidemiological and laboratory data were analysed. Three assessments were performed: 1E (pre-COVID-19, considered baseline and reference for comparisons), 2E (approximately 3 months after the acute phase), and 3E (approximately 20 months after the acute phase). A triglyceride-to-glucose (TyG) index [&ge;]8.74 was considered IR. Albumin-to-globulin ratio (AGR) and lactate dehydrogenase (LDH) were assessed as inflammatory markers. Bivariate analyses were performed, using nonparametric and repeated measures tests. A subsample without metabolic disorder or CVD (age<median, BMI<25 kg/m2, elevated AGR, TyG index=7.80 [0.5]) was generated to reasonably rule out prior baseline IR that could bias the results. The relationships between PCS and TyG in 3E (TyG3) were modeled in 8 multiple regressions, stratifying by sex and BMI combinations. ResultsA total of 112 subjects (median [IQR] of age= 44 [20] years; 65 women) were analysed. Up to 14.3% was obese and 17% was hypertensive. Significant increases between 1E and 3E were registered regarding (i) basal glycemia (BG), 87 [14] mg/dL vs. 89 [14]; p=0.014, (ii) TyG index (8.25 [0.8] vs. 8.32 [0.7]; p=0.002), and (iii) LDH in 3rd tertile (16.1% vs 32.1%; p=0.007). A total of 8 previously normoglycemic subjects, showed BG2 or BG3 >126 mg/dL. The subgroups with IR highest prevalence at 3E were those of BMI [&ge;]25 kg/m2 and PCS+. The subgroup without CVD presented a significant increase in the TyG index (TyG1=7.80 [0.1] vs. TyG3= 8.28 [0.1]; p=0.017). LDH1 was significantly correlated with TyG3 in both sexes (rho=0.214 in women, rho=0.298 in men); in contrast, LDH2 and LDH3 did not present such an association. In multivariable analysis, PCS has shown to be an independent and predictive variable of TyG index in women with BMI<25 kg/m2, after adjustment for age, hypertension, BMI, Charlson comorbidity index, AGR1, AGR2, LDH1, number of symptoms of acute COVID-19, and number of days of the acute episode ({beta}=0.350; p=0.039). ConclusionsPCS has played a secondary role in predicting IR, showing a modest effect compared to BMI or prior hypertension. A significant increase in IR has been noted 20 months after mild COVID-19, both in cases of previous baseline IR and in those without previous IR. Basal serum LDH has shown to be predictive of current TyG, regardless of elevated LDH after SARS-CoV-2 infection. There were profound differences between women and men, confirming the need for a sex-stratified analysis when addressing the relation between PCS and glycemic alterations.