Impaired antibody responses were observed in patients with type 2 diabetes mellitus after receiving the inactivated COVID-19 vaccines

In this study, 89 adult patients with T2DM and 100 adult healthy individuals were enrolled between July 29, 2021, and October 22, 2021, at the Second Affiliated Hospital of Chongqing Medical University. All participants had received a full-course of inactivated COVID-19 vaccines (CoronaVac and BBIBP-CorV). The eligibility criteria were as follows: diagnosed with T2DM prior to vaccination for the patients group; no recorded disease status for HCs; and 21–105 days had passed after the second dose of inactivated vaccines for all participants. The main exclusion criteria were: SARS-CoV-2 infection or a history of suspected clinical SARS-CoV-2 infection before receipt of the first dose of vaccines; malignant tumor; pregnancy; immunosuppressant administration within 6 months; and acute inflammatory diseases. This study followed the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of the Second Affiliated Hospital of Chongqing Medical University (No. 133). All participants gave their informed consent before their inclusion in the study.

Questionnaires were used to record adverse events within 7 days and 30 days after the second dose of vaccines (We did not record adverse events after the first dose of vaccines). The classification of adverse events referred to the classification scale issued by the National Medical Products Administration of China (version 2019).

The titers of anti-receptor binding domain IgG (anti-RBD-IgG) and neutralizing antibody (NAbs) were detected by capture chemiluminescence immunoassays (CLIA). Samples were evaluated with MAGLUMI 2000 (Snibe, Shenzhen, China). The cutoff values of anti-RBD-IgG and NAbs were 1.0 AU/mL and 0.15 ug/mL, respectively. (The converted log values were used for analysis, and the cutoff values of anti-RBD-IgG and NAbs after conversion were 0 AU/mL and 0 ug/mL, respectively). The antibodies were considered seropositive when the detected values were greater than the corresponding cutoff values; otherwise, they were considered seronegative.

Fresh peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll density gradient centrifugation. After that, the PBMCs were stained with multiple fluorescent-coupled antibodies. Finally, the samples were assessed by flow cytometry. Details and strategy of the procedure can be obtained from supplementary materials.

All continuous variables were checked for normality assumption. Median and interquartile range (IQR) were used to describe continuous variables, while frequency or percentage was used to describe categorical variables. Continuous variables were compared using the T-test or Mann–Whitney U test for two groups and Kruskal–Wallis test for three groups (Bonferroni correction was used for the results of multiple comparisons). Categorical variables were compared using the chi-square or Fisher’s exact test; in addition, we used multiple linear regression analysis to obtain factors that may affect antibody titers. Two-tailed P values were reported, with P < 0.05 indicating statistical significance (*P < 0.05; **P < 0.01; ***P < 0.001). Missing data were not interpolated. Calculations were performed using SPSS 26 software and GraphPad Prism 9.0

This study recruited a total of 189 participants: 89 patients diagnosed with T2DM and 100 adult healthy individuals. There were no statistically significant differences in sex, age, BMI, and days between the two groups. Notably, the median age of participants in both groups was over 60 years old. In the T2DM patients group, 75.3% received the CoronaVac, 23.6% the BBIBP-CorV, and 1.1% the BBIBP-CorV + CoronaVac; in HCs, the percentages were 54.0%, 39.0%, and 7.0%, respectively. The median HbA1c (Hemoglobin A1c) in the T2DM group was 7.4%. In addition, FPG (Fasting Plasma Glucose) in the T2DM group was significantly higher than that in HCs (p < 0.001). The red blood cell count, white blood cell count, and lymphocyte count were similar between the two groups, while the platelet count in the T2DM group was lower than that in the HCs group (Table 1).

The overall incidence of adverse events within 7 days of vaccination was 6.7% in the T2DM group and 6% in the HCs group. Within 30 days of vaccination, the overall incidence of adverse events in the T2DM group and HCs group was 5.6% and 8%, respectively. The differences between the two groups were not statistically significant (7 days: p = 0.835; 30 days: p = 0.518). Local adverse events mainly manifested as pain, swelling, and itching at the site of vaccination. Furthermore, systemic adverse events were uncommon. In the T2DM patients group, systemic adverse events, including lethargy and cough, were reported in 5 patients. However, only 1 participant developed systemic adverse events in HCs, namely, lethargy. Of note, no grade 3 and 4 adverse events were observed among the 189 participants (Table 2). Briefly, the incidence of adverse events in T2DM patients who were vaccinated with a full-course of CoronaVac or BBIBP-CorV (or both) was similar to that of HCs, with only a few patients who developed mild adverse reactions.

Subsequently, we detected the anti-RBD-IgG and NAbs levels of participants after receiving the vaccines. Generally, the titers of anti-RBD-IgG (p = 0.004) and NAbs (p = 0.013) in T2DM patients (n = 89) were significantly lower than those in HCs (n = 100). The mean antibody titers of anti-RBD-IgG in T2DM patients and HCs were 0.89 and 1.65 log₂ AU/mL, respectively, and the mean titers of NAbs were 0.24 and 0.61 log₂ug/mL, respectively (Fig. 1a). The positive rates of both antibodies in T2DM patients were also significantly lower than those in HCs (anti-RBD-IgG: p < 0.001; NAbs: p = 0.006) (Additional file 1: Fig. S1a) (additional figures are given in Additional file 1). Well-controlled BMI and FPG levels benefit the prognosis of patients with T2DM. Therefore, we compared the antibody titers in T2DM patients with different BMI and FPG levels. Consistent with our expectation, the titers of anti-RBD-IgG in the BMI < 24 group (n = 45) were significantly higher than those in the BMI ≥ 24 group (n = 44, p = 0.009). A similar trend was observed in the NAbs titers (p = 0.084) (Fig. 1b). Likewise, the positive rates of both antibodies were also higher in T2DM patients with BMI < 24, although the difference was not statistically significant (anti-RBD-IgG: p = 0.139; NAbs: p = 0.24) (Additional file 1: Fig. S1b). Compared with the FPG ≥ 7 group (n = 45), we found that the antibody titers (anti-RBD-IgG: p = 0.242; NAbs: p = 0.214) (Fig. 1c) and positive rates (anti-RBD-IgG: p = 0.072; NAbs: p = 0.303) (Additional file 1: Fig. S1c) were higher in the FPG < 7 group (n = 26), but the differences were not statistically significant.

In addition, we performed a subgroup analysis of HbA1c in the T2DM group, as HbA1c reflects the average plasma glucose levels over the past 2–3 months. We found that the titers of both antibodies (anti-RBD-IgG: p = 0.688; NAbs: p = 0.495) were similar between the HbA1c ≥ 7% group (n = 35) and the HbA1c < 7% group (n = 29) (Fig. 1d). At the same time, there were no differences in seroprevalence between the two groups (anti-RBD-IgG: p = 0.93; NAbs: p = 0.69) (Additional file 1: Fig. S1d).

Furthermore, we analyzed the antibody levels of T2DM patients treated with insulin (among the hypoglycemic drugs, only insulin was used) and those whose regimens did not include insulin (their regimens also did not contain other hypoglycemic drugs). The results showed that both anti-RBD-IgG and NAbs titers in the insulin group (n = 26) were higher than those in the non-insulin group (n = 36), despite the differences were not statistically significant (anti-RBD-IgG: p = 0.392; NAbs: p = 0.602) (Fig. 1e); the trend in positive rates of antibodies was similar to titers (anti-RBD-IgG: p = 0.326; NAbs: p = 0.265) (Additional file 1: Fig. S1e). To investigate whether coexisting with other independent chronic diseases contributed to inferior antibody titers compared to patients with T2DM alone. We assessed antibody titers in patients who were diagnosed with only T2DM (n = 46) and with other independent chronic diseases (n = 43) (including at least one of the following diseases: primary hypertension, coronary heart disease, chronic hepatitis B, chronic obstructive pulmonary disease, and primary chronic kidney disease). We found no differences in antibody titers (anti-RBD-IgG: p = 0.743; NAbs: p = 0.844) or positive rates (anti-RBD-IgG: p = 0.649; NAbs: p = 0.250) between the two groups (Fig. 1f and Additional file 1: Fig. S1f).

In the sex subgroup analysis, no differences were found in antibody titers and positive rates between male (n = 49) and female (n = 40) patients with T2DM. It was worth noting that the antibody titers of males (n = 56) were lower than those of females (n = 44) in HCs (anti-RBD-IgG: p = 0.006; NAbs: p = 0.007) (Additional file 1: Fig. S2a). The trend in positive rates was similar to that in titers (anti-RBD-IgG: p = 0.042; NAbs: p = 0.017) (Additional file 1: Fig. S1g). In terms of age, antibody titers and positive rates were also similar between participants older than 65 years and younger than 65 years, both in T2DM patients (≥ 65y, n = 43; < 65y, n = 46) and in HCs (≥ 65y, n = 32; < 65y, n = 68) (Additional file 1: Figs. S1h and S2b). We then sought to explore whether the administration of different inactivated vaccines would result in different responses, and we found that the titers of both antibodies in HCs vaccinated with CoronaVac were higher than those vaccinated with BBIBP-CorV (anti-RBD-IgG: p = 0.008; NAbs: p = 0.01). In the T2DM group, there were no differences in anti-RBD-IgG antibody titers (p = 0.078) or positive rates (p = 0.059) between patients who received CoronaVac and those who received BBIBP-CorV, but the NAbs titers (p = 0.006) and positive rates (p = 0.006) of the former were higher than those of the latter (Additional file 1: Figs. S1i and S2c).

In summary, the antibody responses of T2DM patients after inactivated COVID-19 vaccines were significantly impaired compared to those of healthy individuals.

Memory B cells produce accelerated and stronger immune responses in secondary immune responses. Therefore, the frequency of RBD-specific MBCs and their subsets was analyzed. We discovered that HCs showed a higher percentage of RBD-specific MBCs than T2DM patients (p = 0.027). In RBD-specific MBCs subsets, the percentage of RBD-specific resting memory B cells (RBD + rMBCs) (p < 0.001) and RBD-specific intermediate memory B cells (RBD + intMBCs) (p < 0.035) was also lower in T2DM patients than in HCs, but the trend of RBD-specific atypical memory B cells (RBD + atyMBCs) was reversed (p < 0.001) (Fig. 2a). Moreover, the percentage of activated memory B cells (RBD + actMBCs) was higher in the T2DM group, but the differences were not statistically significant (p = 0.397). In the T2DM cohort, no statistically significant differences were found in the percentage of RBD-specific MBCs and their subsets, regardless of BMI, FPG, and HbA1c levels (Fig. 2b–d). In terms of medication, treatment with or without insulin appeared to not affect RBD-specific MBCs (Fig. 2e); similar results were revealed between patients diagnosed with only T2DM and those with other independent chronic diseases (Fig. 2f).

Taken together, RBD-specific MBCs responses were attenuated in T2DM patients compared with healthy individuals after inactivated COVID-19 vaccines.

The trend in antibody titers and the frequency of RBD-specific MBCs over time (21–105 days after vaccination) was also analyzed. On the whole, both antibody titers declined as more days passed after receiving the full-course vaccines in all participants (Fig. 3a). Nevertheless, the percentage of RBD-specific MBCs in HCs remained stable, and the percentage of RBD-specific MBCs in T2DM patients showed a slight upward trend (Fig. 3b).

We pooled all T2DM patients to perform multiple linear regression analysis, in which female sex was a positive factor for antibody titers (anti-RBD-IgG: β = 0.608, p = 0.001; NAbs: β = 0.501, p = 0.014), while BMI was a negative factor for antibody titers (anti-RBD-IgG: β =  − 0.528, p = 0.003; NAbs: β =  − 0.475, p = 0.021). In addition, the days after full-course vaccination were inversely correlated with anti-RBD-IgG (β =  − 0.419, p = 0.013). The interaction terms between antibody titers and the variables (age, FPG, HbA1c, insulin use, and T2DM patients combined with other independent chronic diseases, albumin, and creatinine) were not powered to result in statistically significant p values (Tables 3 and 4).