If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Early humoral response to COVID-19 vaccination in patients living with obesity and diabetes in France. The COVPOP OBEDIAB study with results from the ANRS0001S COV-POPART cohort
Support Unit for Clinical Research and Economic Evaluation, Assistance Publique-Hôpitaux de Marseille, 13385 Marseille, FranceAix-Marseille Univ, EA 3279 CEReSS-Health Service Research and Quality of Life Center, Marseille, France
Inserm, F-CRIN, Innovative Clinical Research in Vaccinology Network (I REIVAC), Paris, FranceService des Maladies Infectieuses et Tropicales, CHU de Nîmes, Nîmes, FranceINSERM U1047 – Université de Montpellier, Nîmes, France
Unite des Virus Emergents, Aix-Marseille Université, Institut de Recherche pour le Développement 190, Inserm 1207, Institut Hospitalo-Universitaire Méditerranée Infection, Marseille, France
Assistance Publique-Hôpitaux de Paris, Service de Nutrition, Hôpital Européen Georges Pompidou, Centre Spécialisé Obésité Ile-de-France Sud, 75015 Paris, France
Assistance Publique-Hôpitaux de Paris, Service de Nutrition, Hôpital Européen Georges Pompidou, Centre Spécialisé Obésité Ile-de-France Sud, 75015 Paris, FranceAssistance Publique-Hôpitaux de Paris (AP-HP), Centre d'Investigation Clinique INSERM 1418, Hôpital Européen Georges Pompidou, Paris, France
Support Unit for Clinical Research and Economic Evaluation, Assistance Publique-Hôpitaux de Marseille, 13385 Marseille, FranceAix-Marseille Univ, EA 3279 CEReSS-Health Service Research and Quality of Life Center, Marseille, France
Unite des Virus Emergents, Aix-Marseille Université, Institut de Recherche pour le Développement 190, Inserm 1207, Institut Hospitalo-Universitaire Méditerranée Infection, Marseille, France
Inserm, F-CRIN, Innovative Clinical Research in Vaccinology Network (I REIVAC), Paris, FranceUniversité Paris Cité; Inserm CIC 1417; Assistance Publique Hôpitaux de Paris, Centre d’investigation clinique Cochin Pasteur, Paris, France
Univ Lyon, CarMeN Laboratory, Inserm, Inrae, Université Claude Bernard Lyon-1, Oullins, FranceCentre de Recherche en Nutrition Humaine Rhône-Alpes, Univ Lyon, CarMeN Laboratory, Université Claude Bernard Lyon-1, Hospices Civils de Lyon, Cens, Fcrin/force Network, Pierre-Bénite, France
Patients with diabetes and obesity are populations at high-risk for severe COVID-19 outcomes and have shown blunted immune responses when administered different vaccines. Here we used the ‘ANRS0001S COV-POPART’ French nationwide multicenter prospective cohort to investigate early humoral response to COVID-19 vaccination in the sub-cohort (‘COVPOP OBEDIAB’) of patients with obesity and diabetes.
Methods
Patients with diabetes (n = 390, type 1 or 2) or obesity (n = 357) who had received two vaccine doses and had no history of previous COVID-19 infection and negative anti-nucleocapsid (NCP) antibodies were included and compared against healthy subjects (n = 573). Humoral response was assessed at baseline, at one month post-first dose (M0) and one-month post-second dose (M1), through percentage of responders (positive anti-spike SARS-CoV-2 IgG antibodies (Sabs), geometric means of Sabs; BAU/mL), proportion of individuals with anti-RBD antibodies, and proportion of individuals with anti-SARS-CoV-2-specific neutralizing antibodies (Nabs). Potential clinical and biological factors associated with weak response (defined as Sabs < 264 BAU/mL) and presence of non-reactive anti-RBD antibodies at M1 were evaluated. Univariate and multivariate regressions were performed to estimate crude and adjusted coefficients with 95 % confidence intervals. Poor glycemic control was defined as HbA1c ≥ 7.5 % at inclusion.
Results
Patients with diabetes, particularly type 2 diabetes, and patients with obesity were less likely to have positive Sabs and anti-RBD antibodies after the first and second dose compared to controls (p < 0.001). At M1, we found Sabs seroconversion in 94.1 % of patients with diabetes versus 99.7 % in controls, anti-RBD seroconversion in 93.8 % of patients with diabetes versus 99.1 % in controls, and Nabs seroconversion in 95.7 % of patients with diabetes versus 99.6 % in controls (all p < 0.0001). Sabs and anti-RBD seroconversion at M0 and M1 were also significantly lower in obese patients than controls, at respectively 82.1 % versus 89.9 % (p = 0.001; M0 Sabs), 94.4 % versus 99.7 % (p 0.001; M1 Sabs), 79.0 % vs 86.2 % (p = 0.004 M0 anti-RBD), and 96.99 % vs 99.1 % (p = 0.012 M1 anti-RBD). The factors associated with low vaccine response (BAU < 264/mL) in patients with diabetes were chronic kidney disease (adjusted OR = 6.88 [1.77;26.77], p = 0.005) and poor glycemic control (adjusted OR = 3.92 [1.26;12.14], p = 0.018). In addition, BMI ≥ 40 kg/m2 was found to be associated with a higher vaccine response (adjusted OR = 0.10 [0.01;0.91], p = 0.040) than patients with BMI < 40 kg/m2.
Conclusion
COVID-19 vaccine humoral response was lower in patients with obesity and diabetes one month after second dose compared to controls, especially in diabetic patients with CKD or inadequate glycemic control. These findings point to the need for post-vaccination serological checks in these high-risk populations.
Graphical abstract
Graphical AbstractEarly humoral response to COVID-19 vaccination in patients living with obesity and diabetes in France
]. Accordingly, COVID-19-related hospital and intensive care unit (ICU) admission rates and COVID-19 severity are higher in people with type 2 diabetes (T2D) and people with obesity [
Obesity doubles mortality in patients hospitalized for severe acute respiratory syndrome coronavirus 2 in Paris hospitals, France: a cohort study on 5,795 patients.
]. A systematic review including 400,000 patients revealed that obesity increased risk for COVID-19 infection by 46 %, risk for ICU admission by 74 %, and risk of mortality by 40 % [
]. Furthermore, healthcare cost studies have evidenced that excess weight contributed disproportionally to the costs of COVID-19, with total direct costs of secondary care estimated at €13.9 billon, of which 76 % for overweight or obesity [
]. In addition, previous studies on vaccines against influenza, hepatitis B, and rabies have shown a reduced immune response in individuals with obesity [
]. Following the onset of the COVID-19 pandemic, a number of COVID-19 vaccines were rapidly developed, which were shown to have high efficacy in phase 3 studies. However, in these large-scale studies, people with class III obesity or diabetes were generally underrepresented compared to other populations [
Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK.
]. Since then, an unprecedented mass vaccination campaign has managed to slow the spread and mortality of SARS-CoV-2. As of August 31st 2022, 66.17 % of the entire population of Europe had received a complete initial protocol of COVID-19 vaccine. However, despite the fact that COVID-19 vaccination is the most powerful tool at our disposal to mitigate the severity of COVID-19 disease, multiple unresolved issues remain with regard to the efficacy and immunogenicity of COVID-19 vaccines in populations at high risk of severe COVID-19, such as people with diabetes and severe obesity [
]. In particular, the determinants of antibody responses to vaccine remain poorly understood in these specific populations. To address this issue, we explored the early antibody response (IgG anti-spike (Sabs), anti-receptor binding domain (RBD), and neutralizing antibodies (Nabs)) to primary COVID-19 vaccination in a subpopulation of subjects with obesity but not diabetes or in people with type 1 (T1D) or type 2 (T2D) diabetes, and compared both populations to controls included in the French national multi-center prospective cohort study ANRS0001S COV-POPART.
2. Methods
2.1 Study design
In the COVPOP OBEDIAB study, we investigated the humoral immune response to COVID-19 vaccines at inclusion (−15 days to first dose), M0 (−7 days to second dose) and M1 (one month after the second dose) in two specific subpopulations at risk of severe forms of COVID-19, i.e. subjects with type 1 or type 2 diabetes and subjects with obesity but without diabetes. The OBEDIAB study specifically addressed humoral immune response to primary COVID-19 vaccination (two doses) during the first phase of French vaccination campaign. The HAS (French National Health Authority) recommendations for social distancing and preventive measures were strong and robust and were strictly followed by the general public and especially the most at-risk populations.
These preliminary analyses included patients and controls who had received at least two vaccine doses. Patients were excluded if they had missing serological results at one of the three studied timepoints, i.e. inclusion, M0 and M1, if they had only one first-dose vaccination at the date before inclusion, and if they had positive Sabs, Nabs or anti-RBD at inclusion. Furthermore, participants who tested positive for SARS-CoV-2 anti-nucleocapsid (NCP) antibodies before vaccination or at any follow-up visit were excluded to ensure that asymptomatic prior SARS-Cov2 infections were removed from the analysis.
2.2 Population studied
ANRS0001S COV-POPART is a French national multicenter prospective cohort study designed to evaluate humoral response at 1, 6, 12 and 24 months after the first two doses of COVID-19 vaccines in 11 specific subpopulations, including patients with obesity (BMI ≥ 30 kg/m2) and patients with diabetes mellitus (either type 1 or type 2), compared to healthy individuals (control groups ≥ 18 years old) [
]. The study was conducted in 36 participating centers in France in collaboration with the I-REIVAC network, 10 national learned societies, and 7 patients associations (‘France Rein’, ‘Transhépate’, ‘ARSEP Foundation’, ‘CNAO’, ‘FFD’, ‘EGMOS’, and ‘TRT5 CHV’). Patients were included during the first phase of the French vaccination campaign, which started rollout in France on 27th December 2020.
Written informed consent was obtained from each participant before enrolment, in full compliance with EU GDPR (General Data Protection Regulation) requirements. The study protocol (No. EudraCT/ID-RCB: 2021-A00348-33) was conducted in accordance with the Declaration of Helsinki principles and French law governing research involving human subjects (known as ‘Loi Jardé’). The protocol was approved by the ‘CPP Nord-Ouest IV’ institutional review board (file number: 21.02.12.47147) and the ‘CNIL’ French national data protection authority (authorization number 921111v1).
The criteria for inclusion in the present study were as follows:
–
diabetes, defined by the American Diabetes Association as a history of diabetes, antidiabetic medication or symptoms of diabetes (polyuria, polydipsia, and unexplained weight loss) plus a casual plasma glucose concentration of ≥200 mg/dL (11.1 mmol/L) or fasting plasma glucose ≥ 126 mg/dL (7.0 mmol/L) and/or 2-h post-glucose load ≥ 200 mg/dL (11.1 mmol/L) during an oral glucose tolerance test or HbA1c ≥ 6.5 % (48 mmol/mol). Type 1 diabetes (T1D) was defined as a history of positive IA-2 or GAD or ZnT8 antibodies. Type 2 diabetes (T2D) was defined as absence of criteria for type 1 diabetes.
–
obesity but without diabetes, defined as a body mass index (BMI) ≥ 30 kg/m2 and HbA1c < 6.5 % without glucose-lowering medication.
The main exclusion criteria were pregnancy or breastfeeding, history of known SARS-CoV-2 infection, acute febrile infection within the previous 72 h, symptoms suggestive of COVID-19 or contact with a case within the last 14 days prior to the inclusion visit, history of severe adverse events post-vaccination or severe allergic manifestations, having received another vaccine within 4 weeks prior to the first injection, or scheduled to receive a licensed vaccine in the 4 weeks following inclusion.
Weight, height and waist circumference were measured at each clinical visit. Dyslipidemia was defined as triglycerides ≥ 1.50 g/L and/or LDL-cholesterol ≥ 1.30 g/L and/or HDL-cholesterol ≤ 0.40 g/L and/or pre-existing lipid-lowering treatment. Non-alcoholic fatty liver disease (NAFLD) was defined as a fatty liver index ≥ 60 [
]. Chronic kidney disease (CKD) was defined as an estimated glomerular filtration rate (eGFR; CKD-EPI formula) < 60 mL/min. Microvascular complications were defined as at least one of retinopathy, nephropathy, or neuropathy. Macrovascular complications were defined as at least one of myocardial infarction, stroke, or peripheral arterial disease. Multiple autoimmune diseases were defined as at least one autoimmune disease among vitiligo, Graves, Addison or Hashimoto diseases. Non-diabetes-related comorbidities were defined as at least one of sleep apnea syndrome, hypertension, dyslipidemia, chronic obstructive pulmonary disease, respiratory insufficiency, or heart failure.
2.3 Serological testing
We evaluated the percentage of responders, i.e. positive for anti-spike SARS-CoV-2 IgG antibodies (Sabs), anti-nucleocapsid (NCP), anti-RBD, and the proportion of participants with anti-SARS-CoV-2-specific neutralizing antibodies (Nabs) (in vitro neutralization assay for the original SARS-CoV-2 strain).
The samples analyzed for the purposes of this study were managed and stored by the ANRS Biobank. All analyses were centralized at IHU Méditerranée (Marseille, France). Anti-SARS-CoV-2 IgG antibodies directed against the S1 domain of the virus spike protein were assessed using the QuantiVac ELISA kit from Euroimmun® (Lubeck, Germany). Anti-NCP antibodies were also assessed using the Euroimmun® kit (Lubeck, Germany), which has 94.6 % sensitivity, 99.8 % specificity, and 98.1 % negative predictive value. Neutralizing antibodies against the D614 Wuhan-related [BetaCoV/France/IDF0372/2020] SARS-CoV-2 viral strain were assessed using a microneutralization test, as previously described [
]. This test uses clinical strains of SARS-CoV-2 (100 TCID50/well) and TMPRSS2-expressing VeroE6 cells and is based on the detection of cytopathic effect (CPE) at 5 days post-infection. It is a virus neutralization titer (VNT100) test, i.e.100 % of wells lysed in duplicate format. The test is automated in an NSB3-level laboratory for all dilution and dispensing steps and for CPE reading. The dilutions tested were 20, 40, 80, 160, 320, 640 and 1280. The range was extended if a titer of 1280 was observed in the first instance.
2.4 Outcomes
The main outcome measures were percentage of responders (positive anti-spike SARS-CoV-2 IgG antibodies (by ELISA), ≥1.1 ratio of optic density between sample and calibrator, as defined by EuroImmun serology), geometric mean titers of anti-spike SARS-CoV-2 IgG antibodies expressed in binding antibody units (BAU)/mL, reactive anti-RBD antibodies, and anti-SARS-CoV-2-specific Nabs (in vitro neutralization assay for the original SARS-CoV-2 strain) at 1 month after the second dose of the primary vaccination regimen.
Proportion of participants with Nabs antibodies (a titer ≥20) were described by percentages. Responders were subcategorized as weak responders (anti-spike SARS-CoV-2 IgG level < 264 BAU/mL) [
], moderate responders ([264 and 1350] BAU/mL), or strong responders (>1350 BAU/mL, i.e. the median IgG anti-Spike level of the whole COV-POPART control group (used as reference), as previously described and validated in the initial ANRS0001S COV-POPART study [
One-month humoral response following two or three doses of messenger RNA coronavirus disease 2019 vaccines as primary vaccination in specific populations in France: first results from the Agence Nationale Recherche contre le Sida (ANRS)0001S COV-POPART cohort.
Categorical variables were expressed as numbers and percentages, and continuous variables were expressed as mean ± standard deviation for normally-distributed variables, or as median (1st quartile–3rd quartile) for non-normally-distributed variables. Baseline characteristics were compared between groups using a chi-square test (or Fisher's exact test, as appropriate) for categorical variables and using a Student's t-test/analysis of variance (ANOVA) (or Mann-Whitney/Kruskal-Wallis test as appropriate) for continuous variables.
Presence of anti-spike SARS-CoV-2 IgG antibodies (Sabs), anti-RBD, and anti-SARS-CoV-2-specific neutralizing antibodies (Nabs) responses to COVID-19 vaccination at one month after the first dose and at one month after the second dose were compared between patients with diabetes versus controls and between patients with obesity versus controls using univariate and adjusted logistic regression (for age and sex).
Potential factors associated with weak response (defined as Sabs < 264 BAU/mL) and presence of non-reactive anti-RBD antibodies at M1 in patients with diabetes or obesity were identified using a chi-squared test using a chi-square test (or Fisher's exact test, as appropriate) for categorical variables and using a Student's t-test (or Mann-Whitney test, as appropriate) for continuous variables. Separate multivariate logistic regression models were then used to identify the potential factors associated with weak response and with presence of non-reactive anti-RBD antibodies. Multivariate models included age and gender in addition to variables with p < 0.20 in the univariate analysis. Poor glycemic control was defined as HbA1c ≥ 7.5 % at inclusion. The multivariate regression models were checked for non-multicollinearity by estimating variance inflation factors. Results are reported as adjusted odds ratios (OR) with 95 % confidence intervals (CI).
A two-sided p-value < 0.05 was considered statistically significant. Statistical analyses were performed using IMB PASW SPSS version 20.0 and R version 4.1.3.
3. Results
3.1 General characteristics of the population
Among the 1882 subjects included in the ANRS0001S COV-POPART subcohorts assessed for eligibility, 1879 fulfilled the eligibility criteria and 1320 were enrolled in this analysis as they had no missing serology values at the three timepoints studied (Fig. 1). Before the first vaccine dose, 6 patients (0.3 %) had detectable Sabs, 27 patients (1.4 %) had detectable anti-RBD antibodies, 16 patients (0.8 %) had positive neutralizing antibodies (Nabs), 6 patients (0.3 %) had received their first dose of vaccine before inclusion, and 63 patients (3.3 %) had detectable NCP antibodies and were excluded from the following analysis in order to rule out prior COVID-19 infection or a non-primary vaccination schedule (Fig. 1).
The sample was divided into three groups: controls (n = 573), subjects with obesity but no diabetes (n = 357), and subjects with type 1 or type 2 diabetes (n = 390). Time to assessment of immunological response was a median 30 days [IQR 28–33] after vaccination.
The key characteristics of the studied population are reported in Table 1. Patients with diabetes were older (53.8 ± 14.7 years old) than patients with obesity (47.1 ± 12.8) and controls (47.2 ± 14.7) (p < 0.001). There was a higher proportion of men in both the diabetes group (50.3 %) and the control group (45.6 %) versus patients with obesity (35.6 %) (p < 0.01). Mean BMI was significantly different between controls (22.2 kg/m2 [20.7–23.5]), patients with diabetes (including type 2 (n = 246) and type 1 (n = 144) patients, 28 kg/m2 [24.5–33.3]), and patients with obesity (35.5 kg/m2 [32.5–40.3]; p < 0. 001), and a total of 92 patients (25.8 %) had class III obesity with a BMI ≥ 40 kg/m2 (Fig. 1)). Regarding comorbidities, patients with obesity more often had sleep apnea syndrome and NAFLD, whereas patients with diabetes more often had arterial hypertension, CKD and heart failure (p < 0.001). All patients with obesity or diabetes received an mRNA vaccine (Pfizer or Moderna); 3 % of the healthy controls received an AstraZeneca vaccine (Table 1). As expected, T1D patients were younger and had a lower BMI and fewer comorbidities but higher associated autoimmune diseases compared to T2D patients (Supplementary Table 1).
Table 1Baseline characteristics of the study population.
Characteristics
Controls (n = 573)
Patients with obesity (n = 357)
Patients with diabetes (n = 390)
p‡
Age (years), m ± SD
47.2 ± 14.7a
47.1 ± 12.8
53.8 ± 14.7
<0.001
Gender (M), n (%)
262 (45.8)a
127 (35.6)
196 (50.3)
<0.001
BMI (kg/m2), median (1st Q.–3rd Q.)
22.2 (20.7–23.5)a
35.5 (32.5–40.3)
28.0 (24.5–33.3)
<0.001
BMI ≥ 40 kg/m2, n (%)
0 a
92 (25.8)
38 (9.7)
<0.001
Waist circumference (cm), median (1st Q.–3rd Q.)
–
M 118.0 (110.0–126.2)
M 104.0 (94.0–118.3)
<0.001
F 112.0 (103.0–124.0)
F 98.0 (88.0–113.8)
<0.001
Abdominal adiposity (NCEP-ATP III criteria), n (%)
–
313 (96.9)
226 (66.1)
<0.001
Time since diagnosis (years), m ± SD
–
15.7 ± 11.7
15.9 ± 11.3
0.767
Hypertension, n (%)
31 (5.4) a
78 (21.8)
164 (42.1)
<0.001
Dyslipidemia, n (%)
–
230 (70.6)
257 (66.8)
0.277
HbA1c (≥7.5 %), n (%)
–
0
167 (46.6)
<0.001
History of bariatric surgery, n (%)
–
52 (15.2)
14 (9.3)
0.076
CKD, n (%)
–
15 (4.6)
72 (19.7)
<0.001
COPD, n (%)
5 (0.9)a
7 (3.3)
15 (5.2)
0.312
Sleep apnea syndrome, n (%)
–
54 (54.5)
56 (15.1)
<0.001
Respiratory insufficiency, n (%)
0a
7 (3.3)
10 (3.4)
0.934
Heart failure, n (%)
1 (0.2)a
7 (3.3)
22 (7.5)
0.045
NAFLD, n (%)
–
273 (91.9)
172 (52.4)
<0.001
FIB-4 score
–
0.76 (0.54–1.04)
0.89 (0.61–1.31)
<0.001
History of multiple autoimmune diseases (1/4), n (%)
7 (1.2)a
11 (5.2)
21 (7.2)
0.363
Type of vaccine (1st and 2nd dose), n (%):
Pfizer + Pfizer
488 (85.2)
326 (91.3)
343 (87.9)
NA
Moderna + Moderna
58 (10.1)
26 (7.3)
41 (10.5)
AstraZeneca + AstraZeneca
17 (3.0)
0
0
Pfizer + Moderna
4 (0.7)
4 (1.1)
3 (0.8)
Pfizer + AstraZeneca
0
1 (0.3)
0
Moderna + AstraZeneca
1 (0.2)
0
0
NA (second dose)
5 (0.9)
0
3 (0.8)
Notes: ‡p-value for differences between patients with diabetes or obesity; asignificant difference at p < 0.05 between the three groups. BMI body mass index; CKD chronic kidney disease; COPD chronic obstructive pulmonary disease; NA (second dose): not available for the second dose; NAFLD non-alcoholic fatty liver disease, defined as a fatty liver index ≥ 60. FIB-4 was calculated as a biomarker of hepatic fibrosis. Dyslipidemia was defined as triglycerides ≥ 1.50 g/L and/or LDL-cholesterol ≥ 1.30 g/L and/or HDL-cholesterol ≤ 0.40 g/L and/or pre-existing lipid-lowering treatment. Multiple autoimmune disease was defined as at least one autoimmune disease among vitiligo, Graves, Addison or Hashimoto.
3.2 Comparison of early humoral response to COVID-19 vaccines between patients with diabetes and controls
After the first dose of vaccine, 76.2 % of patients with diabetes developed Sabs antibodies, versus 89.9 % of controls (p < 0.0001), 71.5 % developed anti-RBD antibodies versus 86.2 % of controls (p < 0.0001), and 60.2 % developed positive Nabs versus 68.5 % of controls (p = 0.01) (Fig. 2).
Fig. 2Effect of diabetes mellitus or obesity on humoral response to COVID-19 vaccination in patients without a history of COVID-19 ***p < 0.0001; **p < 0.01; *p < 0.05.
After the second dose of vaccine, Sabs seroconversion was achieved in 94.1 % of patients with diabetes versus 99.7 % in controls (p < 0.0001), anti-RBD seroconversion (reactive IgG anti-RBD) was achieved in 93.8 % of patients with diabetes versus 99.1 % of controls (p < 0.0001), and Nabs seroconversion (positive Nabs antibodies) was achieved in 95.7 % of the patients with diabetes versus 99.6 % in controls (p < 0.0001) (Fig. 2). As shown in Supplementary Table 2, patients with diabetes, and specifically T2D patients, were less likely to have reactive Sabs, anti-RBD and Nabs after the first and second doses of vaccine: OR = 0.36 [0.25;0.51], p < 0.001 for Sabs, OR = 0.40 [0.29;0.56], p < 0.001 for anti-RBD, and OR = 0.69 [0.51;0.94], p = 0.017 for Nabs after the first dose; OR = 0.06 [0.01;0.24], p < 0.001 for Sabs, OR = 0.13 [0.05;0.36], p < 0.001 for anti-RBD, and OR = 0.09 [0.02;0.38], p = 0.001 for Nabs after the second dose (Table 2, Supplementary Table 2 and Fig. 3). Adjustment for age and gender did not significantly change these results at M1 (Table 2 and Supplementary table 2).
Table 2Association between patients with diabetes or obesity (compared to controls) and the presence of anti-spike, anti-RBD, and neutralizing antibody responses to COVID-19 vaccination at one month after the first and one month after the second dose.
Patients with diabetes
Patients with obesity
Non adjusted OR [95%CI]
p
Adjusted OR* [95%CI]
p
Non adjusted OR [95%CI]
p
Adjusted OR* [95%CI]
p
One month after first dose
Sabs (positive)
0.36 [0.25;0.51]
<0.001
0.49 [0.34;0.72]
<0.001
0.52 [0.35;0.76]
0.001
0.45 [0.30;0.67]
<0.001
Anti-RBD IgG (reactive)
0.40 [0.29;0.56]
<0.001
0.54 [0.38;0.76]
0.001
0.60 [0.42;0.85]
0.004
0.52 [0.36;0.76]
0.001
Neutralizing antibodies (positive)
0.69 [0.51;0.94]
0.017
0.80 [0.59;1.10]
0.167
0.83 [0.61;1.12]
0.220
0.82 [0.60;1.12]
0.216
One month after second dose
Sabs (positive)
0.06 [0.01;0.24]
<0.001
0.07 [0.02;0.29]
<0.001
0.12 [0.03;0.56]
0.007
0.12 [0.03;0.56]
0.007
BAU > 1350/mL
0.70 [0.54;0.91]
0.007
0.84 [0.64;1.10]
0.208
1.14 [0.87;1.49]
0.336
1.10 [0.83;1.45]
0.498
Anti-RBD IgG (reactive)
0.13 [0.05;0.36]
<0.001
0.17 [0.06;0.46]
0.001
0.28 [0.10;0.80]
0.018
0.24 [0.08;0.72]
0.010
Neutralizing antibodies (positive)
0.09 [0.02;0.38]
0.001
0.10 [0.02;0.45]
0.003
0.24 [0.05;1.26]
0.092
0.22 [0.04;0.14]
0.072
Notes: *Adjusted for age and gender. Sabs: anti-spike IgG antibodies; BAU: body antibody units.
Fig. 3Effect of the type of diabetes on humoral response to COVID-19 vaccination in patients without a history of COVID-19. T1D: type-1 diabetes. T2D: type-2 diabetes. ***p < 0.0001; **p < 0.01; *p < 0.05.
3.3 Factors associated with vaccine response in patients with diabetes
At one month after the second dose, the percentage of weak responders was significantly higher in patients with diabetes than patients with obesity or controls (12.3 % versus 5.6 % and 3.5 %, respectively, p < 0.0001), and in patients with T2D versus T1D (16.3 % versus 5.6 %, respectively, p < 0.0001) (Fig. 4 and Supplementary Table 2). Factors associated with weak vaccine response (BAU < 264/mL) after the second dose in patients with diabetes were older age, type 2 diabetes, CKD, hypertension, microvascular complications, and NAFLD (Table 3 and Supplementary Table 3). In multivariate analysis, CKD (adjusted OR = 6.88 [1.77;26.77], p = 0.005) and poor glycemic control (i.e. HbA1c ≥7.5 %, adjusted OR = 3.92 [1.26;12.14], p = 0.018) were factors independently associated with a weak vaccine response (BAU < 264/mL) (Table 4). In addition, BMI ≥ 40 kg/m2 was found to be associated with a higher vaccine response (adjusted OR = 0.10 [0.01;0.91], p = 0.040) than patients with BMI < 40 kg/m2.
Fig. 4Percentage of patient responders according to the strength of anti-spike antibodies response at one month after the second dose of anti-SARS CoV-2 vaccine in the group of controls, patients with obesity, diabetes and according to the type of diabetes (type 1 or type 2).
Table 3Determinants of weak responders (BAU < 264/mL) at one month after the second dose.
Patients with diabetes (n = 390)
Patients with obesity (n = 357)
Weak responders (n = 48)
Moderate and strong responders (n = 342)
p
Weak responders (n = 20)
Moderate and strong responders (n = 337)
p
Age (years), m ± SD
60.5 ± 12.0
52.9 ± 14.9
<0.001
47.2 ± 15.2
47.1 ± 12.7
0.986
Gender (M), n (%)
24 (50.0)
172 (50.3)
0.970
9 (45.0)
118 (35.0)
0.365
BMI (kg/m2), median (1st Q.–3rd Q.)
28.6 (24.8–34.1)
28.0 (24.4–33.1)
0.800
32.9 (31.2–37.7)
35.7 (32.5–40.6)
0.025
BMI ≥ 40 kg/m2, n (%)
1 (2.1)
37 (10.8)
0.066
1 (5.0)
91 (27.0)
0.029
Abdominal adiposity, n (%)
23 (76.7)
203 (65.1)
0.200
15 (93.8)
298 (97.1)
0.403
HbA1c (≥7.5 %), n (%)
22 (56.4)
145 (45.5)
0.195
–
–
–
Time since diagnosis (years), median (1st Q.–3rd Q.)
11.0 (4.0–21.0)
15.0 (7.0–21.0)
0.399
8.5 (1.0–23.0)
15.0 (6.0–20.5)
0.094
Type 2 diabetes, n (%)
40 (83.3)
206 (60.2)
0.002
–
–
–
Microvascular complication (1/3), n (%)
19 (45.2)
85 (26.3)
0.011
–
–
–
Macrovascular complication (1/3), n (%)
9 (21.4)
46 (14.2)
0.217
0
11 (3.4)
1
Sleep apnea syndrome, n (%)
5 (11.9)
51 (15.5)
0.545
0
54 (55.7)
0.204
History of bariatric surgery, n (%)
1 (4.2)
13 (10.2)
0.699
2 (11.1)
50 (15.4)
1
CKD, n (%)
22 (48.9)
50 (15.6)
<0.001
3 (17.6)
12 (3.9)
0.037
Hypertension, n (%)
31 (75.6)
133 (52.8)
0.006
6 (54.5)
72 (36.0)
0.336
Dyslipidemia, n (%)
34 (73.9)
223 (658)
0.272
13 (86.7)
217 (69.8)
0.246
COPD, n (%)
4 (9.8)
11 (4.4)
0.242
0
7 (3.5)
1
Respiratory insufficiency, n (%)
1 (2.4)
9 (3.6)
1
1 (9.1)
6 (3.0)
0.315
Heart failure, n (%)
4 (9.8)
18 (7.2)
0.527
0
7 (3.5)
1
History of cancer, n (%)
3 (7.7)
11 (4.5)
0.417
1 (9.1)
5 (2.5)
0.281
Multiple autoimmune diseases (1/4), n (%)
1 (2.4)
20 (8.0)
0.329
0
11 (5.5)
1
Non-diabetes-related comorbidities (1/6), n (%)
42 (87.5)
254 (74.3)
0.045
–
–
–
NAFLD, n (%)
22 (78.6)
150 (50.0)
0.004
11 (91.7)
262 (91.9)
1
Insulin treatment, n (%)
21 (55.3)
219 (70.2)
0.061
–
–
–
mRNA vaccine (1st dose), n (%)
48 (100.0)
342 (100.0)
NA
20 (100.0)
336 (99.7)
1
Notes: BMI body mass index; CKD chronic kidney disease; COPD chronic obstructive pulmonary disease; NAFLD non-alcoholic-fatty liver disease. Microvascular complications were defined as at least one of retinopathy, nephropathy or neuropathy. Macrovascular complications were defined as at least one of myocardial infarction, stroke or peripheral arterial disease for patients with diabetes or as at least one of myocardial infarction, stroke or peripheral arterial disease for patients with obesity. Multiple autoimmune diseases were defined as at least one history of autoimmune disease among vitiligo, Graves, Addison or Hashimoto disease. Non-diabetes-related comorbidities were defined as at least one of sleep apnea syndrome, hypertension, dyslipidemia, chronic obstructive pulmonary disease, respiratory insufficiency, or heart failure.
Table 4Independent factors associated with weak response (BAU < 264 U/mL) at one month after the second dose. Multivariate analysis.
Patients with diabetes
Patients with obesity
OR CI 95 %
p
OR CI 95 %
p
Age (years)
0.99 [0.94;1.04]
0.707
0.99 [0.95;1.03]
0.555
Gender (M)
0.64 [0.24;1.72]
0.377
0.91 [0.28;2.92]
0.872
BMI (≥40 kg/m2)
0.10 [0.01;0.90]
0.040
–
–
BMI (kg/m2)
–
–
0.89 [0.77;1.04]
0.136
Diabetes (type 2)
2.59 [0.46;14.68]
0.281
–
–
Microvascular complications (at least 1/3)
0.77 [0.24;2.50]
0.660
–
–
CKD
6.88 [1.77;26.77]
0.005
5.66 [0.86;37.36]
0.072
Time since diagnosis (years)
–
–
0.97 [0.91;1.03]
0.354
NAFLD
2.65 [0.79;8.91]
0.115
Insulin treatment
0.67 [0.22;2.07]
0.489
HbA1c (≥7.5 %), n (%)
3.92 [1.26;12.14]
0.018
Non-diabetes-related comorbidities (at least 1/6, n (%))
2.14 [0.23–19.95]
0.506
Notes: BAU body antibody units; BMI body mass index; CKD chronic kidney disease; NAFLD non-alcoholic-fatty liver disease. Non-diabetes-related comorbidities were defined as at least one of sleep apnea syndrome, hypertension, dyslipidemia, chronic obstructive pulmonary disease, respiratory insufficiency, or heart failure.
Factors associated with presence of non-reactive anti-RBD antibodies after the second dose in patients with diabetes were older age, CKD, and type 2 diabetes (Supplementary Tables 4 and 5, Fig. 3). In multivariate analysis, CKD (adjusted OR = 22.38 [5.23;95.69], p < 0.001) was the only factor associated with non-reactive anti-RBD antibodies after the second vaccine dose (Supplementary Table 6).
3.4 Comparison of early humoral response to COVID-19 vaccines in patients with obesity
After the first dose of vaccine, 82.1 % of patients with obesity developed Sabs antibodies versus 89.9 % of controls (p = 0.001), 79 % developed anti-RBD antibodies versus 86.2 % of controls (p = 0.004), and 64.3 % developed positive Nabs versus 68.5 % of controls (p = 0.220) (Fig. 2).
After the second dose of vaccine, Sabs seroconversion was achieved in 94.4 % of patients with obesity versus 99.7 % in controls (p = 0.001), RBD seroconversion (reactive IgG anti-RBD) was achieved in 96.9 % of patients with obesity versus 99.1 % in controls (p = 0.012),and Nabs seroconversion (positive Nabs antibodies) was achieved in 98.4 % of patients with obesity versus 99.6 % in controls (p = 0.112) (Fig. 2).
Patients with obesity were less likely to have reactive Sabs and anti-RBD after the first and second doses of vaccine: OR = 0.52 [0.35;0.76], p = 0.001 for Sabs, OR = 0.60 [0.42;0.85], p = 0.004 for reactive anti-RBD after the first dose; OR = 0.12 [0.03;0.56], p = 0.007 for Sabs, OR = 0.28 [0.10;0.80], p = 0.018 for reactive anti-RBD after the second dose (Table 2). Adjustment for age and gender did not change these results (Table 2).
3.5 Factors associated with vaccine response in patients with obesity
Factors associated with weak vaccine response (BAU < 264/mL) after the second dose of vaccine in patients with obesity were lower BMI and CKD (Table 3).
In multivariate analysis, these factors did not remain independently associated to weak vaccine response (Table 4). Factors associated with non-reactive anti-RBD antibodies after the second dose in obese patients were lower BMI and a shorter duration of obesity based on recorded weight history (Supplementary Table 5), but these factors did not remain associated after multivariate analysis (Supplementary Table 6).
4. Discussion
This large, French nationwide population-based multicentric cohort analysis provides unique evidence that the humoral response to primary COVID-19 vaccination based on anti-spike and anti-RBD antibodies was lower in the population with diabetes or obesity than in control subjects, one month after the first and second doses of vaccine during the first phase of rollout of the French vaccination campaign. Anti-SARS-CoV-2-specific neutralizing antibodies were also found in a lower percentage of patients with diabetes than controls after the complete initial ‘double-jab’ protocol without prior SARS-CoV-2 infection. A remarkable yet counterintuitive finding is the fact that diabetic patients with class III obesity were protected against weak vaccine response (BAU < 264/mL), and that no obese patients with a BMI ≥ 40 kg/m2 had non-reactive anti-RBD antibodies, whatever the group (diabetic or obese). Furthermore, in patients with diabetes, poor glycemic control at inclusion was independently associated with weak vaccine response.
Our findings are consistent with an Israeli study on >342 healthcare workers with comorbidities given the BNT162b2 vaccine, which reported that diabetes was significantly associated with lower antibody concentrations (anti-RBD and Nabs) [
BNT162b2 COVID-19 vaccine and correlates of humoral immune responses and dynamics: a prospective, single-centre, longitudinal cohort study in health-care workers.
]. In contrast, Lee et al. reported that T2D patients showed a reduction in CD4+ T-helper-1 (Th1) differentiation following their first dose of BNT162b2 vaccine [
]. However, this initial defect was rectified by the second dose, resulting in levels of CD4+ memory T-cells and CD8+ T cells, anti-RBD IgG, and Nabs that were comparable to healthy individuals at 3–6 months after vaccination [
]. In one recent Kuwaiti study, both T2D (n = 81) and non-diabetic individuals (n = 181) elicited strong immune responses to SARS-CoV-2 BNT162b2 mRNA vaccine; nonetheless, lower levels were seen in people with diabetes, suggesting that continuous monitoring of the antibody levels might be a good indicator to guide personalized needs for further booster shots to maintain adaptive immunity [
]. However, we found no significant association between abdominal adiposity (as assessed by waist circumference) at the time of vaccination and immune responses. This is in contrast with one recent study showing that patients with abdominal obesity and predominant visceral adipose tissue accumulation achieved lower neutralizing IgG antibodies against the Trimeric-complex peak at 1 month compared to individuals without abdominal adiposity [
]. Interestingly, we found that in patients with diabetes, poor glycemic control (i.e. HbA1c ≥ 7.5 %) was independently associated with weak vaccine response (BAU < 264/mL). This is consistent with the Italian study by Marfella et al. (n = 494 T2D) that showed that good glycemic control (HbA1c < 7 %) was associated with a higher virus-neutralizing antibody capacity and a better CD4+ T-cell response compared to poor glycemic control [
]. In this study, HbA1c was linearly associated with the incidence of breakthrough infections. Other studies investigating short-term effects of COVID-19 vaccination on the time spent in different glycemic ranges assessed by continuous glucose monitoring (CGM) in people with type 1 and type 2 diabetes have found that COVID-19 vaccination per se did not impact glycemic control in people with diabetes [
], but that T2D patients with initially poor glycemic control showed improved immune responses when they achieved good glycemic control during the vaccination period [
]. Furthermore, Montefusco et al. nicely evidenced the direct tropism of SARS-CoV-2 on human pancreatic islets and the presence of new-onset hyperglycemia, insulin resistance and long-term beta-cell hyperstimulation after COVID-19 recovery [
]. They further demonstrated that an exacerbated pancreatic inflammatory milieu initiated by a cytokine storm and increased serum secretome (IL-1ß, IL-2, IL-4, IL-7, IL-8, IL-10, IL13, IL-17, G-CSF and IFN-γ) could persist chronically and lead to beta-cell exhaustion and dysfunction [
]. Note that the same team reported an impaired cellular SARS-CoV2-specific cytotoxic immune response and no increase in the release of cytotoxic factors (granzyme A and perforin) or T-cell-related cytokines Il-2 and TNF-α after SARS-CoV2 mRNA vaccines in patients with type 1 diabetes compared to subjects without diabetes, without any significant disruption in CGM parameters, which may indicate a defective ability to inactivate the virus and a less immunogenic vaccination in T1D patients [
Rapid weight loss, central obesity improvement and blood glucose reduction are associated with a stronger adaptive immune response following COVID-19 mRNA vaccine.
]. A recent article from the UK including 9,171,524 participants highlighted that people with underweight and people with obesity remained at greater risk of hospitalization or death even after a second dose of the vaccine, but that people with overweight or obesity who had been vaccinated has a comparable level of protection against severe COVID-19 to people of healthy weight, suggesting that physicians should better target their efforts towards increasing vaccine uptake in people with low BMI (18.5 kg/m2) [
Associations of BMI with COVID-19 vaccine uptake, vaccine effectiveness, and risk of severe COVID-19 outcomes after vaccination in England: a population-based cohort study.
]. Bates et al. reported that BMI had no significant impact on anti-RBD titers and surrogate neutralizing titers 50 days following immunization with the BNT162b2 vaccine [
]. Watanabe et al. interestingly evaluated the adaptive humoral (Sabs) and cell-mediated responses (IFN secretion in response to stimulation with two different SARS CoV-2 peptide mixes, IFN-1 and IFN-2) and found that higher baseline BMI was associated with reduced humoral and cell-mediated responses to BNT162b2 vaccine, while weight loss and normalization of blood glucose induced by a 15-day hypocaloric very-low-carbohydrate diet reversed this negative effect [
Rapid weight loss, central obesity improvement and blood glucose reduction are associated with a stronger adaptive immune response following COVID-19 mRNA vaccine.
]. Finally, Kara et al. recently observed that patients with grade III obesity (BMI ≥ 40 kg/m2,n = 124) generated significantly lower Sabs after CoronaVac and BNT162b2 vaccines compared to healthy individuals [
]. In our study population, among the 130 patients with a BMI ≥ 40 kg/m2, only two were found to be weak responders (BAU < 264/mL). This could possibly be due to the fact that the phenotype of obesity (with ectopic fat deposition, NAFLD, or metabolic complications) may influence the antibody response rather than the simple grade of obesity. For the patients with diabetes, we found that the T2D subgroup had a higher BMI and a higher percentage of patients with grade III obesity than the T1D subgroup, along with a higher percentage of weak responders. However, T2D was not found to be independently associated with a weak vaccine response, unlike the BMI, thus suggesting that the type of diabetes did not affect the association, although this needs to be confirmed in dedicated large-sample studies.
Here we found that CKD was independently associated with lower vaccine response in the subpopulation of patients with diabetes. This finding warrants caution, as part of diabetes complications are known to increase the risk for lower vaccine response. Several studies have reported that patients with CKD have low seroconversion rates in response to mRNA vaccines and fail to develop sustained humoral response, especially those with kidney transplants [
], suggesting a need for ongoing isolation measures and supplementary booster doses in these populations. In a recent systematic review including 16 studies, Boroumand et al. extended our findings that patients with diabetes have lower antibody response after the second dose of vaccine compared to individuals without diabetes, irrespective of vaccine type. Several studies included in this meta-analysis were consistent with our results and reported that optimum glycemic control and higher glomerular filtration rate were associated with higher antibody response among patients with diabetes mellitus [
This study has several limitations. We only studied the early humoral response to COVID-19 vaccines in patients with obesity and diabetes, and not cellular response. Furthermore, we did not assess neutralizing activity against all viral variants of concern. We cannot completely rule out that we included participants with asymptomatic SARS-CoV-2 infection because anti-NCP tests have not a 100 % sensitivity and anti-NCP antibodies wane over time. Thus, humoral responses presented in this study may not reflect vaccine responses only as prior infection could have impacted them [
]. Nonetheless, we do think that the probability is low given the inclusion period for the study and the non-pharmaceutical interventions in place at that time. Most participants in this study were vaccinated with an mRNA vaccine, which limits the generalizability of our findings to other vaccine types, but reflects the use of the mRNA vaccine in France. Future analyses are also warranted to check whether this lower vaccine response in patients with type 2 diabetes persists long-term. Finally, potential confounders such as immunomodulators and associated treatments such as steroids were not considered in this analysis and could have impacted our results.
This study also has several strengths. It enrolled a large-sample nationwide observational multicentric study including >1300 individuals, with a unique representativeness of patients living with class-III obesity (n = 130) and a control group of healthy individuals included during the same period of evaluation. This study included 36 participating centers representing >200 clinical sites in France across France in collaboration with the renowned I-REIVAC network, as well as 10 national learned societies and 7 patient associations. Patients were mainly included during the first phase of the French vaccination campaign rollout, so the findings reflect homogeneous immune response to primary vaccination. Lastly, a major strength pf this study is that all samples were centralized in a unique laboratory, which makes the biological analysis extremely robust and minimizes any variability in the findings.
5. Conclusion
Patients with diabetes or obesity are at higher risk of weak vaccine response at one month after a second COVID-19 vaccine dose. Further characterization of this vulnerable population may help to develop guidance on further protective measures, either by continued social distancing or by additional active or passive vaccinations. In light of these results, we would advise patients with type 2 diabetes and/or obesity to benefit from additional booster doses. Patients with additional comorbidities associated with a weak response, such as CKD or poor glycemic control, should be tested for a post-vaccination serological check.
CRediT authorship contribution statement
Conception and design: BG, LN, XDL, AD, BC, MC, CK, SC, CC, ML. Statistical analysis and/or interpretation of data: BG, SF, NR, ML, OL. Technical analysis of samples: LN, XDL. Drafting of the manuscript or revising it critically for important intellectual content: PL, LW, OL, LE, AB, JL, BG, CS, BC, KC, ML, NR, SF. Final approval of the manuscript submitted: all the authors.
Declaration of competing interest
None.
Acknowledgments
The authors would like to thank all the participants, the dedicated staff at the local centers and their respective biological resource centers, for making this research possible through their continued and committed investment in the cohort. Without their ongoing participation, a cohort of this magnitude would simply not be possible.
We thank the CTU, UMS 54 MART (Methods and Applied Research for Trials) Univ. Bordeaux, INSERM for the methodology, data management and coordination of the ANRS0001S COV-POPART cohort.
We also thank the ANRS SC10 US109 INSERM Biobank for its huge ongoing contribution to this project, the Steering Committee, the Scientific Steering Committee, the ANRS0001S COV-POPART Study Group and the F-CRIN I-Reivac Network team for their support.
Finally, we also thank the ANRS MIE and its COVID-19 scientific committee and CAPNET, the national steering committee for clinical trials and other research on COVID-19 for their support on this study, and the French ministries (Ministère des Solidarités et de la Santé and Ministère de l'Enseignement Supérieur, de la Recherche et de l'Innovation) for their generous ongoing financial support.
We thank the FORCE Network for its help in coordinating the French centers specialized in obesity and diabetes care, and Josep Verdecho Mendez and Chloe Robert for their technical support.
We also thank all the coordinators at the Clinical Investigation Centers, and in particular Bertrand Dussol for his invaluable help.
Obesity doubles mortality in patients hospitalized for severe acute respiratory syndrome coronavirus 2 in Paris hospitals, France: a cohort study on 5,795 patients.
Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK.
One-month humoral response following two or three doses of messenger RNA coronavirus disease 2019 vaccines as primary vaccination in specific populations in France: first results from the Agence Nationale Recherche contre le Sida (ANRS)0001S COV-POPART cohort.
BNT162b2 COVID-19 vaccine and correlates of humoral immune responses and dynamics: a prospective, single-centre, longitudinal cohort study in health-care workers.
Rapid weight loss, central obesity improvement and blood glucose reduction are associated with a stronger adaptive immune response following COVID-19 mRNA vaccine.
Associations of BMI with COVID-19 vaccine uptake, vaccine effectiveness, and risk of severe COVID-19 outcomes after vaccination in England: a population-based cohort study.
00015-X&id=ga1.jpg){kind=link}
00015-X&id=gr1.jpg){kind=link}
00015-X&id=gr2.jpg){kind=link}
00015-X&id=gr3.jpg){kind=link}
00015-X&id=gr4.jpg){kind=link}