ABO and Rhesus Blood Groups Have No Significant Impact on COVID-19 Severity or Mortality in Senegal: Evidence from 295 Patients

1. Introduction

COVID-19, caused by SARS-CoV-2, was declared a global pandemic on March 11, 2020, leading to considerable morbidity and mortality [1,2]. While the majority of patients present with mild forms, a significant proportion develops severe and critical forms associated with high mortality, particularly in elderly patients and those with comorbidities [3].

Since the early weeks of the pandemic, several research groups hypothesized a role of ABO blood groups in COVID-19 susceptibility and severity. A preliminary Chinese study suggested an overrepresentation of blood group A among infected patients and a potentially protective effect of blood group O [4]. These data prompted numerous investigations in various international cohorts, yielding discordant results [5,6,7]. Proposed mechanisms include ABO-SARS-CoV-2 antigenic interactions, differences in von Willebrand factor levels, and variations in infection susceptibility [8].

In Senegal, the Epidemiological Treatment Center (CTE) of Aristide Le Dantec Hospital in Dakar served as one of the main sites for COVID-19 patient management during the first pandemic waves. African data on the role of blood groups in COVID-19 outcomes remain scarce, and to our knowledge no Senegalese study has simultaneously evaluated the association between ABO/Rh groups and COVID-19 mortality.

The objective of this study is to analyze the association between ABO and Rhesus blood groups, clinical severity, and mortality from COVID-19 in the combined cohort of the first pandemic waves at the CTE of Aristide Le Dantec Hospital, Dakar.

2. Materials and Methods

Retrospective, descriptive, and analytical study conducted at the CTE of Aristide Le Dantec Hospital, Dakar. It covers patients hospitalized during the 1st wave (2020) and 2nd wave (2021) of COVID-19.

All patients with RT-PCR-confirmed COVID-19, hospitalized at the CTE, with ABO/Rh blood group documented in their medical records were included. Patients without available blood group data (n=37) were excluded, resulting in an analyzed sample of 295 patients.

Primary severity variable: severe form (Severe and Critical), defined according to WHO criteria. Primary mortality variable: in-hospital death. Independent variable: ABO blood group (O, A, B, AB) and Rhesus status (Rh+, Rh−).

Associations between blood group and severity or mortality were assessed using Fisher’s exact test (comparison of each group vs. all others combined) and global Chi-square (4 ABO groups × 2 outcomes). Bonferroni correction (×4 for the 4 ABO comparisons) was applied. Odds Ratios (OR) with 95% confidence intervals (Woolf method) were calculated. The significance threshold was p < 0.05. All analyses were performed using Python 3 (scipy, pandas).

All data were anonymized prior to analysis, in accordance with good clinical practices and applicable regulations in Senegal. This study was conducted in compliance with the principles of the Declaration of Helsinki.

3. Results

3.1. Population Characteristics

A total of 295 patients with documented blood group were included (out of 332 patients hospitalized at the CTE during the first pandemic waves). Mean age was 57.5 ± 18.1 years. Male predominance was marked (65.1% men). Severe forms accounted for 80 patients (27.1%) and 30 deaths (10.2%) were recorded.

3.2. ABO and Rhesus Blood Group Distribution

ABO distribution: 141 patients of group O (47.8%), 85 of group A (28.8%), 50 of group B (16.9%), and 19 of group AB (6.4%). Rh-positive status accounted for 94.2% (n=278) and Rh-negative for 5.8% (n=17). This distribution is consistent with available data on the general Senegalese population.

Table 1. Distribution of ABO/Rh blood groups, clinical severity and mortality.

Group	n (%)	Mild (n)	Moderate (n)	Severe (n)	Critical (n)	Severe forms (%)	Deaths (%)	p Severe	p Deaths
ABO GROUPS
O	141 (47.8%)	74	26	21	20	29.1%	9.2%	0.513	0.701
A	85 (28.8%)	40	20	13	12	29.4%	9.4%	0.567	1.000
B	50 (16.9%)	32	7	7	4	22.0%	14.0%	0.485	0.312
AB	19 (6.4%)	11	5	2	1	15.8%	10.5%	0.299	1.000
RHESUS GROUPS
Rh+	278 (94.2%)	—	—	—	—	27.7%	10.8%	0.414	0.234
Rh−	17 (5.8%)	—	—	—	—	17.6%	0.0%	—	0.234
TOTAL
Total	295 (100%)	163	58	43	31	27.1%	10.2%	Chi²=2.40	Chi²=1.00

Table 1. Distribution of ABO/Rh blood groups, clinical severity and mortality.

Group	n (%)	Mild (n)	Moderate (n)	Severe (n)	Critical (n)	Severe forms (%)	Deaths (%)	p Severe	p Deaths
ABO GROUPS
O	141 (47.8%)	74	26	21	20	29.1%	9.2%	0.513	0.701
A	85 (28.8%)	40	20	13	12	29.4%	9.4%	0.567	1.000
B	50 (16.9%)	32	7	7	4	22.0%	14.0%	0.485	0.312
AB	19 (6.4%)	11	5	2	1	15.8%	10.5%	0.299	1.000
RHESUS GROUPS
Rh+	278 (94.2%)	—	—	—	—	27.7%	10.8%	0.414	0.234
Rh−	17 (5.8%)	—	—	—	—	17.6%	0.0%	—	0.234
TOTAL
Total	295 (100%)	163	58	43	31	27.1%	10.2%	Chi²=2.40	Chi²=1.00

Figure 1. Distribution of ABO blood groups (n=295).

Figure 1. Distribution of ABO blood groups (n=295).

Figure 2. Proportions of severe forms (A) and deaths (B) by ABO blood group.

Figure 2. Proportions of severe forms (A) and deaths (B) by ABO blood group.

Figure 3. Heatmap of COVID-19 severity distribution by ABO group.

Figure 3. Heatmap of COVID-19 severity distribution by ABO group.

3.3. Association Between ABO Group and Severe Forms

No statistically significant association between ABO blood group and the occurrence of severe forms was observed (Chi² = 2.40; p = 0.494). Proportions of severe forms by ABO group were: 29.1% (group O), 29.4% (group A), 22.0% (group B), and 15.8% (group AB). ORs calculated for each group versus all others combined were all non-significant, both in crude analysis and after Bonferroni correction (Table II, Figure 4).

3.4. Association Between ABO Group and Mortality

Overall mortality was 10.2% (30/295), ranging from 9.2% in group O, 9.4% in group A, 14.0% in group B, to 10.5% in group AB. No difference reached statistical significance (global Chi² = 1.00; p = 0.802).

A non-significant trend toward higher mortality in group B is worth noting (OR = 1.57; 95% CI [0.63–3.89]; p = 0.312; Bonferroni p = 1.000). The limited statistical power due to the small number of deaths (n=30) precludes any definitive conclusion on this point, which will require confirmation in larger cohorts.

3.5. Association Between Rhesus Status and Severity/Mortality

Rh-negative status (n=17) was not associated with severe forms (17.6% vs 27.7% in Rh+; p = 0.414). No deaths were recorded among the 17 Rh− patients (0.0% vs 10.8% in Rh+), with a p-value of 0.234 by Fisher’s exact test. This result, though suggestive, must be interpreted with extreme caution given the very small Rh− subgroup (n=17), which is insufficient for any definitive conclusion.

Table 2. Odds Ratios for severe forms and deaths by ABO/Rh group (n=295).

Comparison	OR	95% CI	p crude	p Bonf.	OR	95% CI	p crude	p Bonf.
	SEVERE FORMS				DEATHS
Group O vs non-O	1.21	[0.72–2.02]	0.513	1.000	0.82	[0.38–1.75]	0.701	1.000
Group A vs non-A	1.17	[0.67–2.05]	0.567	1.000	0.89	[0.38–2.08]	1.000	1.000
Group B vs non-B	0.72	[0.35–1.49]	0.485	1.000	1.57	[0.63–3.89]	0.312	1.000
Group AB vs non-AB	0.48	[0.14–1.71]	0.299	1.000	1.04	[0.23–4.75]	1.000	1.000
Rh+ vs Rh−	1.79	[0.51–6.33]	0.414	—	—	0 deaths/Rh−	0.234	—

OR = Odds Ratio; 95% CI = 95% Confidence Interval (Woolf method); p Bonf. = p after Bonferroni correction (×4 for ABO). Two-sided Fisher’s exact test. — = not calculable (0 events).

Table 2. Odds Ratios for severe forms and deaths by ABO/Rh group (n=295).

Comparison	OR	95% CI	p crude	p Bonf.	OR	95% CI	p crude	p Bonf.
	SEVERE FORMS				DEATHS
Group O vs non-O	1.21	[0.72–2.02]	0.513	1.000	0.82	[0.38–1.75]	0.701	1.000
Group A vs non-A	1.17	[0.67–2.05]	0.567	1.000	0.89	[0.38–2.08]	1.000	1.000
Group B vs non-B	0.72	[0.35–1.49]	0.485	1.000	1.57	[0.63–3.89]	0.312	1.000
Group AB vs non-AB	0.48	[0.14–1.71]	0.299	1.000	1.04	[0.23–4.75]	1.000	1.000
Rh+ vs Rh−	1.79	[0.51–6.33]	0.414	—	—	0 deaths/Rh−	0.234	—

OR = Odds Ratio; 95% CI = 95% Confidence Interval (Woolf method); p Bonf. = p after Bonferroni correction (×4 for ABO). Two-sided Fisher’s exact test. — = not calculable (0 events).

Figure 4. Comparative Forest plot of Odds Ratios for severe forms (●) and deaths (◆) for each ABO/Rh group versus all other groups combined. Logarithmic scale. CTE Aristide Le Dantec Hospital, Dakar (n=295).

Figure 4. Comparative Forest plot of Odds Ratios for severe forms (●) and deaths (◆) for each ABO/Rh group versus all other groups combined. Logarithmic scale. CTE Aristide Le Dantec Hospital, Dakar (n=295).

Figure 5. Individual comparison of each ABO group versus all other groups combined: proportions of severe forms and deaths with Odds Ratios and p-values (crude and Bonferroni). CTE Aristide Le Dantec Hospital, Dakar (n=295).

Figure 5. Individual comparison of each ABO group versus all other groups combined: proportions of severe forms and deaths with Odds Ratios and p-values (crude and Bonferroni). CTE Aristide Le Dantec Hospital, Dakar (n=295).

4. Discussion

The ABO blood group distribution observed in this cohort, with a predominance of group O (47.8%), followed by group A (28.8%), B (16.9%), and AB (6.4%), is consistent with available data on the Senegalese and West African population [4,5]. This distribution differs from those of Caucasian or Asian populations where group A is often more prevalent, which may partly explain discordances between studies.

The present study found no significant association between ABO blood group and COVID-19 severity (p = 0.494). This result is consistent with several recent meta-analyses which, after adjustment for confounding factors (age, sex, comorbidities), do not find an independent effect of blood groups on clinical severity [6,7]. The landmark study by Zhao et al. [4] had suggested a protective role for group O, but this initial work had important methodological limitations (absence of adjustment, unrepresentative sample sizes).

No significant association between ABO blood group and mortality was identified in this cohort (Chi² = 1.00; p = 0.802). However, a trend toward higher mortality in group B (14.0% vs 9.4% for other groups; OR = 1.57 [0.63–3.89]) is worth noting. Although non-significant, this trend is biologically plausible, as studies have suggested a role of the B antigen in hypercoagulability and inflammatory response [8]. The limited statistical power due to the number of deaths (n=30) and the relatively modest cohort size preclude any definitive conclusion.

The absence of deaths among the 17 Rh− patients (0.0% vs 10.8% in Rh+) is a striking finding, but must be interpreted with the greatest caution. The probability that this difference is due to chance remains high (p = 0.234), and the Rh− subgroup (n=17) is clearly insufficient for any conclusion. Further studies with larger Rh− populations will be needed to explore this finding.

Our results suggest that ABO/Rh blood group does not constitute, in clinical practice, a reliable predictive marker of COVID-19 severity or mortality in this Senegalese cohort. In resource-limited settings, other biological parameters such as NLR, neutrophils, or lymphocytes have demonstrated better predictive value and appear more relevant for risk stratification at admission.

Main limitations include: the retrospective study design; the exclusion of 37 patients without available blood group data (possible selection bias); the absence of multivariate adjustment for comorbidities and age; the small Rh− subgroup (n=17) limiting statistical power for this subgroup; and the absence of external validation.

5. Conclusions

This study of 295 COVID-19 patients at the CTE of Aristide Le Dantec Hospital in Dakar, during the first pandemic waves, found no statistically significant association between ABO or Rhesus blood groups and the occurrence of severe forms or deaths from COVID-19.

A non-significant trend toward higher mortality in group B (OR = 1.57; p = 0.312) and zero mortality among Rh− patients deserve exploration in larger cohorts. In clinical practice, blood group cannot replace validated biological markers such as NLR or neutrophil count for COVID-19 risk stratification at admission in the Senegalese context.