The relationship between hemoglobin and (V)over dotO₂max: A systematic review and meta-analysis

Objective There is widespread agreement about the key role of hemoglobin for oxygen transport. Both observational and interventional studies have examined the relationship between hemoglobin levels and maximal oxygen uptake ((V)over dotO(2)max) in humans. However, there exists considerable variability in the scientific literature regarding the potential relationship between hemoglobin and (V)over dotO(2)max. Thus, we aimed to provide a comprehensive analysis of the diverse literature and examine the relationship between hemoglobin levels (hemoglobin concentration and mass) and (V)over dotO(2)max (absolute and relative (V)over dotO(2)max) among both observational and interventional studies. Methods A systematic search was performed on December 6(th), 2021. The study procedures and reporting of findings followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Article selection and data abstraction were performed in duplicate by two independent reviewers. Primary outcomes were hemoglobin levels and (V)over dotO(2)max values (absolute and relative). For observational studies, meta-regression models were performed to examine the relationship between hemoglobin levels and (V)over dotO(2)max values. For interventional studies, meta-analysis models were performed to determine the change in (V)over dotO(2)max values (standard paired difference) associated with interventions designed to modify hemoglobin levels or (V)over dotO(2)max. Meta-regression models were then performed to determine the relationship between a change in hemoglobin levels and the change in (V)over dotO(2)max values. Results Data from 384 studies (226 observational studies and 158 interventional studies) were examined. For observational data, there was a positive association between absolute (V)over dotO(2)max and hemoglobin levels (hemoglobin concentration, hemoglobin mass, and hematocrit (P<0.001 for all)). Prespecified subgroup analyses demonstrated no apparent sex-related differences among these relationships. For interventional data, there was a positive association between the change of absolute (V)over dotO(2)max (standard paired difference) and the change in hemoglobin levels (hemoglobin concentration (P<0.0001) and hemoglobin mass (P = 0.006)). Conclusion These findings suggest that (V)over dotO(2)max values are closely associated with hemoglobin levels among both observational and interventional studies. Although our findings suggest a lack of sex differences in these relationships, there were limited studies incorporating females or stratifying results by biological sex.