Send to

Choose Destination
Physiol Rep. 2019 Jul;7(12):e14153. doi: 10.14814/phy2.14153.

Population data provide evidence against the presence of a set point for hemoglobin levels or tissue oxygen delivery.

Author information

Departments of General Medicine and Endocrinology, Royal Adelaide Hospital, Adelaide, South Australia, Australia.
Department of Cell and Molecular Biology, Karolinska Institutet, Stockholm, Sweden.
Department of Medicine-Cardiology, Karolinska Institutet, Stockholm, Sweden.
Department of Cardiology, University Medical Center Groningen and Rijksuniversiteit Groningen, Groningen, The Netherlands.
Department of Haematology, CHU de Liège and University of Liège, Liège, Belgium.
Department of Internal Medicine, Division of Endocrinology, Diabetology, Vascular Disease, Nephrology and Clinical Chemistry, University Hospital Tübingen, Tübingen, Germany.
Department of Endocrinology, Metabolism and Diabetes, Karolinska University Hospital, Stockholm, Sweden.
Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden.
Menzies School of Health Research and Royal Darwin Hospital, Tiwi, Northern Territory, Australia.
School of Mathematical Sciences and ARC Centre of Excellence for Mathematical and Statistical Frontiers, University of Adelaide, Adelaide, South Australia, Australia.


Hemoglobin levels are believed to be regulated as per a set point model of regulation. This model of regulation, by which specific levels of a parameter are targeted and defended by physiological systems, implies a particular population correlation between the parameter and its controlling hormone. Empirical population correlations of other parameters and their controlling hormones, have denied the presence of such set point-based regulation. To assess if hemoglobin is regulated according to a set point model we performed a systematic search of PubMed/MEDLINE and Web of Science identifying relevant reports published up to November 2018. Population hemoglobin/erythropoietin level correlations were retrieved, and these empirically derived correlations were compared with the positive correlation implied by a set point model of regulation. Authors of papers containing potentially suitable data were contacted with requests for further analyses, and a meta-analysis was performed. Twelve correlations between hemoglobin and erythropoietin levels from eleven papers were analyzed. None of these correlations were significantly positive, three, restricted to the normal range of hemoglobin, were significantly negative. All but one of the other correlations showed a negative trend. New analyses of previously published data sets resulted in similar findings. In particular a new analysis of large data sets of males (n = 2417) and females (n = 2592) with normal range hemoglobin levels, revealed significantly negative correlations. A meta-analysis of our results indicated that the data overall are not consistent with a positive relationship between hemoglobin and erythropoietin (P < 0.0001). Population data indicate that individuals do not have set point levels of hemoglobin.


Erythropoietin; haemoglobin; population correlations; set point

Supplemental Content

Full text links

Icon for Wiley Icon for PubMed Central
Loading ...
Support Center