Relationship Between Levels of Leptin and Hemoglobin in Japanese Men

To the Editor:

Leptin, the ob gene product secreted by adipocyte, decreases food intake while it increases energy expenditure and functions as an important signal for the regulation of body weight.1-3 The leptin receptor is an isoform of the B219 gene product, a member of the hematopoietin receptor family, which is expressed in very primitive hematopoietic cells.4 Recent studies showed that leptin plus erythropoietin acted synergistically to increase erythroid development in vitro.5,6 These findings led us to examine the relationship between the serum levels of leptin and hemoglobin.

We surveyed 708 male workers who were not taking any medication. Information regarding smoking habits, alcohol consumption, and physical activity was obtained by questionnaire and/or from medical records. As for the question regarding physical activity, subjects were asked to choose one from the following four answers: no exercise at all, once or twice per month, once or twice per week, and three times or more per week. Blood was drawn in the morning after a 12-hour or longer fast. Serum leptin and insulin were determined by radio immunoassay (Linco Research Inc, St Charles, MO) and enzyme immunoassay (Dinabot, Tokyo, Japan), respectively.

When the subjects were divided into three groups according to their hemoglobin level (<14.5 g/dL [the lowest quintile], 14.5 to 15.8 g/dL, and ≥15.8 g/dL [the highest quintile]), a negative correlation was observed between the levels of leptin and those of hemoglobin, after being adjusted for age, body-mass index, and physical activity (Table 1). The negative correlation became more apparent after further adjustment for the insulin level. In contrast, when similar analysis was performed to examine the relationship between the levels of leptin and white blood cell counts, no correlation was observed between these two variables after being adjusted for related variables.

This is the first epidemiologic study showing an association between the levels of leptin and those of hemoglobin. Wilson et al7failed to show such a correlation between leptin and red blood cell count without adjusting for leptin-related variables. The gender-dependent difference in serum leptin may be due to the difference of hemoglobin levels, adding to fat mass and sex hormones.8 

Erythropoiesis is thought to be regulated by erythropoietin, which, in adults, is produced mainly in kidneys, in response to hypoxia.9 Leptin production occurs mainly in adipocytes, but there has been no report showing that adipocytes have a sensor for hypoxia. It is interesting that bone marrow contains many adipocytes, the role of which is not clear.

Although the effect of leptin on hematopoiesis may be modest, the results of our epidemiologic study, together with those of previous studies performed in vitro,5,6 suggest that leptin may play some role in hematopoiesis in humans. Further cross-sectional and prospective studies are needed to elucidate the relationship between leptin and hematopoiesis.