Original Research Article Maternal Hemoglobin Depletion in a Settled Northern Kenyan Pastoral Population ELIZABETH M. MILLER* Department of Anthropology, University of Michigan, Ann Arbor, Michigan 48109 Objectives: This study examines maternal hemoglobin depletion in a cross-sectional sample of Ariaal women living in northern Kenya. Maternal hemoglobin depletion occurs when women do not have enough dietary iron to replace the high levels of iron allocated to the fetus during pregnancy. Methods: To study this phenomenon, reproductive histories, socioeconomic status, anthropometry, and hemoglobin levels were collected from a cross-section of 200 lactating Ariaal women in northern Kenya. Results: Ariaal women show increasing levels of hemoglobin with increasing time since birth and lower hemoglobin levels with increasing parity, indicating an incomplete repletion of dietary iron over women’s reproductive lifetime. Women who lived in a more livestock-dependent village had higher hemoglobin levels and lower prevalence of clinical anemia than women who lived in villages more dependent on agriculture, indicating that differences in diet may alleviate the effects of iron depletion. Conclusions: These data demonstrate that Ariaal women are iron depleted due to pregnancy, incompletely replete hemoglobin during the course of lactation, and show depletion of hemoglobin with increasing parity. Women in this community may be able to improve their iron status through a greater reliance on food sources rich in dietary iron. Am. J. Hum. Biol. 22:768–774, 2010. ' 2010 Wiley-Liss, Inc. Evolutionary theory predicts that energy spent on reproduction takes away from an organism’s total energy, reducing the amount of energy available for somatic main- tenance. This forms the basis for life history studies in both humans and other animals (Stearns, 1992). As part of our life history, humans have high energetic depletion during reproduction. Human pregnancy costs up to 300– 500 kcal/day (Butte et al., 2004) and lactation costs up to 500 kcal/day (Goldberg et al., 1991) on top of basal meta- bolic expenditures. In addition, pregnancy and lactation requires a high level of micronutrient investment to support the needs of the growing fetus and infant. In populations with insufficient macro- and micronutrients to replace the resources lost in reproduction, women may experience temporary or permanent depletion. These effects may compound with increasing numbers of off- spring, a phenomenon known as maternal depletion (Jelliffe and Maddocks, 1964). Evidence for maternal depletion in human populations is mixed. Among women in low resource environments, some studies confirm the existence of maternal fat deple- tion with increasing parity (e.g. Adair and Popkin, 1992; Little et al., 1992; Tracer, 1991) while others, particularly women in well-nourished populations, see the opposite pattern of fat deposition with increasing pregnancies; women gain fat stores with each successive pregnancy (e.g. Brown et al., 1992; Harris et al., 1997; Prentice et al., 1981; Smith et al., 1994). Because maternal depletion is not consistent among women, researchers have indicated that a single model of maternal depletion is not appropri- ate. Rather, there are four patterns of reproduction- associated weight change (Shell-Duncan and Yung, 2004; Winkvist et al., 1992). 1. Nondepleted women: food intake is equal to or higher than activity levels and reproductive costs; women can maintain or improve nutritional stores throughout the reproductive cycle and show no parity-related deficien- cies. Women who maintain or gain weight with each pregnancy could be considered nondepleted (e.g. Harris et al., 1997). 2. Repleted women: experience decrease in nutritional stores due to costs of reproduction but are able to completely recover after each reproductive cycle; they do not show long-term parity-related depletion. Women in Lesotho, for example, experience decline in fat stores during lactation, which they then recover before their next pregnancy. They have no parity-related decline in adiposity (Miller and Huss-Ashmore, 1989). 3. Incompletely repleted women: demonstrate decrease in nutritional status and only partially replete between reproductive cycles; these women show patterns of both short-term and long-term depletion. One example would be Au women in Papua New Guinea, who experience declines in nutritional status that become more pronounced with greater parity (Tracer, 1991). 4. Nonrepleted women: have chronic depletion of nutri- tional stores independent of reproductive costs; costs of reproduction are masked by high levels of malnutri- tion. For example, a study of mobile and small-town Rendille women of northern Kenya showed no change in adiposity during lactation, indicating that energy Contract grant sponsor: Leakey Foundation and National Science Foun- dation Doctoral Dissertation Improvement Grant; Contract grant number: BCS-0750779. *Correspondence to: Elizabeth Miller, Department of Anthropology, 101 West Hall, 1085 S. University Ave., Ann Arbor, MI 48109-1107, USA. E-mail: emmill@umich.edu Received 20 February 2010; Revision received 1 June 2010; Accepted 2 June 2010 DOI 10.1002/ajhb.21078 Published online 18 August 2010 in Wiley Online Library (wileyonlinelibrary. com). AMERICAN JOURNAL OF HUMAN BIOLOGY 22:768–774 (2010) V V C 2010 Wiley-Liss, Inc.