Detection of breastfeeding and weaning in modern human infants with carbon and nitrogen stable isotope ratios.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 129:279–293 (2006) Detection of Breastfeeding and Weaning in Modern Human Infants With Carbon and Nitrogen Stable Isotope Ratios B.T. Fuller,1,2* J.L. Fuller,3 D.A. Harris,2 and R.E.M. Hedges1 1 Research Laboratory for Archaeology and the History of Art, University of Oxford, Oxford OX1 3QJ, United Kingdom 2 Department of Biochemistry, University of Oxford, Oxford OX1 3QU, United Kingdom 3 Sutter Auburn Faith Hospital, Auburn, California 95602 KEY WORDS stable isotopes; paleodiet, ﬁngernails; d13C; d15N ABSTRACT Carbon (13C/12C) and nitrogen (15N/14N) stable isotope ratios were longitudinally measured in ﬁngernail and hair samples from mother-infant pairs where infants were exclusively breastfed (n ¼ 5), breast- and formula-fed (n ¼ 2), or exclusively formulafed (n ¼ 1) from birth. All exclusively breastfed infants had a dual enrichment in carbon (1%) and nitrogen (2–3%) when compared to maternal values. In contrast, breast- and formula-fed subjects had reduced enrichments compared to exclusively breastfed subjects, and the exclusively formula-fed infant showed no increase in d13C or d15N values. This ﬁnding of a carbon trophic level effect in breastfeeding infants suggests that 13C-enrichments of approximately 1% in archaeological populations are not necessarily the result of the consumption of C4-based weaning foods such as maize or millet. During the weaning process, the d13C results for breastfed infants declined to maternal levels more rapidly than the d15N results. This suggests that d13C values have the potential to track the introduction of solid foods into the diet, whereas d15N values monitor the length of time of breast milk consumption. These ﬁndings can be used to reﬁne the isotopic analysis of breastfeeding and weaning patterns in past and modern populations. Am J Phys Anthropol 129:279–293, 2006. Since the pioneering discovery that the nitrogen stable isotope ratio (15N/14N) in body tissues can be used to detect the age of weaning (Fogel et al., 1989), there have been numerous applications of this technique to reconstruct breastfeeding and weaning patterns in archaeological populations (Katzenberg et al., 1993; Katzenberg and Pfeiffer, 1995; Schurr, 1997, 1998; Herring et al., 1998; Wright and Schwarcz, 1998, 1999; Dupras et al., 2001; Mays et al., 2002; Richards et al., 2002; Schurr and Powell, 2005). These studies of bone chemistry revealed extensive information about the duration of breastfeeding in different cultures and time periods, and give anthropologists a glimpse of societal views and health patterns that are largely invisible in the archaeological record (Katzenberg et al., 1996). However, our understanding and interpretation of d15N values still rely on data from only two controlled studies of three modern infants, even though the use of nitrogen isotopic ratios in ancient infant nutrition has expanded and become routine (Fogel et al., 1989; Millard, 2000). Thus, precise knowledge concerning the magnitude of the isotopic fractionation and turnover time of isotopic signatures during breastfeeding and weaning in infants is limited. In addition, there has been only one published study of how modern human infant d13C values respond to the consumption of breast milk and weaning foods, with this research reporting no change in d13C results between a single breastfeeding mother-infant pair (Fogel et al., 1989). However, subsequent studies of d13C values from breastfed infants in archaeological populations suggested that individuals display a 1% increase in d13C either as a result of breastfeeding or through the consumption of 13C-enriched C4 (the Hatch-Slack photosynthesis pathway)-based weaning foods such as maize and millet (Katzenberg et al., 1993; Katzenberg and Pfeiffer, 1995; Wright and Schwarcz, 1999; Dupras et al., 2001; Richards et al., 2002; Fuller et al., 2003). If d13C and d15N signatures are to be effectively used to reconstruct breastfeeding and weaning patterns in archaeological and forensic material, then detailed longitudinal research on modern infants is necessary for comparison. This study investigates the d13C and d15N signatures in eight mother-infant pairs (Subjects A–H, mothers; Subjects a–h, infants), where infants were exclusively breastfed (n ¼ 5), breast- and formula-fed (n ¼ 2), or exclusively formula-fed (n ¼ 1) from birth. In order to conduct a longitudinal study of the duration of breastfeeding and the weaning period, we analyzed ﬁngernail and hair samples taken at birth to a minimum age of 1 year from mother-infant pairs residing in the town of Auburn, California. The proteins in sequential sections of ﬁngernails and hair provided a record of isotope ratios existing when that material was synthesized (Fogel et al., 1989; O’Connell and Hedges, 1999). In addition, bovine milk-based (n ¼ 5) and soy-based (n ¼ 2) formulas were also analyzed for d13C and d15N values to determine if unique isotopic signatures exist that could identify modern infants being fed C 2005 V WILEY-LISS, INC. C V 2005 Wiley-Liss, Inc. *Correspondence to: Benjamin Fuller, P.O. Box 707, Auburn, CA 95604. E-mail: email@example.com Received 4 May 2004; accepted 29 October 2004. DOI 10.1002/ajpa.20249 Published online 31 October 2005 in Wiley InterScience (www.interscience.wiley.com). 280 B.T. FULLER ET AL. these different supplements. The ﬁndings of this research have important implications regarding the isotopic interpretation of breastfeeding and weaning patterns in modern and past populations. ISOTOPIC RECONSTRUCTION OF BREASTFEEDING AND WEANING PATTERNS The use of ratios from light stable isotopes (C, N, O, S, and H) for both modern and ancient dietary reconstruction is predicated on the principle that ‘‘you are what you eat’’ plus a small positive fractionation factor known as a trophic level effect (DeNiro and Epstein, 1978; Kohn, 1999). Thus, the isotopic composition of the body will reﬂect the isotopic signatures of foods and liquids consumed during life. In this case, the carbon and nitrogen isotopic results of ﬁngernails and hair will primarily reﬂect the protein portion of the diet. For detailed reviews, see Schwarcz and Schoeninger (1991), Schoeninger (1995), and Kelly (2000). Stable isotope results are measured as the ratio of the heavier isotope to the lighter isotope (13C/12C or 15N/14N) and reported as d values in parts per 1,000 or ‘‘per mil’’ (%) relative to internationally deﬁned standards for carbon (Vienna Pee Dee Belemnite, vPDB) and nitrogen (ambient inhalable reservoir, AIR) (Schwarcz and Schoeninger, 1991). Since the groundbreaking experiments of Fogel et al. (1989), d15N measurements have emerged as one of the most successful techniques to elucidate breastfeeding and weaning patterns in archaeological populations. Building on previous research where each step up the food chain was shown to increase d15N values by approximately 2– 4% (DeNiro and Epstein, 1981; Minagawa and Wada, 1984; Schoeninger and DeNiro, 1984), Fogel et al. (1989) discovered that breastfeeding infants are also one trophic level above their mothers in the food chain, since they are essentially consuming their mother’s tissue through the ingestion of breast milk. Thus, the d15N values of nursing infants are elevated by 2–3% compared to the mother. During the weaning process, the consumption of supplementary foods causes a decline in infant d15N values. When a child is fully weaned (cessation of breastfeeding), its protein d15N values are nearly identical to those of its mother, assuming the mother-child pair has similar diets. MATERIALS AND METHODS Subjects Eight mother-infant pairs (Subjects A–H, mothers; Subjects a–h, infants) participated in this study. All infants were singleton births and were healthy during the period of study. Written informed consent was obtained from the mothers after a full explanation of the research. The protocol was approved by the Sutter Health Central Area Institutional Review Committee (Sacramento, CA). The mother of each subject recorded infant dietary habits, including the timing of the introduction of solid foods (duration and types of foods consumed) and the cessation of breastfeeding. Fingernail and hair analysis Fingernails are an ideal medium to study dietary patterns in modern infants since they are metabolically inert, resistant to degradation, isotopically representative of the body protein pool, and have a fast synthesis rate. Fingernails were collected from subjects at various intervals dur- ing the period of study. When sample size permitted, only nail clippings from the three middle ﬁngers were combined for analysis to reduce errors due to variations in individual nail growth rates. Since ﬁngernails grow out from the cuticle but are collected at the ﬁngertip, there is a lag between sampling time and time of formation. For infants, this lag is estimated to be 2–3 months, such that the ﬁrst sections analyzed were formed in utero (Fogel et al., 1989), while the lag for adult ﬁngernails is estimated to be slightly longer at 4–6 months (Runne and Orfanos, 1981). For cleaning, ﬁngernails were shotblasted to remove surface debris (dirt). Samples (1–1.5 mg) were then placed in tin capsules for isotopic analysis, and combusted to CO2 and N2 in an automated carbon and nitrogen analyzer (Carlo Erba, Milan, Italy) coupled to a continuous-ﬂow isotope-ratio-monitoring mass spectrometer (cf-irm-ms) (PDZ Europa Geo 20/20, Cheshire, UK). When sample size permitted, all specimens were run in triplicate, and analytical errors were typically less than 60.2% for d13C and d15N results. For Subjects H and h, ﬁngernail samples could not be obtained, and hair was analyzed instead. The isotopic values recorded in hair are similar to those of ﬁngernails (O’Connell et al., 2001). Maternal hair samples (30–50 strands) were obtained at 8 weeks postpartum and were measured in triplicate. Infant hair samples (20–40 strands) were obtained at three intervals: birth, age 10 weeks, and age 52 weeks. The hair from Subjects H and h was then cut into 1-cm sections, corresponding to approximately 4 weeks of growth (Saitoh et al., 1969), and prepared for isotopic analysis following the protocol of O’Connell and Hedges (1999). Due to limited sample size, isotopic measurements were only performed once for Subject h. Replicate measurement errors on known standards within the run were less than 0.2% for d13C and 0.2% for d15N. Formula analysis Infant formulas were collected from supermarkets in Auburn, California and isotopically analyzed. The seven different formulas were derived from either cow’s milk (n ¼ 5) or soybeans (n ¼ 2), with various supplements added such as iron. As the ratio of carbon to nitrogen was high, samples had to be analyzed separately for d13C (2–3 mg) and d15N (10–12 mg) values. All samples were run in triplicate, and analytical errors were typically less than 60.2% for d13C and d15N results. RESULTS Exclusively breastfed infants: Subjects a–e The d13C and d15N results from mother-infant pairs (Subjects A–E, mothers; Subjects a–e, infants) where the infant was exclusively breastfed from birth are illustrated in Figures 1–5 and listed in Table 1. Speciﬁc information on the duration of breastfeeding and the timing and types of solid foods consumed for these subjects is listed in Table 2. During breastfeeding, Subjects a–e show a 13Cenrichment of approximately 1% above the mother’s ﬁngernails. The infant d15N values are somewhat more variable, with approximate 15N-enrichments between 1.7– 2.8% compared to the mother’s ﬁngernails. With the addition of solid foods to the diet, the infant d13C values quickly decrease to levels below the maternal d13C values (Figs. 1A, 2A, 3A, 4A, 5A). The infant d15N values also decrease during the weaning process, but in comparison ISOTOPIC SIGNATURES OF BREASTFEEDING AND WEANING 281 Figs. 1A,B. Fingernail d13C (a) and d15N (b) results from Subject A (mother) and Subject a (infant) where Subject a was exclusively breastfed between 12–30 weeks. Speciﬁc dietary information for Subject a is listed in Table 2. to the d13C results, the decline to maternal levels is generally slower and more gradual, with the exception of Subject e (Table 3). Subject e had the most abrupt weaning period (14 weeks) and displayed a simultaneous drop in carbon and nitrogen isotope ratios (Fig. 5A,B). At the cessation of breastfeeding, d13C values tend to increase slightly and then display no general pattern. The d15N results stabilize and change little after the cessation of breastfeeding. Breast- and formula-fed infants: Subjects f–g The results from the two infants who were breast- and formula-fed from birth are presented in Figures 6 and 7 282 B.T. FULLER ET AL. Figs. 2A,B. Fingernail d13C (a) and d15N (b) results from Subject B (mother) and Subject b (infant) where Subject b was exclusively breastfed between 12–32 weeks. Speciﬁc dietary information for Subject b is listed in Table 2. and listed in Table 1. Subject f was breast- and formulafed with Similac1 with iron (180 ml/day) from birth, and displays a small increase in d13C (0.5%) and d15N (1.0%) values during the period of breast- and bottlefeeding (Fig. 6). The start of solid food consumption (rice cereal, chicken, lamb, potatoes, pineapple, and spaghetti) at 30 weeks causes d13C and d15N results to decrease to maternal values. Between 35–45 weeks, Subject f began eating all of the adult table foods in addition to breastand formula-feeding. Subject f was still breastfeeding (mainly for comfort) at time of last sampling at 51 weeks, and the d15N values remained elevated at this point, while the d13C results were below those of Subject F (Fig. 6). Subject g was breast- and formula-fed with Isomil1 soy (30–60 ml/day) from birth and displayed an increase in d13C (0.5%) and d15N (1.4%) during this period (Fig. ISOTOPIC SIGNATURES OF BREASTFEEDING AND WEANING 283 Figs. 3A,B. Fingernail d13C (a) and d15N (b) results from Subject C (mother) and Subject c (infant) where Subject c was exclusively breastfed between 14–47 weeks. Speciﬁc dietary information for Subject c is listed in Table 2. 7). At 25 weeks, Subject g was changed to exclusive nursing for a period of 9 weeks. During this time of exclusive breastfeeding, 13C-enrichment (1%) and 15N-enrichment (3%) were observed. Solid foods (oatmeal, green beans, apples, and pears) were introduced to the diet at week 34, and this resulted in a decrease in d13C and d15N values. Cessation of breastfeeding occurred during week 58, and the subject was eating all of the adult table foods by week 60. Exclusively formula-fed infant: Subject h The results from the only exclusively formula-fed infant are illustrated in Figure 8 and listed in Table 1. Since ﬁngernails were not obtained from Subject h, hair was analyzed. During the bottle-feeding period (350 ml/day of Carnation Good Start1), there was no increase in d13C or d15N results, but rather a decrease of 0.5% for carbon 284 B.T. FULLER ET AL. Figs. 4A,B. Fingernail d13C (a) and d15N (b) results from Subject D (mother) and Subject d (infant) where Subject d was exclusively breastfed between 12–32 weeks. Speciﬁc dietary information for Subject d is listed in Table 2. and 2.0% for nitrogen. Solid foods were ﬁrst introduced to the diet at week 22 (rice cereal, bananas, peaches, apricots, and plums), and at week 26, vegetables (sweet potatoes, potatoes, green beans, corn, and squash) were introduced. The addition of fruits and vegetables to the diet caused d13C and d15N values to recover to those seen at birth. Subject h began consuming meat (chicken and turkey) and dairy products (milk) at week 35, and by week 44 all adult table foods were part of the diet. At time of last sampling, d13C and d15N values were relatively stable at 17.9% and 8.9%, respectively, and Subject h was still consuming formula in addition to table foods. Formula The results from the seven formulas analyzed for d13C and d15N values are presented in Table 4. The formulas were synthesized from either cow’s milk (n ¼ 5) or soybeans (n ¼ 2), and represent four different commercial brands found in Northern California. The majority of milk-based supplements had d13C values between 21.9% and 23.2%, and d15N values between 4–5%. However, the Enfamil1 milk-based formula had signiﬁcantly enriched d13C (19.9%) and d15N (7.5%) results. Both of the soy-based formulas had d13C values that were similar to those derived from cow’s milk, but the d15N values were ISOTOPIC SIGNATURES OF BREASTFEEDING AND WEANING 285 Figs. 5A,B. Fingernail d13C (a) and d15N (b) results from Subject E (mother) and Subject e (infant) where Subject e was exclusively breastfed between 12–19 weeks. In contrast to Subjects a–d, Subject e displays a simultaneous decrease in d13C and d15N values during the weaning period. Speciﬁc dietary information for Subject e is listed in Table 2. signiﬁcantly lower at 0.5%. Since d15N values increase in tandem with each step up the food chain, this difference between soy- and cow milk-based formulas can be attributed to this food chain phenomenon. DISCUSSION Pregnancy and possible in utero effects in mother-infant pairs The mothers’ ﬁngernail and hair d13C results (Subjects A–H) do not show a systematic variation during preg- nancy and lactation. In contrast, during the ﬁrst 12–14 weeks (ﬁngernails formed in utero), 7 of 8 infants (except Subject f) have fetal d13C values that decrease toward time of birth (Figs. 1A, 2A, 3A, 4A, 5A, 6A, 7A, 8A). We speculate that this decline in d13C could be the result of a small change (0.5%) in carbon metabolism in either the mother or fetus, or a variation in the placental uptake and/or transport of carbon. Further research is necessary to fully understand this observation. All mothers (except Subject B) have d15N results that decrease by approximately 0.5–1% between 10 to 20 weeks of infant age (Figs. 1B, 2B, 3B, 4B, 5B, 6B, 7B, 8B). 286 B.T. FULLER ET AL. TABLE 1. C and N results for mother-infant pairs1 13 Subject A a B b C 15 TABLE 1. (Continued) Sampling time (weeks) d13C (%) 6SD d15N (%) 6SD Subject 25 12 6 3 7 12 15 27 36 43 54 0 6 12 14 22 27 29 33 37 45 59 63 67 0 4 8 14 21 28 33 40 44 47 52 57 66 72 77 80 1 8 9 13 16 19 22 25 30 34 39 43 48 52 57 61 66 70 74 79 10 2 5 14 23 34 54 81 18.1 17.9 17.8 18.0 17.8 18.0 18.2 18.3 17.7 17.8 18.2 17.4 17.5 17.6 17.1 16.5 16.7 17.1 18.0 18.9 19.0 18.5 17.7 18.1 18.1 18.0 18.1 18.1 17.7 17.6 17.7 17.7 17.7 17.6 17.8 17.6 18.0 17.7 17.7 17.8 17.5 17.7 17.8 17.2 17.1 17.1 17.2 17.2 17.4 17.4 17.9 18.3 18.8 18.7 18.1 18.1 18.6 18.6 18.5 19.0 18.2 18.4 18.6 18.2 18.4 18.6 18.2 18.4 0.1 0.2 0.2 0.1 0.2 0.1 0.1 0.2 0.1 0.2 0.1 nd nd 0.1 0.1 0.1 0.2 nd nd 0.1 0.2 0.4 0.2 0.2 0.2 0.1 0.1 0.1 <0.05 0.2 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 nd 0.1 0.1 0.1 0.1 0.1 0.3 0.2 0.2 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.2 0.1 0.1 0.1 0.1 nd 0.1 9.3 9.2 9.4 9.4 9.0 8.7 8.8 9.1 9.3 8.9 8.7 8.5 8.6 8.7 9.7 11.4 11.8 11.9 11.3 10.6 10.0 10.3 9.6 9.6 9.2 9.2 9.2 9.1 9.4 9.0 9.0 8.9 9.0 8.9 8.8 9.2 9.1 9.2 9.3 9.3 9.4 9.3 9.6 10.2 10.4 10.6 10.6 11.2 11.2 11.3 11.3 10.6 9.1 8.7 8.5 8.4 9.1 8.7 8.8 8.7 9.2 9.0 8.8 9.0 9.2 8.9 8.9 9.0 0.1 <0.05 0.1 <0.05 0.2 0.1 0.1 0.2 0.2 <0.05 0.3 nd nd 0.2 0.2 0.1 0.1 nd nd 0.3 0.2 0.1 0.2 0.3 <0.05 0.1 <0.05 <0.05 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.2 <0.05 0.1 0.1 0.1 0.2 0.2 0.1 0.4 0.3 nd 0.2 0.2 0.1 0.2 0.1 0.3 0.1 0.1 0.3 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.1 nd 0.1 c D d E e Sampling time (weeks) d13C (%) 6SD d15N (%) 6SD 2 7 9 11 14 18 22 25 31 34 40 50 55 81 4 0 7 17 27 37 45 54 61 76 80 0 8 17 28 38 41 44 48 52 55 61 66 78 85 14 7 3 1 10 14 20 24 31 33 61 65 71 1 8 10 13 15 20 23 25 29 32 37 41 47 51 62 66 18.1 18.6 18.2 18.5 18.5 17.6 17.8 17.9 17.8 17.8 17.7 17.5 18.0 18.2 18.8 19.0 19.2 18.4 18.4 18.4 18.4 18.5 18.1 18.1 18.0 18.5 19.0 17.6 17.9 18.2 18.9 18.9 19.5 19.5 19.4 19.1 19.0 19.1 18.5 18.5 18.0 17.8 18.0 18.1 17.7 17.7 17.8 18.0 18.2 17.7 18.0 17.6 17.5 17.6 17.5 17.0 16.6 16.7 17.1 17.3 18.4 18.5 18.2 18.0 17.8 18.0 18.8 19.1 0.3 0.1 0.2 0.1 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.3 0.1 0.3 0.1 nd nd nd 0.2 0.2 0.3 0.3 0.1 0.1 0.3 0.1 0.1 0.1 0.1 0.2 nd 0.1 0.1 0.2 0.1 0.2 0.3 0.2 0.2 <0.05 0.1 <0.05 nd 0.1 0.2 0.2 0.2 0.1 <0.05 0.2 0.4 0.1 0.1 0.1 0.1 0.1 0.3 0.1 9.8 9.8 10.0 9.8 9.4 11.0 10.9 10.6 11.0 11.1 11.5 11.7 11.4 10.4 9.0 8.4 8.1 8.2 8.0 7.6 7.6 8.0 8.1 8.4 8.6 8.4 8.7 10.3 10.9 10.9 10.4 10.1 9.8 9.7 9.5 9.1 9.0 9.3 9.3 9.6 9.2 9.2 9.0 8.7 8.8 9.1 9.3 9.3 9.3 9.1 9.3 9.3 9.5 9.4 9.5 9.8 10.9 10.8 10.4 9.9 8.1 8.1 9.0 9.0 9.1 9.1 9.4 9.6 0.1 0.2 0.3 0.3 0.1 0.2 0.1 0.2 0.2 0.1 0.2 0.4 0.1 0.2 0.3 0.1 0.1 0.1 0.1 0.1 0.2 0.4 0.2 0.2 0.1 nd nd nd 0.3 0.2 0.2 0.1 <0.05 <0.05 0.3 0.3 0.2 0.2 0.1 0.3 nd 0.3 0.2 0.2 0.1 0.1 0.2 0.1 <0.05 0.2 0.2 <0.05 nd 0.2 0.1 0.3 0.1 0.1 0.1 0.3 0.3 0.2 0.3 <0.05 0.3 0.4 0.1 0.2 ISOTOPIC SIGNATURES OF BREASTFEEDING AND WEANING TABLE 1. (Continued) Subject F f G g H h Sampling time (weeks) d13C (%) 6SD d15N (%) 6SD 18 14 4 0 3 10 15 18 26 34 48 3 10 14 15 19 24 27 32 34 48 51 22 8 3 3 17 22 44 52 2 7 16 22 26 29 32 38 48 77 20 16 12 8 4 0 4 8 4 0 6 10 20 24 28 32 36 40 44 48 52 19.2 19.0 18.3 18.7 18.5 18.8 18.9 18.5 18.9 18.7 18.7 19.0 18.9 18.2 18.2 18.5 18.4 18.5 18.8 18.9 19.4 19.1 18.2 17.7 17.8 17.6 18.3 18.0 17.2 17.3 17.8 18.1 17.1 17.4 17.2 16.9 17.1 17.6 18.4 18.3 17.1 17.0 16.8 16.8 17.1 16.8 17.3 16.9 16.4 17.1 16.5 16.7 17.3 16.9 16.5 16.5 17.1 17.7 18.0 17.9 17.9 0.1 0.3 0.1 0.2 <0.05 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.1 <0.05 0.1 0.2 0.1 0.3 0.1 0.2 0.1 0.1 0.1 0.2 0.1 0.3 nd 0.1 nd 0.1 0.1 0.1 0.1 0.2 0.1 0.4 <0.05 0.1 <0.05 <0.05 <0.05 <0.05 0.2 0.3 0.3 <0.05 nd nd nd nd nd nd nd nd nd nd nd nd nd 9.5 9.6 8.9 8.9 9.1 9.4 9.7 9.8 9.7 9.6 9.4 10.0 10.0 10.5 10.3 10.0 10.7 11.0 10.6 10.9 10.1 10.5 9.4 9.8 9.6 9.4 8.8 9.6 9.5 9.7 9.9 9.9 10.7 11.3 11.8 12.2 12.4 11.8 10.4 9.2 9.1 9.2 9.1 8.9 8.7 8.5 8.9 9.4 9.4 9.0 9.0 9.3 7.2 8.1 8.6 9.2 9.0 9.0 8.8 9.0 8.9 0.1 0.3 0.2 0.1 0.2 <0.05 <0.05 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.3 0.2 0.4 0.1 0.4 <0.05 0.1 <0.05 <0.05 0.1 0.2 nd 0.1 nd <0.05 0.1 0.1 0.1 0.1 0.1 0.4 0.3 <0.05 0.1 0.1 <0.05 <0.05 0.1 0.2 <0.05 0.1 nd nd nd nd nd nd nd nd nd nd nd nd nd 1 Subjects A–H, mothers; Subjects a–h, infants. Measurements were made in triplicate except where indicated (nd). A sampling time of 0 ¼ birth; negative week values are sampling times before birth. This decline in maternal d15N values can be attributed to the pregnancy effect, and is likely the result of the maternal conservation of nitrogen for the growing fetus during 287 gestation (Fuller et al., 2004). Two subjects (c and h; Figs. 3B, 8B) have in utero d15N results that seem to track the decrease in maternal values, and Millard (2000) noted that the three infants in his study all had lower fetal d15N values than predicted. These ﬁndings suggest the possibility that fetal d15N values parallel the decline in maternal d15N results during pregnancy. However, more speciﬁc research is needed to discover if there is a link between the falling d13C and d15N values and maternal metabolism, and to understand why some subjects show evidence of a decrease in fetal d15N values, while others do not. Breastfeeding, weaning, and infant d15N values The ﬁngernails of exclusively breastfed subjects display the expected d15N increase during breastfeeding, but there is a wide range in 15N-enrichments (1.7–2.8%). This greater variability can be attributed to the fact that in addition to diet, d15N values are inﬂuenced by the nitrogen balance of an individual. Positive nitrogen balance (growth) was shown to decrease d15N values, and negative nitrogen balance (nutritional stress) results in increased d15N values (Hobson and Clark, 1992; Hobson et al., 1993; Gannes et al., 1997; Fuller et al., 2004). Additional research is necessary to better understand the relationships among infant weight gain, diet (maternal breast milk protein content), and 15N-enrichment during breastfeeding. During the weaning process, all of the infant d15N values decline toward the maternal values, and this reﬂects the consumption of foods with lower d15N values than breast milk protein (d15N ¼ 8–9%, Fuller, 2003). This decrease in d15N during weaning was slower compared to the d13C results, and this will be discussed in detail below (Table 3). The d15N values of the two breast- and formula-fed subjects (f and g) are also 15N-enriched by approximately 1.0% and 1.4%, respectively. This reduced 15N-enrichment shows that the infant tissue nitrogen is being derived from both the breast milk protein (d15N ¼ 8–9%; Fuller, 2003) and the bovine milk formula (d15N ¼ 4–5%). It is signiﬁcant to note that after being breast- and bottlefed from birth, Subject g switched to a period of exclusive breastfeeding between 25–34 weeks (Fig. 7b). During this interval, the d15N ( 3%) results increased to levels seen in infants who were exclusively breastfed from birth. Thus, as seen in a previous study of horse hair (West et al., 2004), the infant ﬁngernail data here also support the ﬁnding that consumer protein is very quick to register the isotopic signals of a dietary change. The only exclusively formula-fed infant (Subject h) shows no increase in d15N, and actually decreases by approximately 2.0% (Fig. 8B). This decrease in d15N during exclusive formula-feeding conﬁrms that the 15Nenrichment observed during exclusive breastfeeding is the result of the trophic level effect (Fogel et al., 1989). The formula (Carnation Good Start1) consumed by Subject h had a d15N value of 4.3% (Table 4). Thus, the d15N value of 7.2% at 20 weeks is 2.9% enriched above the formula (Fig. 8b), and this is near the high end of 15N-enrichments (1.7–2.8%) for exclusively breastfed Subjects a–e. Since infant formula is usually higher in protein content compared to breast milk, it is hypothesized that bottle-fed infants may have higher diet-to-tissue-trophic-level fractionation factors compared to breastfed infants. More research is necessary to explore this possibility. 288 B.T. FULLER ET AL. TABLE 2. Timing of weaning process and the foods consumed by exclusively breastfed Subjects a–e Subject Period of exclusive breastfeeding1 a 12–30 weeks b 12–32 weeks c 14–47 weeks d 12–32 weeks e 12–19 weeks Timing of weaning process and foods consumed Cessation of breastfeeding Weaning week 30: Rice cereal started once a day. Week 35: Peas and bananas started once a day. Week 48: Rice cereal, bananas, peas, applesauce, and sweet potatoes. Week 57: Oatmeal, peas, bananas, spaghetti, applesauce, chicken pasta, and 440 ml of Carnation Good Start1 a day. Week 66: All adult table foods eaten and 440 ml/day of Carnation Good Start1. Weaning week 32: Rice cereal started once a day. Week 42: Rice, oat, and barley cereal, green beans, and corn tortillas. Week 44: Started 120–240 ml/day of Enfamil1 with iron. Week 52: Rice and oat cereal, applesauce, chicken, peas, carrots, sweet potatoes, peaches, pears, bananas, and 120–240 ml/day of Enfamil1 with iron. Week 60: All adult table foods eaten plus 120–240 ml/day of Enfamil1 with iron. Week 64: All adult table foods eaten plus 300 ml/day of Enfamil1 with iron, and 300 ml/day of lowfat cow’s milk. Week 76: All adult table foods eaten. Weaning week 47: 1 cup of rice or wheat cereal in breast milk started once a day, carrots, and squash. Week 60: 1 cup of rice or wheat cereal in breast milk, carrots, squash, apples, pears, crackers, grape juice, and apple juice. Week 68: Rice or oat cereal, cheese, crackers, yogurt, apples, pasta, tomatoes, chicken, rice, corn chips, and bagels. Week 78: All adult table foods eaten. Weaning week 32: Rice cereal, oatmeal, apples, bananas, peaches, apricots, and blueberries. Week 46: Turkey and egg yolk started. Week 50: Cottage cheese, tofu, sweet potatoes, green beans, broccoli, cauliﬂower, and yogurt. Week 60: All adult table foods eaten. Weaning week 19: Started 120 ml/day of Similac1 with iron. Week 30: Rice cereal started once a day. Week 40: Rice cereal, carrots, peas, apples, bananas, and formula-feeding with Similac1 with iron. Week 66 Week 56 Subject still breastfeeding at every meal at time of last sampling. Week 67 Week 33 1 There is a time lag between period of ﬁngernail growth and sampling. Real-time isotopic signatures can be obtained by subtracting 12 weeks (Subjects a, b, d, e) or 14 weeks (Subject c) from times listed. Breastfeeding, weaning, and infant d13C values The ﬁndings of this study clearly demonstrate an intrinsic 13C-enrichment of 1% in the ﬁngernail proteins of exclusively breastfed human infants. The breast- and formula-fed subjects also displayed 13C-enrichments of approximately 0.5%, while the exclusively bottle-fed individual showed no increase in d13C values (Fig. 8A). This discovery of a carbon trophic-level shift in modern breastfeeding infants is in contrast to the results of Fogel et al. (1989), and suggests that 13C-enrichments of approximately 1% in archaeological populations are not necessarily indicative of the consumption of C4-based weaning foods such as maize or millet. During the weaning process, the d13C results for breastfed infants declined to maternal levels more rapidly than the d15N results (Table 3). This faster return of infant d13C values to maternal d13C values was seen in all subjects except Subject e. Subject e was weaned the most abruptly and displayed a nearly simultaneous decrease in both d13C and d15N values (Fig. 5A,B). The cause of this faster decline in d13C values during weaning is not fully understood, but we speculate that it could be the result of a combination of two possible mechanisms. First, infants’ 13C-enrichment over the mothers’ ﬁngernails is only 1% compared to 2–3% for 15N-enrichment. Thus, this greater increase in infant d15N values during breastfeeding also means that it will take longer for d15N values to return to maternal levels during weaning. Second, since the weaning foods (rice cereal, oatmeal, fruits, and vegetables) that these subjects consumed are more rich in carbohydrates than protein, the dilution in the infant carbon pool will be greater when compared to 289 ISOTOPIC SIGNATURES OF BREASTFEEDING AND WEANING TABLE 3. Time when breastfeeding infant C and N values return to maternal levels during the weaning period1 13 15 Subject Time when infant d13C values return to maternal d13C levels Time when infant d15N values return to maternal d15N levels a 33 weeks b c 37 weeks 81 weeks d 39 weeks e f 27 weeks 32 weeks g h 38 weeks NA Still elevated (0.5%) at 67 weeks 50 weeks Still elevated (1.4%) at 81 weeks Still elevated (0.9%) at 85 weeks 27 weeks Still elevated (1.1%) at 51 weeks 50 weeks NA 1 For d15N results, some subjects are still elevated at time of last sampling. Subject h was formula-fed from birth, so no results could be obtained (NA). the nitrogen pool, which is still being primarily derived from the high-quality protein in breast milk. While no isotopic measurements were made on the weaning foods, d13C results from a previous study (Schoeller et al., 1986) of cereals, fruits, and vegetables from Chicago, Illinois, showed that these foods (which are similar to the weaning foods) are 13C-depleted (23% to 29%) compared to the breast milk protein (19.5%) and whole breast milk (22% to 23%) that these infants were consuming (Fuller, 2003). Thus, this faster decrease in d13C could be the result of the more pronounced dilution of breast milk carbon with the 13C-depleted carbon content in the weaning foods. It is hoped that advances in the isotopic measurement of carbon and nitrogen single amino acids (Fogel and Tuross, 2003; Corr et al., 2005; McClelland and Montoya, 2002; Schmidt et al., 2004) and carbohydrates will permit a more detailed investigation of this topic by tracing how breast milk and weaning foods are incorporated into infant protein. The ﬁndings that d13C values are elevated by approximately 1% during exclusive breastfeeding and that d13C values return more quickly than d15N values to maternal levels during weaning have important ramiﬁcations for the interpretation of breastfeeding and weaning patterns in archaeological populations. Currently, the most accepted method for determining the age of weaning is by measuring d15N values in bone collagen from different age classes of infants and children, and noting the age at which d15N values return to those of the adult female population (Katzenberg et al., 1993; Schurr, 1997; Dupras et al., 2001; Mays et al., 2002; Schurr and Powell, 2005). Little signiﬁcance has been attributed to bone collagen d13C values, as they rarely display unique isotopic patterns similar to those of d15N results, although a number of stud- 13 ies noted C-enrichments (0.5–1.4%) and attributed these to either a possible tropic-level effect of breastfeeding or the consumption of C4-based foods (maize or millet) (Katzenberg et al., 1993; Katzenberg and Pfeiffer, 1995; Wright and Schwarcz, 1999; Dupras et al., 2001; Richards et al., 2002). In addition, research from dentine serial sections from the Medieval site of Wharram Percy, where no C4 dietary items were present, identiﬁed a 13C-enrichment (1%) that was likely the result of breastfeeding (Fuller et al., 2003). Based on the d13C results from this current study, we suggest that infant bone collagen d13C values might yield information that is as important as d15N values in determining ancient breastfeeding and weaning patterns. In particular, the fact that d13C values return faster than d15N values to maternal levels suggests that d13C values can be used as a possible indicator for the introduction of solid foods to the diet, whereas d15N values seem to be better suited to monitor the duration of breast milk consumption or the nitrogen trophic-level effect of breastfeeding. Subject a (Fig. 1A,B) provides a perfect example for this argument. For Subject a, the point at 33 weeks is higher than the mean maternal d15N value by 2.2% but equal to the mean maternal d13C value. If the d13C and d15N values for Subject a at point 33 were bone collagen results, the elevated d15N value would be interpreted as evidence of exclusive breastfeeding, but there would be no mention of d13C value, since it is equal to the mean maternal d13C value. However, using the present study as a model, it would be possible to correctly identify this individual as eating solid foods while still breastfeeding. The fact that many archaeological studies on weaning found elevated d15N values without increased d13C values suggests the possibility that the infants were dying while still breastfeeding but were also in the process of consuming solid foods. From a medical point of view, this observation is not surprising or unfounded, since the introduction of solid foods to the infant diet is clearly linked to an increased risk of infant mortality because of the exposure to new pathogens, nutritional stress, and a decrease in immunity levels (Knodel and Kintner, 1977; Katzenberg et al., 1996; Lönnerdal, 2000). However, given the modest sample size of this study, caution should be applied when using these results as the only models for archaeological interpretation. More isotopic research is needed on modern infants from outside North America (from countries where there is less consumption of C4 foods) to conﬁrm if d13C values can be used as a unique marker for the introduction of solid foods during weaning. In addition, it would be interesting to examine how the consumption of 13Cenriched C4-derived weaning foods (such as maize and millet) affects the isotopic results of modern infants. In conclusion, the combined analysis of infant d13C and d15N values has the potential to become an enhanced technique for examining both the timing of the weaning process and the cessation of breastfeeding in modern and past populations. TABLE 4. 13C and 15N results from bovine milk-based and soy-based infant formulas from Northern California Formula d13C (%) 6SD d15N (%) 6SD Enfamil (soy-based) Enfamil1 (milk-based) Enfamil1 (milk-based with iron) Similac1 (milk-based with iron) Similac1 (milk-based low iron) Isomil1 (soy-based) Carnation Good Start1 (milk-based) 19.8 19.9 23.1 23.2 22.7 21.7 21.9 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.5 7.5 4.5 4.9 4.7 0.2 4.3 0.3 0.4 0.3 0.2 0.3 0.4 0.1 1 290 B.T. FULLER ET AL. Figs. 6A,B. Fingernail d13C (a) and d15N (b) results from Subject F (mother) and Subject f (infant) where Subject f was breastand formula-fed between 12–30 weeks. Compared to the exclusively breastfed infants (Subjects a–e), Subject f displayed reduced 13 C- and 15N-enrichments. Speciﬁc dietary information for Subject f is discussed in Results. Formula The d13C and d15N results from infant formulas show that there are isotopic distinctions between the soy- and cow milk-based products (Table 4). While the number of formulas studied was small (n ¼ 7), modern infants who consume soy-based formulas should display tissue d15N values that are lower compared to individuals fed bovine milk-based supplements, and this could potentially distinguish among subjects fed these different formulas. However, more isotopic research is necessary on infant formulas from different locations to determine the extent of differences in isotopic signatures between soy- and milkbased formulas. ISOTOPIC SIGNATURES OF BREASTFEEDING AND WEANING 291 Figs. 7A,B. Fingernail d13C (a) and d15N (b) results from Subject G (mother) and Subject g (infant) where Subject g was breast- and formula-fed between 12–25 weeks. Between 25–34 weeks, Subject g was exclusive breastfed and displays increased 13C- (1%) and 15 N-enrichments (3%) compared to the breast- and formula-fed period. Speciﬁc dietary information for Subject g is discussed in Results. CONCLUSIONS We conducted a longitudinal study of d13C and d15N signatures in eight mother-infant pairs where infants were exclusively breastfed, breast- and formula-fed, or exclusively formula-fed from birth. All exclusively breastfed subjects displayed enrichments in 13C (1%) and 15N (2–3%) compared to maternal values. Individuals who were fed both breast milk and formula showed reduced enrichments in 13C and 15N, but the infant who exclusively consumed formula from birth displayed no trophic-level enrichments in carbon or nitrogen compared to maternal values. These ﬁndings demonstrate a dual increase in d13C and d15N values of tissue proteins during nursing, and show that 13Cenrichments of 1% in archaeological populations are not necessarily the result of a weaning diet based on maize or millet. 292 B.T. FULLER ET AL. Figs. 8A,B. Hair d13C (a) and d15N (b) results from Subject H (mother) and Subject h (infant) where Subject h was exclusively formula-fed between 0–22 weeks and does not show elevated d13C and d15N values compared to maternal levels. Speciﬁc dietary information for Subject h is discussed in Results. During the weaning process, the d13C results for breastfed infants declined to maternal levels more rapidly than the d15N results. Thus, based on the results of this small study, d13C values seem to be more sensitive to the introduction of solid foods to the diet, whereas d15N values appear to monitor the duration of breast milk consumption by an infant or the trophiclevel effect of breastfeeding. While more research is needed from breastfed infants fed isotopically different diets, it is likely that the combined analysis of d13C and d15N values will reﬁne the interpretation of infant nutrition in past populations. In addition, this technique is simple, inexpensive, and noninvasive, and can be used in medical and forensic research to detect breastfeeding and weaning patterns in modern human populations. ISOTOPIC SIGNATURES OF BREASTFEEDING AND WEANING ACKNOWLEDGMENTS We are greatly indebted to all parents and subjects who participated in this study. Without their help, interest, and enthusiasm, this research would not have been possible. We gratefully acknowledge the help of Nancy Sage, and all the nurses of the OB/GYN Department at Sutter Auburn Faith Hospital in the collection of dietary histories and samples. Thanks go to Peter Ditchﬁeld, Martin Humm, Angela Bowles, and Christine Tompkins for technical assistance and advice during isotopic analysis. This manuscript was greatly improved by the comments and suggestions of Mandy Jay, Vaughan Grimes, and three anonymous reviewers. LITERATURE CITED Corr LT, Sealy JC, Richards MP, Evershed RP. 2004. A novel marine dietary indicator utilising compound-speciﬁc bone collagen amino acid d13C values of ancient humans. J Archaeol Sci 32:321–330. DeNiro MJ, Epstein S. 1978. Inﬂuence of diet on distribution on carbon isotopes in animals. Geochim Cosmochim Acta 42:495–506. DeNiro MJ, Epstein S. 1981. Inﬂuence of diet on the distribution of nitrogen isotopes in animals. Geochim Cosmochim Acta 45:341–351. Dupras TL, Schwarcz HP, Fairgrieve SI. 2001. Infant feeding and weaning practices in Roman Egypt. Am J Phys Anthropol 115:204–211. Fogel ML, Tuross N. 2003. Extending the limits of paleodietary studies of humans with compound speciﬁc carbon isotope analysis of amino acids. J Archaeol Sci 30:535–545. Fogel ML, Tuross N, Owsley D. 1989. Nitrogen isotope tracers of human lactation in modern and archaeological populations. Carnegie Inst Yr Bk 88:111–117. Fuller BT. 2003. Isotopic investigation of pregnancy and breastfeeding in modern and archaeological populations. D.Phil. dissertation, University of Oxford, UK. Fuller BT, Richards MP, Mays SA. 2003. Stable carbon and nitrogen isotope variations in tooth dentine serial sections from Wharram Percy. J Archaeol Sci 30:1673–1684. Fuller BT, Fuller JL, Sage NE, Harris DA, O’Connell TC, Hedges REM. 2004. Nitrogen balance and d15N: why you’re not what you eat during pregnancy. Rapid Commun Mass Spectrom 18:2889–2896. Gannes LZ, O’Brien DM, Martinez del Rio C. 1997. Stable isotopes in animal ecology: assumptions, caveats, and a call for more laboratory experiments. Ecology 78:1271–1276. Herring DA, Saunders SR, Katzenberg MA. 1998. Investigating the weaning process in past populations. Am J Phys Anthropol 105:425–439. Hobson KA, Clark RG. 1992. Assessing avian diets using stable isotopes II: factors inﬂuencing diet-tissue fractionation. Condor 94:189–197. Hobson KA, Alisauskas RT, Clark RG. 1993. Stable nitrogen isotope enrichment in avian tissues due to fasting and nutritional stress: implications for isotopic analyses of diet. Condor 95:388–394. Katzenberg MA, Pfeiffer S. 1995. Nitrogen isotope evidence for weaning age in a nineteenth century Canadian skeletal sample. In: Grauer AL, editor. Bodies of evidence: reconstructing history through skeletal analysis. New York: Wiley-Liss. p 221–235. Katzenberg MA, Saunders SR, Fitzgerald WR. 1993. Age differences in stable carbon and nitrogen isotope ratios in a population of prehistoric maize horticulturists. Am J Phys Anthropol 90:267–281. Katzenberg MA, Herring DA, Saunders SR. 1996. Weaning and infant mortality: evaluating the skeletal evidence. Yrbk Phys Anthropol 39:177–199. Kelly JF. 2000. Stable isotopes of carbon and nitrogen in the study of avian and mammalian trophic ecology. Can J Zool 78:1–27. 293 Knodel J, Kintner H. 1977. The impact of breastfeeding patterns on the biometric analysis of infant mortality. Demography 14:391–409. Kohn MJ. 1999. You are what you eat. Science 283:335–336. Lönnerdal B. 2000. Breast milk: a truly functional food. Nutrition 16:509–511. Mays SA, Richards MP, Fuller BT. 2002. Bone stable isotope evidence for infant feeding in Mediaeval England. Antiquity 76:654–656. McClelland JW, Montoya JP. 2002. Trophic relationships and the nitrogen isotopic composition of amino acids in Plankton. Ecology 83:2173–2180. Millard AR. 2000. A model for the effect of weaning on nitrogen isotope ratios in humans. In: Goodfriend G, Collins M, Fogel M, Macko S, Wehmiller J, editors. Perspectives in amino acid and protein geochemistry. New York: Oxford University Press. p 51–59. Minagawa M, Wada E. 1984. Stepwise enrichment of 15N along food chains: further evidence and the relation between d15N and animal age. Geochim Cosmochim Acta 48:1135–1140. O’Connell TC, Hedges REM. 1999. Investigations into the effects of diet on modern human hair isotopic values. Am J Phys Anthropol 108:409–425. O’Connell TC, Hedges REM, Healey MA, Simpson AHRW. 2001. Isotopic comparison of hair, nail, and bone: modern analyses. J Archaeol Sci 28:1247–1255. Richards MP, Mays S, Fuller BT. 2002. Stable carbon and nitrogen isotope values of bone and teeth reﬂect weaning age at the Mediaeval Wharram Percy Site, Yorkshire, UK. Am J Phys Anthropol 119:205–210. Runne U, Orfanos CE. 1981. The human nail: structure, growth, and pathological changes. Curr Probl Dermatol 9:102–149. Saitoh M, Uzuka M, Sakamoto M, Kobori T. 1969. Rate of hair growth. In: Montagna W, Dobson RL, editors. Hair growth. Oxford: Pergamon Press. p 183–201. Schmidt K, McClelland JW, Mente E, Montoya JP, Atkinson A, Voss M. 2004. Trophic-level interpretation based on d15N values: implications of tissue-speciﬁc fractionation and amino acid composition. Mar Ecol Prog Ser 266:43–58. Schoeller DA, Minagawa M, Slater R, Kaplan IR. 1986. Stable isotopes of carbon, nitrogen, and hydrogen in the contemporary North American human food web. Ecol Food Nutr 18:159–170. Schoeninger MJ. 1995. Stable isotope studies in human evolution. Evol Anthropol 4:83–98. Schoeninger MJ, DeNiro MJ. 1984. Nitrogen and carbon isotopic composition of bone collagen from marine and terrestrial animals. Geochim Cosmochim Acta 48:625–639. Schurr MR. 1997. Stable nitrogen isotopes as evidence from the age of weaning at the Angel Site: a comparison of isotopic and demographic measures of weaning age. J Archaeol Sci 24: 919–927. Schurr MR. 1998. Using stable nitrogen isotopes to study weaning behavior in past populations. World Archaeol 30: 327–342. Schurr MR, Powell ML. 2005. Role of changing childhood diets in the prehistoric evolution of food production: an isotopic assessment. Am J Phys Anthropol 126:278–294. Schwarcz HP, Schoeninger MJ. 1991. Stable isotopic analyses in human nutritional ecology. Yrbk Phys Anthropol 34:283– 321. West AG, Ayliffe LK, Cerling TE, Robinson TF, Karren B, Dearing MD, Hleringer JR. 2004. Short-term diet changes revealed using stable carbon isotopes in horse tail-hair. Funct Ecol 18:616–624. Wright LE, Schwarcz HP. 1998. Stable carbon and oxygen isotopes in human tooth enamel: identifying breast-feeding in prehistory. Am J Phys Anthropol 106:1–18. Wright LE, Schwarcz HP. 1999. Correspondence between stable carbon, oxygen and nitrogen isotopes in human tooth enamel and dentine: infant diets at Kaminaljuyu. J Archaeol Sci 26:1159–1170.