CARBON ISOTOPE COMPOSITION OF BLOOD FROM A LARGE POPULATION (N=200) OF HUMANS: IMPLICATIONS FOR INTERPRETATION OF DIET

Carbon isotope composition of living and ancient human tissues is commonly used for interpretation of diet. The largest datasets assembled, even from modern populations, approach only 26 subjects (Wilkinson et al, 2007), yet these results are often extrapolated to communities of hundreds to thousands of living and ancient human beings. In order to quantify the variation in δ¹³C value in tissues from a large population of humans of unknown, and therefore presumably varied diet, we performed carbon isotope analysis on 200 individuals using conventional blood-draw techniques. We report the range of these natural abundance carbon stable isotopes in human blood, and discuss the implications of these results on modern dietary studies.

200 blood samples (from 100 females, 100 males) were collected through the Johns Hopkins Medical Institutions. No blood was drawn specifically for this study; all samples were surplus from daily collection of blood. Samples were de-identified, and no information was available to the investigators beyond date of the blood sampling and the gender of the patient. Blood was drawn into tubes without anti-coagulation or other additives, such as EDTA, in order to avoid any carbon contribution from sources other than the subject's blood. For centrifugation, blood was transferred into 15mL polystyrene tubes and separated into serum and clot. Samples were combusted to CO₂ in a Eurovector elemental analyzer configured with a Stable Isotope Ratio Mass Spectrometer.

Carbon isotope determinations for serum and clot revealed the following range in δ¹³C value: -15.6‰ to -23.0‰. Median values as well as means, standard deviations, and ranges did not show significant differences between male and female, or serum and clot. Serum and clot δ¹³C values from each subject were highly correlated (females R² = 0.93; males R² = 0.92).

In summary, the total population studied shows a range in blood δ¹³C value of 7.2‰. Our finding that there is a range of δ¹³C values among various individuals is important to the hypothesis that δ¹³C values reflect dietary intake in modern or ancient human populations. The magnitude above range (7.2‰) is approximately half of the total range attributed across C3 and C4 plants. We discuss the application of these results in the health community as a possible biomarker for dietary intake.