Characteristics of a scientifically sound dietary biomarker include, but are not
limited to, a measure that has a strong correlation with a diet exposure, is reliable
and thus stable on repeat measurement, has known and acceptable individual and
population variance, can be easily accessed in terms of the necessary biological
tissue/sample for measurement, and is cost-effective (de Vries et al. 2013 ;
Greenwald et al. 2000 ). When selecting an appropriate biomarker for nutritional
research, several criteria should be considered (Table9.1). Interestingly, many
biomarkers that are adopted for use do not meet the selection criteria, but are used
regardless, especially as thefield is evolving.
Numerous biomarkers of dietary intake and/or nutritional status have been
developed. Table9.2summarizes the biomarkers of exposure for several nutrients
in the diet. The use of macronutrient dietary biomarkers has recently been reviewed
(Hedrick et al. 2012 ). Biomarkers of exposure generally correlate well with dietary
intake and for most people the lower circulating levels/concentrations are associated
with dietary inadequacy.
In addition, functional biomarkers have been developed that capture nutrient
status through measurement of a select biological function that is dependent on the
specific nutrient. Table9.3summarizes common functional biomarkers employed
in nutritional science. Functional biomarkers reflect the impact of a nutrient on
Table 9.1 Criteria for quality diet/nutritional biomarkers
Concentration of nutrient in circulating pool must be an estimate of overall status
Homeostatically regulated nutrients are not indicators of status (e.g., serum calcium)
Known coefficient of variance (below 10%)
Assessment for error structure (the rate at which randomness may be contributing to
results/findings)
Established validity
Established reliability
Acceptable sensitivity
Temporality with outcome demonstrated
Tested for colinearity utilizing statistical tests such as partial least squares regression (PLS) or
principal components analysis
Known confounders that influence measurement results (adiposity, age, gender, etc.)
Established modulation/change in health outcome or surrogate endpoint with specific dietary
exposures
Established association with disease or intermediary of disease risk
Differential expression of the biomarker in well-nourished/normal versus malnourished/high-risk
tissue/fluids
Mechanistic basis for effect; biological plausibility
Coherence with current knowledge, evidence
Dose-responsive, exposure-responsive/biological gradient
Technology available for measurement
Accessible in human tissue/fluids; small quantity of tissue specimen required
Cost-effective178 T.E. Crane and C.A. Thomson