limited to urinary sucrose and fructose. These biomarkers have been shown to relate
to intake (Tasevska et al. 2005 ), but imprecision has been demonstrated, suggesting
that only about 41.7% of excreted fructose plus sucrose can be explained by esti-
mations of intake (Song et al. 2013 ). Identification of reliable and valid biomarkers
to predict sucrose and fructose intake is of value given compelling evidence that
high intake of sugar and/or foods and beverages concentrated in sugars is likely
contributory to obesity (Te Morenga et al. 2013 ).
Concentration and Replacement Dietary Biomarkers
Concentration biomarkers are surrogate indicators of nutrient or bioactive“levels”
in the body. Generally measured in plasma or serum, these measures correlate with
nutrient intake but generally with lower correlation coefficients (<0.60) than
recovery or predictive biomarkers. The lower correlations are thought to be a result
of the effects of metabolism and bioavailability on concentration measures. In fact,
an advantage of concentration biomarkers over dietary self-report is the fact that
such measurements provide a more biologically relevant exposure estimate given
that concentration values reflect individual variance introduced by differences in
absorption, microbiota influences, nutrient–nutrient interactions, medication use,
tissue turnover, metabolism, and excretion (Holst and Williamson 2008 ). Common
concentration biomarkers include serum or plasma vitamins (e.g., serum ascorbic
acid, plasma B12, or plasma 25-OH vitamin D), RBC fatty acids (EPA, DHA, etc.),
and serum or plasma carotenoids (as indicators of fruit and/or vegetable intake or
supplementation). Recovery biomarkers are similar to concentration markers in
terms of correlating with intake of a select food nutrient and/or food constituent.
However, these concentration biomarkers generally refer to biomarkers used for
foods where nutrient composition data are lacking (Grace et al. 2004 ). Bioactive
compounds in foods such as phytoestrogens and alkylresorcinols as measures of
soy and grain intake, respectively, fall into this category of biomarkers.
Concentration biomarkers have been widely applied in nutritional epidemio-
logical research in an effort to quantify and qualify diet–disease associations. As an
example, a study of women previously treated for breast cancer showed a 43%
reduced risk for breast cancer recurrence in women who entered the study with the
highest plasma carotenoid concentrations (Rock et al. 2005 ). Of note, self-reported
carotenoid intake was not significantly associated with plasma levels or breast
cancer recurrence risk in this same sample of women (Natarajan et al. 2006 ).
Contrary to this, in the CAREDS trial, where serum as well as dietary intake of
lutein and zeaxanthin were assessed in relation to the formation of nuclear cataracts
in older adults, both diet and serum measures were inversely associated with risk,
although risk reduction was higher in relation to serum concentrations (32% lower
risk) as compared to dietary associations (23% lower risk) (Moeller et al. 2008 ).
Concentration biomarkers for antioxidant nutrients have also been investigated in
relation to surrogates of morbidity and mortality, including telomere length (Sen
9 Biomarkers of Diet and Nutritional Health 183