Approaches of Measuring Vitamin C in Epidemiological Studies

METHODOLOGICAL APPROACHES OF MEASURING VITAMIN C IN EPIDEMIOLOGICAL STUDIES 

There is no instrument available by which vitamin C exposure can be perfectly assessed, particularly if one assumes that chronic diseases require a long induction period. Nutritional epidemiologists were able to demonstrate that for many nutrients one measurement during a lifetime correlates well with long-term intake (Friedenreich et al.y 1992). They concluded that nutrition behavior during adulthood is a stable phenomenon related to the individual subject. This correlation of previous with current intake is also valid for vitamin C. Therefore, measurement instruments that cover even a short time period will be suitable tool to assess lifelong vitamin C exposure. 

Vitamin C exposure can be determined on different levels. One level relates to the amount that is ingested. Another level relates to the amount of vitamin C that is available in the body. A third level relates to particular metabolic functions which depend on vitamin C availability in the tissue. Intake measurements and blood 

The intake of vitamin C is estimated from food intake. Usually, first the intake of a food item is obtained, subsequently the amount of vitamin C is calculated by multiplying the amount of food consumed with its average vitamin C concentration. The latter information is usually derived from existing food tables or nutrient data banks (Sinha et aL, 1993). Food tables only reflect the average concentration however, in fruits and vegetables large variations in vitamin C concentrations can occur even if obtained from the same source (Vanderslice and Higgs, 1991). Table I shows the five major sources of vitamin C in different populations. 

Food frequency questionnaires are the research instruments most often used in epidemiological studies because of their easy application. However, the amount of misclassification of dietary exposure by this instrument may be greater than more elaborate measurement techniques such as biomarkers or long-term food records. The amount of misclassification of the more simple food frequency questionnaire is estimated in validation studies which show a high correlation between the different methodological approaches (Table II). The results of validation studies can be used to correct for misclassification in the statistical analyses. 

---