Use of Biological Markers to Validate Dietary Intake Measurements

Nutritional biomarkers are those elements or compounds in biological samples capable of reflecting relationships between diet, nutritional status, and disease processes. Not all biomarkers are suitable for use in dietary validation studies. One of the key features of a marker should be its ability to reflect intake over a wide range of intakes.

Figure 2 shows the sequence whereby food or drink containing potential biomarkers (nutrients and nonnutrients) may be ingested, absorbed, distributed, and excreted. The stages in the top of Figure 2 are not usually amenable to sampling (e.g., taking samples of gastric contents), and it is only in the later stages in which compounds are in circulation, present or stored in tissues, or excreted that they are more readily sampled. The complexity of the relationship between intake and the measured levels of biomarkers in these lower stages, however, may limit the usefulness of certain compounds in validation studies. Figure 3 illustrates for four nutrients the varying relationships between tissue levels

Food, drink, nutrient/^ \or non-nutrient /

Ingestion

Digestion

Absorption

Transportation

Distribution Utilization Storage Mobilization Metabolism Excretion

Figure 2 Stages in the pathway between intake and measurement of biomarkers.

and levels in diet. Vitamin E shows a more or less linear relationship between blood and dietary levels across a broad range of intakes. Riboflavin appears in urine only after tissues are saturated and any excess is excreted. An alternative measure to assess riboflavin, erythrocyte glutathione reductase activity coefficient), is a sensitive biomarker of intake only at low levels. Vitamin C in blood is a poor marker at low levels of intake, increases in sensitivity as a marker of intake across the middle range of intakes, and is poor once again at high levels of intake where excess vitamin is excreted in the urine. Retinol is stored in the liver and its level in blood is controlled

homeostatically. It is therefore a poor marker above relatively low intakes. A further problem is the point in time or span of intake reflected by the marker. Figure 4 shows that some markers relate to intakes days or weeks prior to the sampling point (e.g., energy and doubly labeled water), whereas other may reflect intake over months (iron intake reflected by ferritin) or years (calcium intake reflected by bone mass). The influence of other factors relating marker to intake (e.g., hem versus non-hem iron in the diet, and the influence of vitamin C or dietary fiber on absorption) may undermine the ability to conclude that a low measurement of a marker is necessarily a reflection of low dietary intakes. Thus, the value of markers in assessing the validity of intake measurements is often limited to specific ranges of intake in diets of known composition. This may not be sufficient for epidemiological purposes, where the entire range of intake may be of interest in relation to disease risk.

The two most widespread uses of biomarkers for the assessment of the validity of measures of diet are to identify under- or overreporters and to assess the correctness of ranking of individuals according to their nutrient intake.

Techniques for Identifying Under- and Overreporters

Doubly labeled water The scientific basis that underlies the doubly labeled water method for estimating energy expenditure relies on the differential rates of loss of hydrogen and oxygen from the body at different levels of energy expenditure. Hydrogen is lost primarily in water, whereas oxygen is lost in

Intake

Intake

High

Intake

Intake

High

High

High

Figure 3 Associations between intake and biomarker over a wide range of intakes: vitamin E:cholesterol ration, vitamin C, riboflavin, and retinol. EGRAC, erythrocyte glutathione reductase activity coefficient. (Adapted from Kohlmeier L (1991) What you should know about your biomarker. In: Kok FJ and van't Veer P (eds.) Biomarkers of Dietary Exposure. London: Smith Gordon.)

Days | Weeks | Months | Years

Energy

Doubly labeled water

Body weight

Fatty acids

Cholesterol Erythrocyte esters membranes

Adipose tissue

Tocopherols

Serum

Adipose tissue

Retinol

Liver tissue

Carotenoids

Plasma

Adipose tissue

Vitamin C

Urine Leucocytes Plasma

Iron

Hemoglobin

Ferritin

Calcium

Urine

Bone mass

Selenium

Erythrocyte Plasma glutathione peroxidase

Toenails

Figure 4 Time scale over which different biomarkers may reflect the relationship with intake.

Figure 4 Time scale over which different biomarkers may reflect the relationship with intake.

both water and carbon dioxide. The relative rates of loss can be used to estimate energy expenditure in free-living subjects over a period of approximately 2 weeks, thereby providing a reference measure for energy intakes over a similar period. Provided the subject is in energy balance (neither gaining nor losing weight due to changes in body composition), the measures of expenditure and intake should agree. The technique allows for identification of both under- and overreporters. The boundaries of acceptability of the test measures (e.g., to within ±10% of habitual energy intake) need to be chosen according to the needs of the study in which the validity of the test measure is being assessed. The level of agreement between test and reference measure will dictate both the precision of the estimate of mean intake for an individual or subgroup and the extent to which subjects will be misclassified when ranked according to level of intake.

The main disadvantage of the doubly labeled water technique is its high cost. In a large-scale study, it is not feasible to use doubly labeled water with every subject in order to assess the completeness of dietary records. The technique is therefore usually used to assess validity of the test measure in a sample of subjects who are taken to be representative of the sample for the main study.

Another disadvantage is that doubly labeled water provides a marker for energy only. The diet recorded could differ substantially from the subject's usual diet but have an energy content in agreement with the estimate of energy expenditure. In the absence of additional information about usual patterns of food consumption, such a record would be regarded as valid. A further problem is that not all food consumption or nutrient intake correlates strongly with energy intake. For example, fruits and vegetables and their associated nutrients (e.g., vitamin C, beta-carotene, and potassium) may be overreported in a dietary assessment in which energy intake agrees well with energy expenditure, but the over-reporting would not be identified. These comments are summarized in Table 4.

Urinary nitrogen and potassium excretion A second technique for identifying under- or overreporters is to collect 24-h urine samples and compare the amounts of nitrogen and potassium excreted with the amount ingested. Allowing for incomplete absorption and losses of nitrogen from the gastrointestinal tract (digestive juices and shed epithelial cells), hair, skin, and sweat, the amount of nitrogen excreted should be approximately 81% of the nitrogen ingested. Allowing for daily variations in intake and excretion, if daily recorded intake of nitrogen is less than 70% of the corresponding urinary nitrogen excretion over the following 24 h, the respondent is likely to have underreported his or her usual consumption; anyone whose recorded intake of nitrogen is more than 100% of their excretion is likely to have overreported their consumption. The more days of intake and excretion data that are collected, the better the agreement over the recording period should be for subjects who are in nitrogen balance. If urinary nitrogen is to be used as a marker for the completeness of dietary recording, it is helpful to have at least 4 days' worth of data (diet and urine). For potassium, the expected urinary excretion is 95% of the intake, with limits of 80 and 110% for 'good' reporting.

As with doubly labeled water, it is assumed that the subject is in balance, neither losing nor gaining body nitrogen or potassium.

Table 4 Limitations of biological reference methods appropriate for validation of dietary assessment measures

Reference method

Limitations

Doubly labeled water

Urinary nitrogen: completeness of samples confirmed using PABA Urinary nitrogen only

Biochemical measurements of nutrients in blood or other tissues

Energy intake:BMR ratio

Energy only

Assumptions of model regarding water partitioning may not apply in cases of gross obesity, high alcohol intake, or use of diuretics Very expensive

Analysis technically demanding Protein only

PABA analysis affected by paracetamol and related products Protein only

Danger of incomplete samples

Complex relationship with intake mediated by digestion, absorption, uptake, utilization, metabolism, excretion, and homeostatic mechanisms Cost and precision of assays Invasive

Imprecision of estimate of BMR based on body weight and regression equations Single cutoff point (e.g., EI:BMR <1.1) will not identify low-energy reporters with higher habitual energy expenditures Higher estimates of cutoff (e.g., EI:BMR <1.2) captures more true low energy reporters but also more good reporters

It is important to ensure that the urine collections are complete. This necessitates the use of an inert metabolic marker (para-amino benzoic acid (PABA)), which is rapidly absorbed and excreted. Subjects take a divided dose of 240 mg PABA throughout the day. At least 85% of the PABA should be recovered in the urine in a 24-h collection. If the amount recovered is less than 85%, the urine sample may be regarded as incomplete and therefore not suitable for analysis of nitrogen in order to check the completeness of the dietary record. Because paracetamol and related compounds interfere with the PABA assay, a measure of excretion over 115% of the administered dose would be suspect.

As with doubly labeled water, the principal weakness of urinary nitrogen as a marker for the completeness of dietary records is that many foods contain low levels of nitrogen but may be important sources of other nutrients. Any check for the completeness of dietary records based on nitrogen will not assess the presence or absence of these other foods. Also, the issue of dietary distortion is not addressed. Potassium is more widespread in foods, although the largest contributors are usually fruits and vegetables. Using urinary nitrogen and potassium in combination gives a better assessment of the completeness of the recording than any single marker.

Ratio of energy intake to basal metabolic rate Doubly labeled water and urinary nitrogen excretion are particularly useful for assessing the validity of prospec-tively recorded diets because the time frame of the test and reference measures can be made to coincide. A third technique for assessing validity can be used with both prospective records and recalls of diet. It is based on energy and thus has the limitations of a validating marker relating to a single dietary factor. It has the advantage, however, of being able to be applied to all subjects in a dietary survey because no external reference measure is needed. It is a biomar-ker in the sense that it relies on a biological measure (body weight) and is best applied when measures of physical activity at the individual level are also available. The assumption is that there should be reasonable agreement between estimated requirement and estimated intake.

Schofield equations can be used to estimate basal metabolic rate (BMR) based on age, gender, and body weight. An individual whose reported energy intake is below the level of energy expenditure likely to be needed to carry out day-to-day activities has probably underreported his or her diet. A typical cutoff point for an acceptable ratio of the energy intake to BMR ratio in an individual is 1.2, taking into account daily variations in energy intake over a period of 7 days of dietary recording and allowing for the inaccuracies of the estimate of BMR based on the Schofield equations. A cutoff of 1.2 will identify only those subjects who under-report and whose levels of activity are low. For subjects with higher levels of activity (e.g., estimated from questionnaire responses), proportionately higher cutoff points are appropriate. It is also possible to estimate an upper probable level of energy expenditure (e.g., 2.5 times BMR, depending on habitual level of activity) and subjects with reported levels of intake over this value may be regarded as being overreporters.

Diet Tweak System

Diet Tweak System

Trying To Lose Weight Can Be Tough. But... Not Losing Weight and Gaining What You Lost Back, Sucks. If you've ever felt that no matter what you do to lose weight nothing seems to work. If you've ever felt that there has got to be some kind of a system or way to lose weight...but just have not found it yet.

Get My Free Ebook


Post a comment