While most people are looking forward to the festive break and the prospect of a good deal of feasting, there are others who dread this period of the calendar. Their trepidation has nothing to do with religion and everything to do with food. For those people suffering from eating disorders, Christmas time, or any other period of celebration, can be very difficult.

It is likely that we have all heard of eating disorders but think that they affect a small proportion of the population. This is strictly true, with only an estimated 1-4% of women in the US suffering from the two major disorders of anorexia nervosa and bulimia at some stage in their lives. But translate that percentage into real numbers for a US population of around 300 million and the figures become much more significant.

Anorexia nervosa sufferers see themselves as overweight, even though they may be disturbingly thin and adopt sparse eating habits. People with bulimia nervosa have sporadic periods of binge eating, often followed by compensatory behaviour such as self-induced vomiting to prevent the consequent gain in weight.

Currently, diagnosis relies on input from the patient regarding their eating patterns, as well as objective measures such as the body mass index. In many cases, patients do not recognise or admit to having any problems, making it more difficult for doctors to give the correct diagnosis. A clinical test would be of enormous help but there are no tests for eating disorders at this time.

Spurred on by this deficiency, researchers from Brigham Young University led by Kent Hatch have shown that a patient's hair may provide all the evidence needed to diagnose an eating disorder. The team were guided by archaeological studies on hair as well as studies on living humans which had shown that the dietary history of an individual is stored in the hair in the form of the stable isotope ratios of carbon and nitrogen.

In the normal anabolic state of the human body, dietary proteins are eventually incorporated into the hair. During this process, ¹⁴N is excreted at a faster rate than ¹⁵N, so the latter isotope becomes more concentrated in the hair. When the body is in a catabolic state and begins to use up its reserves of energy and protein, the ¹⁵N/¹⁴N ratio is increased further. So, the researchers decided to see whether nutritional stress could be identified from the isotope ratios in hair.

Hair was collected from 20 patients with anorexia nervosa and/or bulimia nervosa and from 23 non-clinical volunteers. Five hairs were pulled out by the roots, cleaned, and lined up from the root ends before being cut into 1-cm segments. The same section segments from one person were combined for analysis by isotope ratio mass spectrometry to determine the isotope ratios for carbon and nitrogen (delta-¹³C and delta-¹⁵N).

In a visual examination of the values of the isotope ratios, without knowing which patients had an eating disorder, Hatch correctly differentiated between patients and controls in 70% of the cases. While this is fairly good, the success rate is inadequate for a clinical test. So the team applied discriminant analysis to the data using a simple algorithm to process the data.

The algorithm calculated the sensitivity and specificity of the test, where sensitivity is defined as the probability that the test is positive from someone with the disease and specificity is the probability that the test is negative from someone who does not have the disease. The values for both were 80%, indicating a good, if not exceptional, ability for diagnosis.

Hatch points out that the algorithm they devised should only be applied to the North American population, since their diet tends to include more corn than that of people in other regions of the world. Corn imposes a different carbon isotope ratio in the body than crops such as wheat and rice. However, it should not be difficult to calculate algorithms for other societies.

Other inaccuracies might have been introduced by dietary variations, by the relatively small number of hairs sampled and by the 1-cm cut length. However, the researchers argue that the 80% accuracy figures show that the method is basically sound.

The method can already predict which patients have an eating disorder with reasonable certainty and, with further refinement, could lead to even better predictions and the ability to discriminate between anorexia and bulimia. It is patient friendly, requiring only a few hairs in a non-invasive sampling procedure and could develop into a valuable diagnostic tool for clinicians.

Related links:

• National Institute of Mental Health on eating disorders
• Department of Integrative Biology, Brigham Young University
• Rapid Commun. Mass Spectrom. 2006, 20, 3367: "An objective means of diagnosing anorexia nervosa and bulimia nervosa using ¹⁵N/¹⁴N and ¹³C/¹²C ratios in hair"

Article by Steve Down 

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.

 Kent Hatch