Blood Test May Distinguish Anorexia From 'Constitutional Thinness'

Liam Davenport

March 07, 2018

NICE, France – A simple, inexpensive blood test can help clinicians to distinguish between anorexia nervosa (AN) and constitutional thinness (CT) in severely underweight young women, new research suggests.

Since the introduction of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), the diagnostic criteria for AN have relied on psychological factors, making it harder to distinguish AN from CT.

However, a new study of more than 150 young women showed that blood levels of a thyroid hormone can be used to distinguish the two conditions with high sensitivity and specificity.

In cases in which the exclusive use of psychological criteria is not sufficient, the diagnosis of AN "should be completed by at least free T3 determination, which is a cheap and accessible" test, note Natacha Germain, MD, PhD, Endocrinology and Eating Disorders Reference Center, Centre Hospitalier Universitaire de Saint-Étienne, France, and colleagues.

The findings were presented here at the European Psychiatric Association (EPA) 2018 Congress.

Diagnostic Confusion

Germain told meeting attendees that using criteria from the DSM-IV to differentiate between AN and CT in a young woman whose body mass index was less than 18.5 kg/m2 was "quite easy."

That's because the criteria for AN included the presence of an eating disorder, resistance to food intake, and amenorrhea. In contrast, CT was "a state with no eating disorder, no resistance to body weight gain, no undernutrition, and no amenorrhea."

However, the amenorrhea criteria were removed with the introduction of the DSM-5, making it easier to confuse AN and CT. This confusion "is highly detrimental, especially for CT, because of the social stigmatization, health insurance penalty, and the inefficient, aggressive therapy applied in this population," said Germain.

To identify distinguishing markers that could be used to easily tell the difference between the two conditions, the investigators studied 40 patients with AN, 56 patients with CT, and 54 women with normal weight, who served as controls. DSM-IV criteria were used to define AN and CT.

They evaluated hormonal and nutritional parameters, and administered a series of psychiatric questionnaires to find robust, distinguishing markers that would fit with criteria from the DSM-5.

Performing receiver operating characteristic curve analysis, they found that only the restrained eating subscale of the Dutch Eating Behavior Questionnaire could be used to distinguish between AN and CT. It had a specificity of 94% and a sensitivity of 95% at a cutoff of 0.74.

"Unfortunately, it's only one dimension in a complex questionnaire, and it's sometimes biased by denial," said Germain. She added that in clinical practice, the availability of the questionnaire is very low.

Objective Biological Markers

Estradiol was, as expected, the best marker for distinguishing between the two populations. It had a sensitivity and specificity of 87% and 80%, respectively, at a cutoff of 24. However, using estradiol as a biomarker in this context is incompatible with the DSM-5, Germain noted.

Leptin was also very good at distinguishing between the conditions, with a specificity of 80% and a sensitivity of 84% at a cutoff of 2.58. "Unfortunately, leptin testing has a low availability and high cost, so it's not very easy to use it as a marker," she said.

The researchers found that free triiodothyronine (T3) levels could be used to distinguish between the patient groups It had a sensitivity of 80% and a specificity of 91% at a cut-off of 3.3 pmol/L (P < .0001).

Moreover, the test for free T3 levels is cheap, and the results are simple to interpret. It is also widely available.

Germain concluded that for a young, thin woman, a free T3 level of less than 3.3 pmol/L indicates AN with undernutrition.

In the postpresentation discussion, she was asked by an audience member whether it's possible to make a diagnosis of AN solely on the basis of the patient's history.

"Yes, it could be possible," she said. "It would be very simple to have the weight history, but most of the time, we don't have it. So sometimes it's not that easy, and we need biological, objective markers."

"Subjective Category"

After the session, Medscape Medical News spoke to session chair Sébastien Guillaume, MD, from the University of Montpellier, France, who was not involved in the research.

Asked about the impact of the change in diagnostic criteria for AN in the current DSM-5, he said that "it was not good, but it was better than doing nothing, compared with DSM-IV.

"At least we have some patients included who were as serious as we had before but were excluded from the diagnostic category," he added. "We didn't know what to do with them, and basically they have the same disease, so I think the good thing is there is [now] less 'eating disorder not otherwise specified.' "

Guillaume noted that "the category remains very subjective, but that's like all the other categories of the DSM. Maybe one day we'll have some biomarkers to better characterize it."

The researchers and Dr Guillaume have disclosed no relevant financial relationships.

European Psychiatric Association (EPA) 2018 Congress. Abstract OR0174, presented March 6, 2018.

For more Medscape Psychiatry news, join us on Facebook and Twitter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.