Diabetes Risk Following Mental Disorders

This is the third in a series of posts looking at a recent research study of mental disorders and risk for later medical illnesses.

In this graph I has summarized data from the manuscript below on mental disorders associated with later risk of diabetes mellitus.

Two eating disorders, bulimia nervosa and binge eating disorder were both associated with a tripling of the rate for a later diagnosis of diabetes mellitus. Both of these eating disorders are linked to overweight and obesity possibly explaining the later association with diabetes.

Mood disorders and anxiety disorders panic disorder and agoraphobia also were associated with a later diagnosis of diabetes mellitus. One concern with this finding is the potential for drugs used for mood and anxiety disorders to increase weight.

I will finish out this series in the next post on mental disorders linked to later chronic paid disorder.

Follow the author on Twitter WRY999

Scott KM, Lim C, Al-Hamzawi A, Alonso J, Bruffaerts R, Caldas-de-Almeida JM, Florescu S, de Girolamo G, Hu C, de Jonge P, Kawakami N, Medina-Mora ME, Moskalewicz J, Navarro-Mateu F, O'Neill S, Piazza M, Posada-Villa J, Torres Y, & Kessler RC (2015). Association of Mental Disorders With Subsequent Chronic Physical Conditions: World Mental Health Surveys From 17 Countries. JAMA psychiatry, 1-9 PMID: 26719969

Stroke Risk Following Mental Disorders

An important recent manuscript published in JAMA Psychiatry looked at medical illnesses rates following diagnosis of a brain (mental) disorder.

This very large international study examined over 47,000 subjects followed between 2001 and 2011.

Baseline psychiatric assessment was completed using the Composite International Diagnostic Interview. Physical illness was assessed using a self-report of physician's diagnosis.

The manuscript is not free but I was able to obtain a professional courtesy copy of the manuscript from the lead author. 

The study found that a diagnosis of a mental disorder increased rates for a variety of later physical conditions.

I have put together a graph of the mental disorders associated with higher rates of later stroke (above).

Bulimia nervosa showed a three fold increase in later stroke. I am not aware of this association previously being noted. Alcohol and drug use disorders are known to increase stroke risk and this was found in the study lending face validity support.

Additional mental disorders associated with later stroke were bipolar disorder and panic/agoraphobia. 

Bipolar disorder has been linked to higher risk of hypertension and smoking.

The implications of these findings include need for enhanced medical illness screening and risk factor reduction in those with a mental disorder diagnosis. 

I will look at some the other individual medical conditions in future posts.

Follow the author on Twitter WRY999

Scott KM, Lim C, Al-Hamzawi A, Alonso J, Bruffaerts R, Caldas-de-Almeida JM, Florescu S, de Girolamo G, Hu C, de Jonge P, Kawakami N, Medina-Mora ME, Moskalewicz J, Navarro-Mateu F, O'Neill S, Piazza M, Posada-Villa J, Torres Y, & Kessler RC (2015). Association of Mental Disorders With Subsequent Chronic Physical Conditions: World Mental Health Surveys From 17 Countries. JAMA psychiatry, 1-9 PMID: 26719969

Parental Education As Risk Factor For Eating Disorders

Genetic and environmental risk factors contribute to the risk for anorexia nervosa and other eating disorders.

Known risk factors for anorexia nervosa include female gender, young age, family member with anorexia nervosa, weight loss, and participation in weight sensitive sports or activities, i.e. gymnastics, dancing.

There has also been evidence that anorexia nervosa is more common in higher socioeconomic classes. This finding has made it one of the few brain disorders more common with this category.

A recent study using the Swedish medical registry sheds some light on increased risk related to socioeconomic status in anorexia and other eating disorders.

The Swedish medical registry used in this study contained over 2,000,000 records and from this group the research team were able to identify 15,474 girls and women (1.5%) with an inpatient eating disorder diagnosis along with 1051 boys and men (0.1%)

The key findings from the study included:

• Rates of eating disorder diagnosis increased with those born in 1982 and later
• Female eating disorder diagnosis rates were predicted by greater educational level in the father, mother and maternal grandparents
• Rates of anorexia nervosa and bulimia nervosa were nearly doubled in children of parents with post-graduate education compared to those without secondary education
• Parental social class and income were not linked to increased eating disorder risk

The authors note their study results finding links between eating disorders and socioeconomic status may be due to the higher education risk factor.

The authors also note the mechanism for this association is unclear. The speculate that high educational expectations for children may overlap with familial perfectionism, a trait known to increase eating disorder risk. Additionally, they note there is the potential for their study to be confounded by genetic factors.

This is an important study that will prompt further study of the link between parental education and eating disorder risk.

Readers with more interest in this topic can access the free full-text manuscript by clicking on the PMID link below.

Follow the author on Twitter @WRY999 

Photo of black bellied whistling duck is from the author's files.

Goodman A, Heshmati A, & Koupil I (2014). Family history of education predicts eating disorders across multiple generations among 2 million Swedish males and females. PloS one, 9 (8) PMID: 25162402

Brain Response to Food in Anorexia vs Bulimia

The eating disorders anorexia nervosa and bulimia nervosa share some common features.  Excessive fear of being fat is a core feature of both disorders.  However, key clinical differences between the disorders exist.

Most patients with bulimia nervosa are in the normal to overweight category.  Anorexia nervosa by definition requires being of low weight, often to such extreme levels to pose a danger of death due to the effects of starvation.

Now we have a study using fMRI that examines differences between these two eating disorder in how the brain responds to food stimuli.  Samantha Brooks of Uppsala University in Sweden along with colleagues from England and Germany have recently published this study in Plos One.  Here are key elements of the design of their study:

• Subjects: 8 women with bulimia nervosa, 18 women with anorexia nervosa (11 with restricting subtype and 7 with binge purging subtype) and 24 healthy control women
• Stimuli: Color photographs of a variety of foods along with control photographs of non-food items
• Analysis: Comparison of brain BOLD effect activation between food photos compared to control photos and comparisons between eating disorder diagnostic groups

The researchers found in healthy controls food stimuli (in contrast to non-food stimuli) activated the following brain regions:

• right insular cortex
• right middle temporal gyrus
• left cerebellum
• left caudate
• right somatosensory area (binge purge anorexia subtype only)

Brain activation to food stimuli differed in those with anorexia as well as those with bulimia compared to health controls.  In the anorexia group food stimuli activated the following brain regions:

• right precuneus
• right dorsolateral prefrontal cortex
• left cerebellum

Activation to food stimuli in those with bulimia nervosa was found in the following brain regions:

• right insular cortex
• left precentral gyrus
• left dorsolateral prefrontal cortex
• right visual cortex

The image below from the manuscript compares the bulimic group with the anorexia subjects.  Areas activated in bulimia nervosa to a greater extent than in those with anorexia nervosa included the right caudate, right superior temporal gyrus and right insula.

The authors note that activation of the left dorsolateral prefrontal cortex with food stimuli in both eating disorders support activation of a "cognitive control network".  This suggests those with eating disorders have to "think about eating food" in addition to processing an appetitive response.

The authors note that enhanced activation of the caudate and precentral gyrus in bulimia nervosa "suggests and= increased appetitive response to food images".  This may reflect food craving as these areas have been noted to become activated in healthy controls after fasting.

Clinically, food and carbohydrate craving are more often features of bulimia.  Many anorexia nervosa patients report no or limited feelings of hunger.

Look for more functional imaging studies in those with eating disorders.  The studies suggest interventions that normalize brain activation responses to food stimuli may be an pathway to reduction of dysfunctional eating behaviors.

Photo of filtered sunrise in Juno Beach, Florida from the author's private collection.

fMRI image from cited manuscript Brooks et.al used under terms of the Creative Commons Attribution License.

Brooks, S., O′Daly, O., Uher, R., Friederich, H., Giampietro, V., Brammer, M., Williams, S., Schiöth, H., Treasure, J., & Campbell, I. (2011). Differential Neural Responses to Food Images in Women with Bulimia versus Anorexia Nervosa PLoS ONE, 6 (7) DOI: 10.1371/journal.pone.0022259

Cognitive Biomarkers in Eating Disorders

I previously posted on a review of the neuropsychological domain of set-shifting as a possible biomarker for eating disorders.  In that post, set shifting as measured by the Wisconsin Card Sort tests was impaired in those with an anorexia nervosa and continued impaired despite weight restoration.  Additionally, sisters of those with an eating disorder, also showed some impairment on this measure even when they had no eating disorder problem.  This supports a potential role for set shifting as a potential biomarker for eating disorders.

Now an additional study of neuropsychological performance in women with eating disorders and their healthy sisters has been published in the journal Comprehensive Psychiatry.  In this study, the authors examine two computerized tasks tapping the symmetry of performance of the brain’s left and right hemisphere.   

The test of right hemisphere function involved a two-bar visual bar graph stimulus.   Participants were asked to identify the presentation as odd or even dependent on the location of an indicator arrow.  In the test of left hemispheric function task, four letter strings were presented and participants were asked to identify the words as real work or a pseudoword.

The subjects with eating disorders included young women with anorexia nervosa, restricting subtype, anorexia nervosa, binge-purge subtype and bulimia nervosa.  The neuropsychological tests were assessed for accuracy and speed of response.  The key finding from the study were:

• Patients with eating disorders showed higher error rates and slower response times on tests of both the right and left hemisphere
• Anorexia nervosa, restricting subtype tended to be associated with the greatest level of impaired functioning
• Sisters of the eating disorder subjects also showed impairment despite scores similar to controls on measures of depression, impulse control, obsessional thinking and eating disorder symptoms

The authors note that there results support “general individual differences in cognitive processes that may run in families, irrespective of the eating disorder condition of the family member”.    The authors also note that their findings support the continuum model of eating disorders proposed by Michael Strober.  This model proposes that anorexia nervosa, restricting subtype is the most severe category of the eating disorders with bulimia nervosa being the least severe (on a general basis, it is possible for some individuals with bulimia nervosa to have a more severe eating disorder than some individuals with anorexia nervosa, restricting subtype).   The data from this study indicate the anorexia nervosa restricting subtype demonstrated the most severe neurocognitive impairment with bulimia nervosa the least impairment.

Look for more studies examining neurocognitive function in those with eating disorders.  Strategies that combine neuropsychological assessment with fMRI and genetic data may be the most powerful strategy for research in these disorders.  


Photo of rescued sea turtle from the Loggerhead Marine Center in Juno Beach, Florida courtesy of Yates photography

Rozenstein MH, Latzer Y, Stein D, & Eviatar Z (2011). Neuropsychological psychopathology measures in women with eating disorders, their healthy sisters, and nonrelated healthy controls. Comprehensive psychiatry PMID: 21397219

Why Is Anorexia Nervosa Neglected for Drug Development?

Atypical Antipsychotic Olanzapine

Eating disorders have been a neglected area for high-quality psychopharmacologic research.  There are probably several reasons for this.  The classic eating disorder anorexia nervosa is relatively rare and identifying 500 to 1000 subjects for a clinical trial would likely be a significant (but not impossible) research challenge.   There are currently no FDA approved drugs indicated for the treatment of anorexia nervosa.

One drug in the U.S. has FDA approval for bulimia nervosa, the antidepressant fluoxetine.  But this approval occurred in the late 1980’s meaning we are approaching twenty-five years without a new drug approval for bulimia nervosa.  Pharmaceutical company interest in bulimia may be tempered somewhat by the experience of a trial using the drug bupropion.  Bupropion appeared effective in reducing binge eating in bulimia nervosa but a the clinical trial participants on bupropion had an increased risk of seizures during the trial.  The electrolyte disruption seen in bulimia (from bingeing and purging behaviors) may contribute to an increased risk of seizure—particularly for drugs with a known risk of reducing seizure thresholds.

So is anorexia nervosa neglected because there just are no potential candidates?  There is some evidence of the potential for the atypical antipsychotic medications in anorexia nervosa.  This evidence comes from outside the FDA approval process and typically involves small numbers of subjects.  McKnight and Park from the Department of Psychiatry at the University of Oxford recently summarized the research knowledge base in this area.

Why should atypical antipsychotics be considered for an eating disorder?  McKnight and Park propose three reasons:

• Atypicals reduce agitation and anxiety—common hindrances in refeeding underweight patients with anorexia nervosa
• Atypical often cause weight gain
• Some features of anorexia nervosa resemble psychosis—persistent belief of being overweight despite starvation and weight loss

Only three double-blind trials have been published according to this review.  All involve olanzapine versus placebo.  Two of these studies found and increase in BMI (weight) with the drug.  All of the these studies found some favorable effect on anxiety, depression or eating disorder psychological symptoms.

One single blind study compared the atypical antipsychotic drug amisulpride to fluoxetine and placebo.  Amisulpride produced a significant weight gain compared to the other two treatment arms. 

Four open label non-blinded studies suggest the potential for quetiapine to have a favorable effect on weight gain and/or reduction in eating disorders psychological variables.

Evidence for the use of other atypical agents (i.e. risperidone, aripiprazole) is limited to a few case reports, but these have generally been favorable.  One case report noted development of hyperglycemia in adult with anorexia receiving 15 mg of olanzapine daily.

I searched the ClinicalTrials.gov website for anorexia nervosa and found only two active clinical trials recruiting subjects:

• Olanzapine versus placebo for outpatients with anorexia nervosa (an NIMH-sponsored study) conducted by Cornell, University of Pittsburgh, Johns Hopkins and the University of Toronto with a targeted enrollment of 160
• Aripiprazole versus placebo—a phase III study being conducted at the University of Barcelona in Spain with a targeted enrollment of 60

So the history of neglecting anorexia nervosa for drug development seems to be continuing.  What will it take for more research attention to this important disorder so that clinicians will have more to offer the patients and their families?

Image of Chemical Structure of Olanzapine provided in the public domain by author Ben Mills.

McKnight RF, & Park RJ (2010). Atypical antipsychotics and anorexia nervosa: a review. European eating disorders review : the journal of the Eating Disorders Association, 18 (1), 10-21 PMID: 20054875