If older patients are suddenly craving sweets, gaining weight, and developing swallowing difficulties, consider a diagnosis of frontotemporal dementia (FTD), a new study suggests.

Results show that patients with certain types of FTD eat significantly more carbohydrates and sugar than healthy controls or those with Alzheimer’s disease (AD), and that these changes don’t appear to be explained by being hungrier.

Patients presenting with such eating behaviors should raise a red flag, study author Olivier Piguet, PhD, associate professor, University of New South Wales, and Principal Research Fellow, Neuroscience Research, Australia, told Medscape Medical News.

“Someone in their 50s or early 60s showing these changes in eating preferences and the amount of food that they eat would certainly indicate that something might be going on in their brain that needs exploring further.”

The study, published in the December issue ofJAMA Neurology, is the first to quantify abnormal eating behaviors in patients with FTD, said Dr Piguet.

Eating Disturbance

“Changes in eating behavior are part of the criteria for the diagnosis of behavior variant FTD, but no one has really looked at exactly what that means,” said Dr Piguet. “This study is really the first one to try to measure what it means when we say these patients have an eating disturbance.”

The analysis included 75 patients with dementia: 21 with personality or behavioral disturbance (behavioral variant FTD [bvFTD]), 26 with language disturbances (semantic dementia or SD), and 28 with AD, as well as 18 age- and education-matched healthy controls.

Caregivers completed the Appetite and Eating Habits Questionnaire (APEHQ), which includes 34 questions examining changes in eating behaviors with regard to swallowing, appetite, eating habits, food preferences, and other oral behaviors (eg, eating objects such as cigarette butts). For each question, researchers calculated a composite score that included frequency and severity and derived an overall score for each domain.

Investigators found the bvFTD group had significantly higher scores than the AD group for all 5 APEHQ domains: swallowing (P = .003), appetite change (P = .007), eating habits (P = .001), food preferences (P = .001), and other oral behaviors (P = .009).

Caregivers also completed the Cambridge Behavioral Inventory, which includes four questions related to eating behaviors: sweet preference, eating the same foods, changes in appetite, and table manners.

The table manners item was included because of anecdotal evidence that patients with behavioral disturbances lose this etiquette. “Caregivers will report that ‘my husband is stealing food from someone else’s plate’ or piling up food on their plate,” said Dr Piguet.

There were significantly greater changes related to sweet preference (P < .001), eating the same foods (P = .001), and table manners (P = .007) in the bvFTD group compared with the AD group.

Some of the findings were unexpected.

“The finding that the behavior variant patients changed their preferences in terms of the foods they like and their tendency to focus more on sweet foods confirms something we knew already, although we were able to quantify that,” commented Dr Piguet. “But we also found that this tendency to prefer sweet foods was also present in the group with semantic dementia, which was surprising.”

Although patients with semantic dementia increased their carbohydrate intake to some degree, the increase among those with bvFTD was “more prominent,” said Dr Piguet, “again confirming anecdotal evidence that we were able to measure with the questionnaire.”

Caregivers also measured patients’ level of hunger and satiety using a visual analogue scale before and after meals during a 24-hour period (with higher scores indicating more hunger). The bvFTD group had a significantly higher overall hunger-satiety index score than the semantic dementia (P= .02) and AD (P = .03) groups, but not the control group (P = .38).

These results suggest that “these patients don’t eat because they feel full,” but simply because “the food is there and they eat it,” said Dr Piguet.

Weight Concerns

The bvFTD and SD groups had significantly greater waist circumference compared with the control group, and the bvFTD group had significantly greater waist circumference compared with those with AD.

These patients, said Dr. Piguet, are “getting into the danger zone” with body mass indexes “hovering around the 30 mark, which raises concerns about their general health, their cardiovascular health and risks for related illnesses such as diabetes.”

An earlier paper by the same research group found atrophy in the hypothalamus — the area of the brain that plays a central role in eating regulation — in patients with FTD, said Dr Piguet. “The brain is receiving incorrect messages from the periphery in terms of hunger and satiety and is then responding to these messages in an incorrect manner.”

The changes, he added, can lead to disordered behavior when it comes to eating. “These patients tend to want to eat the food that they have in front of them; they have difficulty in inhibiting or stopping their eating.”

In FTD, atrophy predominantly affects the frontal lobes and temporal brain areas; in AD, in contrast, other brain regions are affected. “You rarely see these behavior changes in patients with AD,” noted Dr Piguet.

Reached for comment on these findings, Ronald Petersen, MD, PhD, director, Mayo Alzheimer’s Disease Research Center, Mayo Clinic, Rochester, Minnesota, said the study results illustrate disinhibition on the part of patients with bvFTD.

“Foods that are sweet are very attractive to most of us, and people with bvFTD lack the ability to deny themselves the pleasure,” said Dr Petersen. “They can’t reason that this type of excessive ingestion of sweets will lead to weight gain and other health consequences.”

These patients, he added, lack the ability to foresee the consequences and act impulsively. “This is due to frontal lobe dysfunction since that part of the brain is involved in our ability to judge the consequences of behavior.”

This work was supported by a National Health and Medical Research Council (NHMRC) project grant, the Australian Research Council (ARC) Centre of Excellence in Cognition and its Disorders, a Royal Australasian College of Physicians scholarship and MND Australia scholarship, an ARC Discovery Early Career Research Award, an ARC Federation Fellowship, and an NHMRC of Australia Career Development Fellowship. The authors have disclosed no relevant financial relationships.

JAMA Neurol. 2014;71:1540-1546. Abstract