NASHVILLE – Advising patients in a comprehensive weight loss intervention to moderate their alcohol consumption did not change how much they drank over the long term. At the same time, abstinent patients kept off more weight over time than those who were classified as heavy drinkers, in a new analysis of data from a multicenter trial.
Abstinent individuals lost just 1.6% more of their body weight after 4 years than those who drank (P = .003), a figure with “uncertain clinical significance,” Ariana Chao, PhD, said at the meeting presented by the Obesity Society and the American Society for Metabolic and Bariatric Surgery. “The results should be taken in the context of the potential – though controversial – benefits of light to moderate alcohol consumption,” she added.
Alcohol contains 7.1 kcal/g, and “calories from alcohol usually add, rather than substitute, for food intake,” said Dr. Chao. Alcohol’s disinhibiting effects are thought to contribute to increased food intake and the making of less healthy food choices. However, existing research has shown inconsistent findings about the relationship between alcohol consumption and body weight, she said.
Reducing or completely cutting out alcoholic beverage consumption is common advice for those trying to lose weight, but whether this advice is followed, and whether it makes a difference over the long term, has been an open question, said Dr. Chao.
She and her collaborators at the University of Pennsylvania, Philadelphia, used data from, “a multicenter, randomized, clinical trial that compared an intensive lifestyle intervention (ILI) to a diabetes support and education (DSE) control group,” for 5,145 people with overweight or obesity and type 2 diabetes, explained Dr. Chao and her coinvestigators.
Dr. Chao and her colleagues looked at the effect that the lifestyle intervention had on alcohol consumption. Additionally, to see how drinkers and nondrinkers fared over the long term, they examined the interaction between alcohol consumption and weight loss at year 4, hypothesizing that individuals who received ILI would have a greater decrease in their alcohol consumption by year 4 than those who received DSE. The investigators had a second hypothesis that, among the ILI cohort, greater alcohol consumption would be associated with less weight loss over the 4 years studied.
To measure alcohol consumption, participants completed a questionnaire at baseline and annually thereafter. The questionnaire asked whether participants had consumed any alcoholic beverages in the past week, and how many drinks per week of wine, beer, or liquor per week were typical for those who did consume alcohol.
Respondents were grouped into four categories according to their baseline alcohol consumption: nondrinkers, light drinkers (fewer than 7 drinks weekly for men and 4 for women), moderate drinkers (7-14 drinks weekly for men and 4-7 for women), and heavy drinkers (more than 14 drinks weekly for men and 7 for women).
At baseline, 38% of participants reported being abstinent from alcohol, and about 54% reported being light drinkers. Moderate drinkers made up 6%, and 2% reported falling into the heavy drinking category. Females were more likely than males to be nondrinkers.
Heavy drinkers took in significantly more calories than nondrinkers at baseline (2,397 versus 1,907 kcal/day; P less than .001).
Individuals who had consistently been heavy drinkers throughout the study lost less weight than any other group, dropping just 2.4% of their body weight at year 4, compared with their baseline weight. Those who were abstinent from alcohol fared the best, losing 5.1% of their initial body weight (P = .04 for difference). “Heavy drinking is a risk factor for suboptimal long-term weight loss,” said Dr. Chao.
Even those who were consistent light drinkers lost a bit less than those who were abstinent, keeping off 4.2% of their baseline body weight at 4 years (P = .04).
Look AHEAD included individuals aged 45-76 years with type 2 diabetes mellitus and a body mass index of at least 25 kg/m2, or 27 kg/m2 for those on insulin. Excluded were those with hemoglobin A1c of at least 11%, blood pressure of at least 160/100 mm Hg, and triglycerides over 600 mg/dL. A total of 4,901 patients had complete data available in the public access data set and were included in the present analysis. Dr. Chao and her colleagues used statistical techniques to adjust for baseline differences among participants.
The three-part ILI in Look AHEAD began by encouraging a low-calorie diet of 1,200-1,500 kcal/day for those weighing under 250 pounds, and 1,500-1,800 kcal/day for those who were heavier at baseline. Advice was to consume a balanced diet with less than 30% fat, less than 10% saturated fat, and at least 15% protein.
Patients were advised to strive for 10,000 steps per day, with 175 minutes of moderate-intensity exercise each week. Exercise was unsupervised.
Behavioral modification techniques included goal-setting, stimulus control, self-monitoring, and ideas for problem solving and relapse prevention. The intervention used motivational interviewing techniques.
With regard to alcohol, the ILI group was given information about the number of calories in various alcoholic beverages and advised to reduce the amount of alcohol consumed, in order to reduce calories.
The DSE group participated in three group sessions annually, and received general information about nutrition, exercise, and general support.
A potentially important limitation of the study was that alcohol consumption was assessed by self-report and a request for annual recall of typical drinking habits. An audience member from the United Kingdom commented that she found the overall rate of reported alcohol consumption to be “shockingly low,” compared with what her patients report drinking in England. The average United States resident drinks 2.3 gallons of alcohol, or 494 standard drinks, annually, according to the National Institute on Alcohol Abuse and Alcoholism, said Dr. Chao.
The midlife age range of participants, their diabetes diagnosis, and the fact that depressive symptoms were overall low limits generalizability of the findings, said Dr. Chao, adding that psychosocial factors, other health conditions, and current or past alcohol use disorder could also cause some residual confounding of the data.
Dr. Chao has received research support from Shire Pharmaceuticals.
SOURCE: Chao A et al. Obesity Week 2018, .