Weight Loss with Dietary Counseling Fizzles Over Time

BOSTON, July 2 -- Dietary counseling produces modest weight loss, but the effect disappears within about five years, a meta-analysis showed.
On average, weight loss programs that use group or individual counseling, or both, helped participants drop 6% of their body weight after one year, found Michael L. Dansinger, M.D., of Tufts-New England Medical Center here, and colleagues.
At three years, however, half of the weight lost was regained, and at about 5.5 years participants were back to baseline, they reported in the July 3 issue of Annals of Internal Medicine.
Nonetheless, "even small, relatively short-term weight loss may have an important clinical effect," they wrote.
"As many of the included studies show," they added, "dietary and lifestyle changes resulting in modest weight loss consistently reduce such cardiovascular risk factors as hyperglycemia, hypercholesterolemia, and hypertension."
Although weight loss programs with a counseling component have long been known to produce modest, transient effects, the magnitude and persistence of the weight loss across clinical trials was less clear.
So the researchers gathered the relevant studies from a prior systematic review covering 1980 to 1997 and searched the literature for randomized trials published from 1997 through 2006.
The 46 trials included reported original data comparing dietary counseling interventions with control, either usual care or minimal counseling (general verbal or written advice). Participants had to be adults with a body mass index of 25 kg/m2 or greater at baseline.
Overall, about 6,386 people underwent dietary counseling and 5,467 received usual care or minimal counseling. Mean age ranged from 27 to 68, and mean BMI ranged from 25 to 40 kg/m2. Participants were healthy in 12 trials, but had hypertension in another 12 trials, type 2 diabetes in 10, impaired glucose tolerance in five, heart disease in three, dyslipidemia in two, and other conditions in six.
Nearly all of the trials promoted exercise, but interventions varied. Seventeen targeted calorie intake, 25 restricted fat intake, and others individualized dietary recommendations to achieve a specific weight loss goal. Group meetings were used in 18 trials, individual meeting in 13 trials, a combination in 11, and three trials used the Internet.
Dr. Dansinger and colleagues found no studies of low-carb diets that included a control group, so none of were included in their review.
Active intervention ranged from 2.5 to 48 months. Although about two-thirds of the trials were of at least fair quality, 28% were poor quality, the researchers said.
They found the expected pattern of weight loss during the active phase of trials and weight regain during the maintenance phase.
On average during active intervention, BMI dropped by 0.12 units per month during months three to six and 0.04 units per month from months three to 12. Meta-regression suggested an overall loss of 0.08 units per month to one year (P=0.007).
The net weight loss effect was about two BMI units at one year (6% of baseline weight, or about 11 lbs).
Diet and exercise showed no consistent benefit over diet alone, although the difference significantly favored exercise with diet at 12 months. The rate of weight loss with diet and exercise was 0.23 BMI units per month from months three to 12 averaged across three trials compared with essentially no change in the diet groups alone (P=0.009).
Studies of people with diabetes showed that dietary counseling intervention was half as effective as in studies of nondiabetics (P<0.001).
During maintenance, the average weight regain increased with time: 0.01 BMI units per month from months six to 12, 0.02 to 0.03 units per month from 12 to 18 months, and 0.04 units per month from 24 to 30 months.
About half the initial weight loss was regained at three years, Dr. Dansinger and colleagues noted.
Meta-regression suggested, "patients would return to their baseline weights after approximately 5.5 years," they wrote.
Independent predictors of greater weight loss were:
A lower daily caloric intake target (P=0.009).
More frequent support meetings (P<0.001).
Absence of diabetes (P<0.001).
Notably, whether the intervention included exercise was not a significant factor (P=0.50).
Predictors of slower weight regain were lower calorie targets (P=0.011) and the absence of diabetes (P=0.012), but not type of intervention (P=0.62).
Only a third of the studies reported recommended daily caloric intake, however, and the predictive findings can only be considered hypothesis generating because of the retrospective nature of the meta-analysis and lack of patient-level data.
The study was also limited by differences between interventions studied, "although such heterogeneity may enhance the generalizability of our findings," the investigators wrote.
Long-term trials are needed to confirm the findings, the researchers concluded.
The study was funded by the Agency for Healthcare Research and Quality. Dr. Dansinger reported support by a grant from the National Institutes of Health. The researchers disclosed no potential financial conflicts of interest. Primary source: Annals of Internal MedicineSource reference: Dansinger ML, et al "Meta-analysis: The effect of dietary counseling for weight loss" Ann Int Med 2007;147:41-50.