Article Text
PDF
Statistics from Altmetric.com
Q In overweight or obese adults, are dietary/lifestyle, pharmacological, and surgical interventions effective for inducing long term weight loss?
METHODS
Data sources:
Medline (1966 to September 2003), HealthSTAR (1975 to September 2003), Cochrane Controlled Trials Register (1990 to September 2003), bibliographies of relevant articles, and researchers in the field.
Study selection and assessment:
randomised controlled trials (RCTs) or cohort studies that had an English abstract and investigated a prespecified dietary/lifestyle, pharmacological, or surgical weight loss method in overweight or obese adults (body mass index ⩾25 kg/m2) 18–65 years of age. Study exclusion criteria included enrolment of non-consecutive or <100 participants, use of an intervention that was not approved for clinical use, and follow up <2 years. (1 year of follow up was allowed for pharmacological studies because only 4 studies were identified with follow up ⩾2 y).
Outcomes:
weight loss efficacy defined by absolute weight loss and the proportion patients with ⩾5% weight loss.
MAIN RESULTS
Dietary/lifestyle therapy. 16 studies (n = 5698) met the selection criteria. Mean patient age ranged from 40 to 59 years, and mean weight at baseline ranged from 78 to 116 kg. Dietary/lifestyle interventions comprised very low calorie diets (<1100 kcal/d) or low calorie diets (1200–1500 kcal/d), sometimes combined with behaviour counselling/lifestyle modification. In 13 trials, weight loss was assessed during a 3–18 month induction phase followed up by a 12–36 month weight loss maintenance phase after the intervention. In 3 trials, weight loss was assessed only during the 24–48 month induction period. Meta-analysis was not done because of study heterogeneity. Weight loss was usually <5 kg after 2–3 years and 1.8–10.0 kg after 4–7 years.
Pharmacological interventions. 19 RCTs (n = 9953) met the selection criteria. Mean patient age ranged from 40 to 58 years, and mean weight at baseline ranged from 87 to 102 kg. All RCTs compared a dietary/lifestyle regimen plus a drug (orlistat or sibutramine) with a dietary/lifestyle regimen alone. Meta-analysis was done using a random effects model. More patients in the pharmacological (orlistat or sibutramine) intervention group than in the placebo group attained weight loss ⩾5% at 1 or 2 years (13 RCTs; odds ratio 2.94, 95% CI 2.47 to 3.50).
Surgical therapy. 9 studies (3 RCTs and 6 cohort studies) (n = 3622) met the selection criteria. Mean patient age ranged from 34 to 49 years, and mean weight at baseline ranged from 110 to 142 kg. Surgical interventions comprised restrictive procedures that induced early feelings of fullness or diversionary procedures that decreased caloric absorption. Meta-analysis was not done because of study heterogeneity. Weight loss was 28–76 kg after 2 years, 17–73 kg after 3 years, and 20 kg after 8 years.
CONCLUSIONS
In overweight or obese adults, dietary intervention plus pharmacological treatment (orlistat or sibutramine) is more effective than dietary intervention alone for inducing long term weight loss. Benefit of weight loss after surgical interventions must be assessed against risks.
Commentary
Obesity and concomitant cardiovascular risks, type 2 diabetes, hypertension, and hyperlipidaemia have reached epidemic proportions.1 Clinicians need evidence and guidelines on successful strategies to counsel patients on long term weight loss. The review by Douketis et al showed very small weight loss with diet and lifestyle treatment and modest to large long term weight loss with pharmacological and surgical treatment, accompanied by reductions in cardiovascular risk factors in high risk groups only. Clinical applicability is limited by the use of low or very low calorie restricted diets, including pre-packaged foods; lack of detail regarding behavioural counselling; and either lack of measurement of exercise levels or lack of restriction of exercise to the treatment group. It was difficult to assess the unique effects of caloric restriction, exercise, or the elements of behavioural counselling as they were bundled together. Of course, such combinations are often used by dieters within or outside of formal programmes.
The review by Curioni and Lourenco concluded that dietary intervention plus exercise results in statistically significant and clinically meaningful initial weight loss compared with dietary intervention alone. However, p values and confidence intervals were not within conventional values of statistical significance.
These 2 reviews show that primary research on weight loss continues to be hampered by methodological issues of high drop out rates, where the results are based on completers only or last observation carried forward and overestimate treatment effect; lack of attention to high risk groups; lack of appropriate comparison groups; and need for longer follow up.
Although both reviews described the strategies used to effect weight change, neither considered the mechanisms of the interventions reviewed. Theory based interventions have been shown to contribute significantly to success in weight management and physical activity.2
We are left with patients needing to reduce weight to reduce health risks but little evidence to support many specific interventions. Clinicians who suggest weight loss or weight maintenance strategies need to take into account individual patient preferences and situations and to help them plan methods of dietary practices and food preparation to encourage healthy eating, along with planned physical activity. Although the review by Douketis et al showed the effectiveness of pharmacological and surgical interventions, these approaches should be recommended only after careful consideration of harms and benefits, particularly in light of the limited information on cardiovascular risk factor reduction and cardiac event follow up.
Footnotes
For correspondence: Dr J D Douketis, Department of Medicine, McMaster University, Hamilton, Canada. jdouket{at}mcmaster.ca
Source of funding: not stated.