I was very excited to see a recent paper (Sanchez et al. 2013) published in the British Medical Journal that reports enhanced weight loss among obese women who consumed a probiotic+prebiotic intervention (also known as a “synbiotic”), compared to those consuming a placebo.
Since Ruth Ley first published a study (Ley et al. 2006) showing that certain human fecal microbes were associated with obesity, and that weight loss was accompanied by a shift to “lean” fecal colonization, there has been much hope that changing colonization patterns could help with weight management. Indeed, colonizing microbiota may function almost like an organ that contributes to energy utilization and storage. Since probiotics and prebiotics are useful approaches to changing the colonizing microbiota, perhaps they can help with weight management.
The following is my attempt to familiarize you with the Sanchez study, providing insight into its strengths and weaknesses, and share with you my conclusion that – although a respectable first study – we require additional evidence to understand what role a probiotic, prebiotic and/or synbiotic can have on facilitating weight loss. Details of the study are shown in the box at the end of this post.
To consider this study carefully, understanding some jargon is helpful:
- Primary endpoint: The main result that is measured in a study. Primary endpoints should be clearly stated during the study design phase. The primary endpoint is the measure upon which the sample size calculation is based.
- Sample size calculation: When conducting a study, a researcher must determine how many subjects are needed in each arm of a study to enable statistical judgments about differences between those arms that are accurate and reliable. The appropriate sample size is one that would be expected to enable detection of a difference between the test and placebo groups. These calculations are based on effect sizes observed in other (sometimes pilot) studies.
- Intention-to-treat analysis: Analysis of study results that includes every subject who is randomized in the study, regardless if they complete the study. This analysis typically renders a conservative estimate of treatment effect.
- Per protocol population: A subset of the ITT population who completed the study without any major protocol violations. Study dropouts are not included.
- Null result: A result that leads to the conclusion that the study does not provide evidence for a treatment effect, but also cannot prove that there is no effect. It is generally impossible to prove a negative.
In this study, no difference was observed in the primary endpoints of body weight, body composition or waist circumference for subjects in the synbiotic group compared to the placebo group.
Although the study was not designed to compare results in men and women separately from the total group, after the study was completed researchers observed that women in the synbiotic group lost more weight at both 12- and 24-week measures (9.7 and 11.5 pounds, respectively) than women on the placebo (5.7 and 5.5 pounds, respectively). Women taking the synbiotic also lost 3.3% fat mass at the 24-week measure, which was a statistically significant improvement compared to the placebo. This is a very interesting observation, and why this happened will surely lead to forthcoming hypotheses from researchers. Perhaps women are more responsive to the synbiotic. Or perhaps this is an accident of the study.
Women dropped out of the study at a higher rate than the men (30% overall, 28% of placebo and 31% of synbiotic group for women; 18.8% overall 16.7% of placebo group, 20.8% of synbiotic group for men). To the extent that the reason for dropout was tied to the study endpoints (perhaps women who did not lose weight were more likely to drop out), this could bias the study findings. Since the subgroup analysis was only presented for the subjects who finished the study (per protocol) and not for all subjects entered into the study (intention-to-treat), it leaves open the question that this biased the results in favor of a positive outcome. As succinctly stated by Bell, et al. (2012) “Dropout in longitudinal randomised controlled trials is common and a potential source of bias….”
The authors conclude that the synbiotic “can accentuate body-weight loss in women submitted to energy restriction.” In my opinion, this may be an overstatement of the study findings, considering that the study was not designed for this endpoint and the study exhibited a high dropout rate. The observation that women were more responsive than men to the synbiotic instead leads to a hypothesis for a follow-up study, designed for this endpoint.
One interesting aspect of this study is the rationale for the prebiotic component of the intervention. The product contained 300mg of a prebiotic mixture. Since daily doses of in the range of at least 4-5 grams of prebiotic are typically needed for a detectable impact on indigenous microbes, this dose seems too low to be effective. However, the prebiotic was added to improve the survival of the probiotic through the GI tract. Evidence for this from an in vitro analysis was provided in the study.
I eagerly await the next study that tests the effect of probiotics on weight loss. The list of clinical trials (on clinicaltrilas.gov) shows multiple studies at different stages of completion that examine probiotics and some aspect of weight loss. (ClinicalTrials.gov Identifiers: NCT00880958, NCT01106924, NCT01433120, NCT01445704, NCT01870544, NCT01927107, NCT01978691, NCT02037295.) Maybe I won’t have to wait too long.
- Study design: Randomized, double-blind, placebo-controlled trial
- Primary endpoints: pre and post measures of body weight, body composition and waist circumference. (Note that these are not clearly stated in the study publication but are clearly stated in the trial registration)
- Subjects: Healthy (i.e., without associated co-morbidities, such as hyperlipidemia, diabetes or hypertension) obese men and women. (Note that ‘obese’ is defined as a body mass index of 30 kg/m2 or greater. In this study, subjects were recruited with BMIs between 29 and 41.)
- Number of subjects: A total of 125 subjects were randomized into the study (63 placebo, 62 intervention). 93 subjects completed the trial (48 placebo; 45 intervention).
- Dropout rate: 25.6% overall, no difference in dropout between placebo and intervention groups. Only 54 of 77 (29.9%) women completed the study; 40 of 48 (16.7%) men completed the study; Overall, 93 of 125 (25.6%) completed the study.
- Intervention: 2 capsules per day (before breakfast and before dinner), each capsule containing 1.6×108 cfu L. rhamnosus CGMCC1.3724 + 300 mg of a mix of oligofructose and inulin (70:30, v/v)]
- Duration: 24 total weeks, comprising 2-12 week phases: energy restriction for 12 weeks followed by weight maintenance for 12 weeks. (A 2-week washout preceded the 24 week study.)
- Supervision provided during study: dietary records and physical activity records collected for 3-day periods during each study phase. A personalized diet plan based on calculated energy requirements was developed by a dietician for each subject with 500 kcal/d restriction for first 12 weeks and then maintenance for the second 12 weeks. Compliance to the diet plan was measured every 2 weeks by a 24 h dietary recall.
- Trial was registered at ClinicalTrials.gov
- The study was sponsored by Nestlé Research Center, Lausanne, Switzerland.
- Some information not included in the publication: sample size calculation, statement of primary endpoints, clear comparison of intention to treat and per protocol results for subgroup analysis based on sex.
Sanchez M, Darimont C, Drapeau V, Emady-Azar S, Lepage M, Rezzonico E, Ngom-Bru C, Berger B, Philippe L, Ammon-Zuffrey C, Leone P, Chevrier G, St-Amand E, Marette A, Doré J, Tremblay A. Effect of Lactobacillus rhamnosus CGMCC1.3724 supplementation on weight loss and maintenance in obese men and women. Br J Nutr. 2013 Dec 3:1-13. [Epub ahead of print]
Ley RE, Turnbaugh PJ, Klein S, Gordon JI. Microbial ecology: human gut microbes associated with obesity. Nature. 2006 Dec 21;444(7122):1022-3.