Lessons from Studying the Most Successful Weight Loss Dieters
by Dr. Mike Israetel, Chief Sport Scientist / Anshuman Radhakrishnan / |
Jan 31, 2021
Over 30,000 people use the RP Diet Coach App every month. They pick the foods they want to eat, and the app tells them how much to eat and when. App users choose diet duration and body weight goals, and the app makes recommendations for altering the diet (increasing, decreasing, or leaving food amounts the same) throughout the process in order to keep the user on target.
In our quest to make the app a better coach and further our understanding of what dieting parameters best predict success, we’ve been collecting (anonymous) data from app users. We now have over 10,000 completed diets to analyze for clues about what approaches work best. Our first investigation sought to determine the likely best duration, loss-rate, and level of adherence to the diet; it can be found here. This, our second investigation, focuses on what choices are common to more or less successful dieters. In order to study this, we broke dieters into four groups:
- Group 1: Dieters that achieved their weight loss goal plus additional weight loss (> 0.5% beyond their goal)
- Group 2: Dieters that lost within +-0.5% of their weight loss initial goal.
- Group 3: Dieters that lost more than half of their initial goal loss, but didn't lose enough to make the Group 2 cutoff.
- Group 4: Dieters that lost less than half of their initial goal loss.
First, here is a list of the variables in our analysis that did not in any meaningful way correlate with diet achievement:
- Starting weight
- Proportion of whole vs. packaged and restaurant foods consumed
- Initial body fat levels
- Diet diversity
- Daily activity levels
- Sleep amounts
- Number of meals per day
Now, it’s not that those variables don’t affect diet success at all. Many of them do! You can’t just get four hours of sleep every night and expect to be just as successful with your diet as you would with eight hours nightly. In many cases the groups did not differ along the above parameters, so we could not hypothesize whether those variables impacted success. For example, Group 1 input an average of 8.69 hours of sleep per night, Group 2 8.71 hours, and Groups 3 and 4 both input an average of 8.74 hours per night. All groups averaged 4-5 meals per day (in fact, the average was ~4.53 meals per day for all groups!).
Before we look at the variables that did differ between groups and therefor might be good predictors of success, let’s look at how diet outcomes differed between our four groups. Plotted below are weight lost as a percent of initial weight and absolute weight lost, on average, for each group:
Group 1 dieters lost almost 7% of their body weight on average, amounting to an average of about 13lb. In contrast, Group 4 dieters lost just under 2% of their body weight on average, amounting to an average of just over 3lb.
The first difference between Groups 1 and 4 is in their planned diet durations. Group 1 dieters (the overachieving group) were twice as likely to choose a shorter, six week diet than they were a 12 week diet. Group 2 dieters, those that pretty exactly met their goals, were three times more likely to choose a six rather than 12 week diet. Group 3 was still more likely to diet for six weeks than for 12, but by a much smaller margin. Group 4 dieters, in contrast, tended to choose ta longer duration diet; they were more than twice as likely to set their diet duration to 12 weeks versus six. The most successful dieters favored shorter diets.
*The higher the ratio, the SHORTER the average diet was.
Something very notable about these data is that even though the more successful dieters very often opted for the shorter duration diets, they still lost more weight on average than those opting to diet for longer! The most successful group lost in six weeks what the least successful failed to lose in 12. In addition, Group 2 (the “on target” dieters) lost almost as much weight as Group 3, and thus managed to get about the same results while dieting for less time on average.
|Planned Diet Length
While we cannot conclude causation, it seems that the most successful dieters, whether we measure this by how close they got to their goal, the percentage of body weight they lost, or how much total weight they lost, chose diets of shorter duration (~6 weeks). Those who chose longer diets (closer to 12 weeks) tended to be less successful. Does this mean that those who want to be successful should diet for shorter durations? It may or may not, but it certainly makes trying a shorter duration diet worth considering.
The second very meaningful difference we found between the groups was for likelihood of adherence to the app AI’s recommendations. Group 1, the diet over-achievers, chose to take the app AI’s advice on raising, maintaining, or lowering food amounts during the course of their diets about 92% of the time! Group 2 and 3 took the app AI’s suggestions about 89% and 86% of the time respectively. Group 4 dieters, the least successful with respect to their initial goals, followed the app’s guidance only about 74% of the time. It seems that the most successful dieters did what the app’s AI asked of them more often than the other groups.
Adjustments Followed Proportion
- Group 1: 0.918
- Group 2: 0.890
- Group 3: 0.861
- Group 4: 0.736
We cannot say conclusively based on these data that dieting for shorter periods of time or following the app’s AI recommendations causes an increase in success, but the most successful dieters exhibited both of these behaviors. We do know from other research that following a plan and limiting hard dieting to shorter durations both have benefits for weight loss. If you have had some less than successful diets (or even fairly successful ones), perhaps try a shorter diet and take the app’s recommendations seriously next time. A chance at more success across a shorter dieting period is worth a shot!