3 ms·

Even the abstract directly contradicts the conclusion: > The mean weight loss from baseline at 12 months was −8.0 kg (95% confidence interval [CI], −9.6 to −6.

by teilo 4y ago

Even the abstract directly contradicts the conclusion: > The mean weight loss from baseline at 12 months was −8.0 kg (95% confidence interval [CI], −9.6 to −6.4) in the time-restriction group and −6.3 kg (95% CI, −7.8 to −4.7) in the daily-calorie-restriction group. So in other words, beyond caloric restriction, the time-limited group lost more weight. The study seems to have ignored other measures, such as inflammation markers. Further, a 16/8 restriction is barely a restriction at all. It's trivial to consume 3 meals in that window. It's also too wide a window to measure the results of autophagy, one of the key reasons to do intermittent fasting.

ProjectArcturis 4y ago
The confidence intervals overlap by quite a bit. The difference is not statistically significant. If this were a clinical drug trial, the drug would not get approved. >The study seems to have ignored other measures, such as inflammation markers. They also ran a panel of metabolic markers, which the abstract says were in line with the overall weight loss results.
- foldr 4y ago
  This seems to be a (very typical) misuse of NHST. All we want to know in this case is which of the following statements is the more probable: (1) Intermittent fasting has benefits. (2) Intermittent fasting does not have benefits. Instead, what we're being told is that we can't reject (2) with a high degree of confidence. When you look at the sizes of the groups in the study, this is pretty obviously going to be the case. With so few participants there'd have to be an enormous difference between groups in order to be sure that (2) is false. They need hardly have bothered doing the study at all.
gamerDude 4y ago
Well, the trick is that the confidence intervals overlap, and do so pretty heavily. So, while on average they may have lost more, their test group size was either quite small or had high variability such that it could come down to randomness that one group lost more than the other. At the end of the study they had 61 in calorie restricted and 57 in intermittent fasting. So, not huge groups. And looking at their confidence range knowing their numbers, it likely means on both sides multiple people lost 7kg on both sides. I didn't see a way to download that study to get the raw data, which would also be helpful, as I think it would also be useful to look at % body fat loss, etc.

by teilo 4y ago

ProjectArcturis 4y ago

The confidence intervals overlap by quite a bit. The difference is not statistically significant. If this were a clinical drug trial, the drug would not get approved. >The study seems to have ignored other measures, such as inflammation markers. They also ran a panel of metabolic markers, which the abstract says were in line with the overall weight loss results.

foldr 4y ago
This seems to be a (very typical) misuse of NHST. All we want to know in this case is which of the following statements is the more probable: (1) Intermittent fasting has benefits. (2) Intermittent fasting does not have benefits. Instead, what we're being told is that we can't reject (2) with a high degree of confidence. When you look at the sizes of the groups in the study, this is pretty obviously going to be the case. With so few participants there'd have to be an enormous difference between groups in order to be sure that (2) is false. They need hardly have bothered doing the study at all.

gamerDude 4y ago

Well, the trick is that the confidence intervals overlap, and do so pretty heavily. So, while on average they may have lost more, their test group size was either quite small or had high variability such that it could come down to randomness that one group lost more than the other. At the end of the study they had 61 in calorie restricted and 57 in intermittent fasting. So, not huge groups. And looking at their confidence range knowing their numbers, it likely means on both sides multiple people lost 7kg on both sides. I didn't see a way to download that study to get the raw data, which would also be helpful, as I think it would also be useful to look at % body fat loss, etc.