Page 1 of 2

5:2 Questionnaire results...stats and everything! Must read!

PostPosted: 09 Dec 2013, 19:33
by carorees
Back in April 2013, I together with a few forum members (notably @PhilT and @eakman) created a questionnaire to try to get to the bottom of why some people lose weight faster than others with 5:2. Lots of you kindly filled in the questionnaire and have been patiently waiting for the results. I was held up by not having the skills to do the necessary statistical analysis but finally @P-JK offered to help and together we have written a report on our findings.

It is interesting stuff! Here is a summary of our report and a link to where you can read the full document: https://www.academia.edu/5360346/Factor ... l_analysis

The results of a questionnaire sent out to members of the 5:2fastdiet forum were analysed statistically to try to discover what are the main factors influencing weight loss with this diet and how great an influence each factor has. As is the case with all diets, gender, starting BMI and duration of time on the diet were, unsurprisingly, key factors influencing the speed of weight loss. Overall, a quarter of the variation in speed of weight loss is accounted for by these three factors. As few men answered the questionnaire and some factors appear to have a different effect for men and women, our full analysis had to be restricted to only the women participants. While physical activity might be expected to enhance weight loss, we were unable to show this in our analysis. This is likely due to insufficient data about change in activity levels before and after starting 5:2. An important factor influencing weight loss is the number of hours with no calorie intake on fast days. Fasting fully (no calories) for more than 20 hours on fast days increased weight loss by around 120 grams per week compared with a 12 to 16 hour fast. Only a few people fast for less than 12 hours or longer than 24 hours but the difference between these extremes in terms of weight loss appears to be even larger at 240 grams per week. Adding an extra fast per week (i.e., 4:3 rather than 5:2) seems to result in faster weight loss, but the participants using 4:3 were too few for this effect to show in the full analysis. Although the main calorie restriction is conferred by the 2 days’ fasting, participants’ eating behaviour on the non-fast days had an important effect on the speed of weight loss. Firstly, bingeing on feed days can, unsurprisingly, slow down the weight loss. Compared with women who never binge, those who binge sometimes or often experience slower weight loss: in the order of 120 grams per week. Bearing in mind that the average weight lost per week for the women in our study was 490 grams per week, an effect size of 120 grams is important. Secondly, a change in dietary habits to a ‘healthier’ diet (defined as eating fewer snacks, less sugary food, fewer ready meals and eating more vegetables) appeared to increase weight loss by around 250 grams per week. Roughly one half of this effect is estimated to be the result of a change in food content, the other half is estimated to be due to eating less food (change in food quantity). Our final analysis, which takes into consideration the five main factors (BMI, weeks on diet, binge behaviour, healthy diet and length of fast) was able to explain almost half (40%) of the differences in rate of weight loss seen in the women who answered our questionnaire.


TL;DR
Apart from the obvious factors of being heavier to start with, being male and being on the diet for longer, we found that longer fasts, together with no bingeing and a change to a healthier diet on feed days seem to explain a lot of the variation in speed of weightloss between people.

Thank you to all those who took part and who helped with the analysis of the results. I'm hoping that, armed with this preliminary information, we can design a bigger, better questionnaire next year that would enable us to dig deeper into these factors!

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 19:44
by cblasz
WOW! That is very impressive and very interesting! It seems to make sense overall. Thank you for doing this!

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 19:51
by Ballerina
That was interesting, thank you Caroline for all your input on this forum and for your amazing in depth knowledge of how our biochemistry works. I know it certainly helps me understand what is happening to my body and why

Ballerina x :heart:

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 19:55
by jools7
wow, thanks so much for your efforts, @carorees! I am so not a researcher, so I absolutely take my hat off to you for your hard work! This is awesome!

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 20:10
by PennyForthem
Very interesting and I echo the thanks

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 20:18
by Sue.Q
:heart: Great read and very interesting indeed thank you Caroline and your assistant for all the hard work :heart:

I didn't get to fill it in because I was at my start and a
complete useless idiot with this site Lol
( not much better now)
However I always believed that me doing 4:3 was the reason for my steady consistent weightloss :heart:
:heart: :heart: 50lbs down + 50lbs to go next year :clover: Sue

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 20:29
by Wmr309
Looking forward to reading the full document but right now off to make myself a nice breakfast. Just reached my goal with this wonderful WOE! :victory: I can't do the 20 hours of no calories, but I have increased my exercises significantly and low carbs on most other days. Thank you @Carorees and @P-JK for your work on this.

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 21:10
by Wendy Darling
Thank you for that Caroline and PJ-K. I'm sorry if I missed it, but do you think that those of us with less to lose, lose at a slower rate? When I compare my stats with some of our forum buddies with more to lose, mine is sooooo slooowww and they are steaming ahead. It is something I've always wondered. Hope that's not a stupid question. Thanks

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 21:32
by carorees
Yes, starting BMI was a significant factor. The analysis suggests that for every 5 BMI points higher your starting BMI the speed of weight loss increases by about 50grams per week. So a person with a BMI of 40 would lose around 150g per week more than someone starting at a BMI of 25.

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 21:56
by P-JK
In addition: it is also to be expected that the effect of lower BMI is not linear when BMI is getting closer to 20 (as for most people there will not be much to lose). There are some indications of this (so a particular small weight loss at BMI levels below 22), but as only very few participants were in this category, we have to be careful with conclusions. Some more info is in the full paper.

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 09 Dec 2013, 23:36
by miffy49
Thank you both. An interesting read. As Ballerina has said, being better informed about things makes it easier to understand the various aspects of how it all works. I for one like to know why I should be following a particular plan rather than just doing it because someone said so!

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 10 Dec 2013, 00:06
by bordergirl
Thank you so much for your work on this. Some things weren't surprising at all while the large effect of the number of hours fasting was an "aha!" moment. I've always been frustrated with studies that are of short duration and limited scope, but I am understanding better the difficulty of isolating or separating the various factors and determining which thing has a given effect. But, even so, real-life anecdotal evidence is also valuable because we don't live in laboratories, we live in the real world. Messy, but real!
As we continue to accumulate information over time and with a broader population, the force of this WOE will also become stronger.

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 10 Dec 2013, 10:32
by P-JK
Just a bump to get some attention (as not all members wil be regular visitors of the NERDY section ...)

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 10 Dec 2013, 11:00
by Moogie
Tweeted about this to bring in some more readers, thanks for all your hard work on this it's amazing!

Re: 5:2 Questionnaire results...stats and everything! Must r

PostPosted: 10 Dec 2013, 11:07
by carieoates
Interesting read and many thanks. I've always suspected that with a BMI of just 29 at the start (27 now) I found a loss very slow going. No steaming ahead for me either. More a case of lose, maintain, maintain, lose.