I know that 99% of the SuppVersityreaders will be smart enough not to believe that calories don’t count. Accordingly, it may be news, but probably not very surprising for most of you to hear that a recent (re-)analysis of food logs from two intermittent fasting studies underlines how effective intermittent fasting really is when it comes to reducing people’s energy intake. In previous studies researchers have already been able to show that the vast majority of dieters do not overeat and consume more than the ~1500kcal/d you may be missing on a fasting day.
A group of British researchers has now readdressed the issue. In their latest paper in “Food Science and Nutrition” (Harvey 2018) they assessed the initially wanted to quantify the “extent of compensatory overeating on unrestricted days” by reanalyzing subjects food logs from two randomised trials of comparisons of IER and CER (Harvie et al., 2010, 2013).
With the first study being conducted in young overweight women and the second one spanning larger age-group, both studies target similar populations, the “fasting” protocols were different, though. with ~650kcal on two fasting days per week in Harvie 2010, and two low-carb days per week in Harvie 2013. Here’s how the scientists summarize the protocols themselves (Harvey 2018):
Study 1 tested a 2 day IER in 53 premenopausal women with overweight or obesity (BMI 24–45 kg/m2) over six months. The IER was designed to provide an overall 25% energy restriction below estimated energy requirements and involved two consecutive days each week of 70% energy restriction below estimated energy requirements, and five relatively unrestricted days /week designed to meet 93% of their estimated daily energy requirements calculated using the Schofield Equation (Schofield, 1985). Restricted days in Study 1 provided 2,700 kJ and 50 g protein and comprised of 1.136 L (2 pints) of semi-skimmed milk, four 80 g portions of vegetables, one portion of fruit, a salty low energy drink, and an over the counter multivitamin and mineral supplement.
Normal eating days advised a Mediterranean‐type diet (30% of energy from fat; [7% saturated, 15% monounsaturated and 7% polyunsaturated fatty acids], 45% low glycaemic load carbohydrate and 25% protein). Participants were advised on maximal quantities of protein, carbohydrate, and fat on these days and weekly guidance for treat foods (3 per week) and alcohol (14 units/week) to prevent over-consumption. Seven-day food diaries were completed after 1 and 6 months on the diets and were checked for completeness with participants and analyzed using Compeat Nutritional Analysis System (Carlson Bengston Consultants, London, UK).
If someone tries to tell you that intermittent fasting will – just like exercise – make you even more susceptible to overeating, show them the results of Buthani’s 2013 study on the “effect of exercising while fasting on eating behaviour and food intake”.
How does the magic happen? If we are honest, we have to admit that we don’t even understand why the overeating that is a commonly observed in rodent studies (in almost all of them the rodents don’t overeat to an overcompensation level btw) doesn’t occur in humans… or, as in these particular studies, in overweight women. When it comes to rationalizing the “carry over effect”, which has likewise been observed by Klempel et al. (5% with alternate day fasting in | Klempel 2010) , does in fact seem to conflict with both common sense and previous short-term fasting studies conducted in laboratory settings amongst subjects not attempting to lose weight… and that’s exactly where an often-ignored argument is coming in: at least, when you’re still significantly overweight, every step in the right direction is going to make it easier to continue on the track.
In part that’s certainly due to behavioural aspects of following an IER as part of a weight loss programme, of which Harvey et al. explain that “[a]necdotal reports from subjects in the two reported trials suggest IER can make individuals more aware of their habitual food intakes and habits, increase awareness of appetite and hunger, and provide assurance that they can function adequately during restricted days without the need for extra energy on surrounding days” (Harvey 2018). That it’s (mainly) a behavioural/learning component would be supported by data from Bhutani et al. (2013) reported increased restrained eating and decreases in uncontrolled eating amongst overweight subjects undertaking alternate day energy restriction (75%).
Before you start celebrating, though, I have to warn you: the targeted weekly calorie deficit in the studies at hand was relatively small and as Harvey et al. point out it is well possible that a similar phenomenon “may not occur if IER diets are presented as spells of severe restriction and spells of feasting as is the case with many commercial IER diets” (Harvey 2018).
Study 2 tested two different versions of a 2-day IER for 3 months of weight loss and 1 month of weight maintenance (with one restricted day per week) amongst 75 pre and postmenopausal women with overweight or obesity (BMI 24–45 kg/m2). One IER was designed to provide an overall 25% energy restriction below estimated energy requirements and involved two consecutive days of energy and carbohydrate restriction each week (energy restricted low carbohydrate IER, 2,500–2,717 kJ/day, <50 g carbohydrate/day). Each restricted day allowed approximately 250 g of protein foods, three servings of low‐fat dairy foods, four 80 g portions of low‐carbohydrate vegetables and one portion of low‐carbohydrate fruit, at least 1,170 ml of other low‐energy fluids, and an over the counter multivitamin and mineral supplement.
The five unrestricted days were as above with advice on maximal portions of foods and treats to ensure participants did not overconsume. The other two day IER regimen tested was similar, but allowed unlimited lean meat, fish, eggs, tofu and unsaturated fats on restricted days (ad libitum low carbohydrate IER). Seven-day food diaries were collected and checked for completeness with each participant at 1, 3 and 4 months and were analyzed via Wisp (Tinuviel Software, Anglesey, Wales).
Aside from 7-day food logs, the scientists also had data available that would allow them to classify the subjects into two groups: those that fasted flexibly, i.e. on different days and those that stuck to a certain weekly schedule… and what shall I say:
Here’s where the interesting stuff starts!
Instead of negative effects on adherence and thus reduced fat loss, a comparison of the two groups across studies revealed a significant positive effect of using a more flexible regimen with less than 50% of restricted days on the same 2 days each week:
“When studies were combined percentage weight loss at 3 months was −5.8 (−6.7 to −4.7) % in the consistent group and −7.4 (−8.7 to −6.2) % in the non‐consistent group (p = .09)” (Harvey 2018)
With the overall large standard deviations, the inter-group difference didn’t reach statistical significance – if we take the relative difference of the means, i.e. 1.6% that’s more than some people manage to lose in total over a three months period and certainly note-worthy (see bottom line for thoughts on why this was the case).
Intermittent fasting is when undereating triggers further undereating!
Ya, I know. It sounds stupid, but the above, i.e. the observation that intermittent fasting does not just fail to trigger energy overconsumption on the following day, but rather
“a spontaneous reduction in energy intake below [subjects’] baseline reported intakes and the prescribed isoenergetic diet during all unrestricted days including the days immediately before and after restricted days” (Harvey 2018)
is a consistent result that was observed across of both of Harvie’s previously cited studies. Now, we are dealing with self-reported food intakes and would have to confirm that in future studies with tighter dietary control. For the time being, however, the (in both studies) significant advantage over a control diet does yet clearly support the notion that all subjects managed to reduce their food intake ~4%-6% below their reported baseline energy intake (Table 1).
Table 1: Energy intake at baseline and on the different unrestricted days of intermittent energy restriction in Study 1 (n = 44) and Study 2 (n = 67) in (kJ/day | Harvey 2018)
In that, a closer analysis of the data in Table 1 informs us that the subjects in study 1 (young, overweight women) were actually restricting their energy intake by 19% on the unrestricted days… voluntarily (they were supposed to eat 7,728 kJ, but ended up with only 6,250 kJ/day). Similar results have been observed in study 2, where only the carbohydrate intake was restricted:
“In Study 2 the mean (95% CI) recommended intake during unrestricted days for both groups was 7,546 (7,399–7,691) kJ, whereas the actual mean intakes were comparable between the energy restricted low carbohydrate IER 5,775 (5,243–6,243) kJ/day and the energy restricted ad lib low carbohydrate IER 6,427 (5,907–6,945) kJ/day (p > .05). The combined groups in Study 2 had a 19% energy restriction below the recommended intake on these days” (Harvey 2018).
As the authors . rightly point out, the “overall energy restriction across the whole week including restricted and unrestricted days in Study 1 and 2 was [thus] 39% below estimated energy requirements, rather than the planned 25% energy prescription” (Harvey 2018)…Don’t get me wrong – Calories count, but they are not the only thing that counts: Food quality obviously counts, as well (some may argue, dependendently, though). The satiety effects of what you eat, for example, and the protein content of our diet are both important parameters, as well. How come? Well, with satiating foods that don’t send you on an insulin rollercoaster you will have a much easier time sticking to the fasts (applies to protein, too; due to its satiety effects) and with sufficient protein you can minimize the loss of lean mass (esp. if you lift at the same time).
… and again: that happened in response to an intervention of which many well-paid experts will still try to tell you that it’s ‘just going to make you overeat’.
Neither on the day before (pre-) nor after (post-)compensatory mechanisms were observed
Figure 1: IF will not make you overeat – the opposite is the case.
In fact, GEE regression analyses of the data from the days immediately before and after fasting days clearly indicate that neither in study 1, where data were available on 44 subjects for 59 days immediately following (after) and 59 days immediately before a restricted day (before) and for 177 other unrestricted days (other), nor in study 2, where data were available on 67 subjects for 166 days immediately after and 73 immediately before a restricted day and for 632 other unrestricted days not including days adjacent to a restricted day, showed significant differences in energy intake on the day immediately before and/or after the two restricted days compared with the ‘other’ unrestricted days (p = .46).
Which weekdays you’re fasting on, doesn’t seem to matter, either; after adjusting for the day of the week the scientists found no significant difference across the months that food diaries were completed (p = .38), and no significant difference between days of the week (p = .09). Moreover, there were no statistical significant differences in energy intake between “before” and “other” unrestricted days (p = .98), after adjusting for the day of the week and no significant difference between days of the week (p = .08) nor between months (p = .78 | all p-values from study 1, but generally the same results in study 2).
Bottom line: The first thing the re-analysis of the studies does for us is to confirm the notion that intermittent fasting is not a trigger of overeating in the average overweight woman (young and older).
The second thing it the paper at hand suggests is that being flexible in terms of when to take those fasting days seems to be an (albeit non-significant) advantage – an observation that seems only logical, but it will allow you to be more socially compatibel by, for example, post-poning your fasting day from Wednesday to Thursday if the former is a friends birthday and you’d have the choice between falling off the dieting wagon, and/or being the social pariah “on that funky diet, where you don’t eat for X days of the week”.
In conjunction with the excellent and (for many people) pain/hunger-free results (at least in the latter phases of the studies at hand, the subjects no longer complained about hunger on the fasting days), the two previously mentioned study outcomes alone would make intermittent fasting worth trying – the fact that it seems to inspire people (in this case overweight women) to reconsider their overall food intake and/or get used to eating less (could also be that this is a downstream effect of increased metabolic health), so that they reduce their food intake by another -14% beyond what’s prescribed by their diet, however, should convince even true skeptics that dieting two days a week could be a feasible, realistic long-term weight loss strategy that’s worth trying out
Bhutani, Surabhi, et al. “Effect of exercising while fasting on eating behaviors and food intake.” Journal of the International Society of Sports Nutrition 10.1 (2013): 50.
Harvey, Jennifer, et al. “Intermittent energy restriction for weight loss: Spontaneous reduction of energy intake on unrestricted days.” Food Science & Nutrition 6.3 (2018): 674-680.
Harvie, Michelle N., et al. “The effects of intermittent or continuous energy restriction on weight loss and metabolic disease risk markers: a randomized trial in young overweight women.” International journal of obesity 35.5 (2011): 714.
Harvie, Michelle, et al. “The effect of intermittent energy and carbohydrate restriction v. daily energy restriction on weight loss and metabolic disease risk markers in overweight women.” British Journal of Nutrition 110.8 (2013): 1534-1547.
Klempel, Monica C., et al. “Dietary and physical activity adaptations to alternate day modified fasting: implications for optimal weight loss.” Nutrition journal 9.1 (2010): 35.