Preliminary reports indicate that ADMF may be an effective strategy to help obese individuals lose weight [13
]. However, the dietary and physical activity adaptations that contributed to this pronounced weight loss by ADMF were not tested previously. We show here, for the first time, that weight loss by ADMF occurred due to change in meal pattern, i.e. obese subjects limited their energy intake to 25% of needs on the fast day with no hyperphagic response on the feed day. This change in meal pattern helped these subjects to achieve a marked degree of energy restriction (37% net daily) which was related to the pronounced weight loss attained (5.6 kg in 8 weeks). This study is also the first to demonstrate that subjects become habituated to the ADMF diet (i.e. feel very little hunger on the fast day) after approximately 2 weeks, and that physical activity habits are not affected by fasting on alternate days.
A key objective of the present study was to examine the degree of energy restriction achieved by ADMF and to investigate how this relates to rate of weight loss. In order to measure energy intake and percent energy restriction, we asked obese participants to complete food records on feed and fast days throughout the trial. Results from the food record analysis reveal that obese subjects were able to consistently limit their energy intake to approximately 25% of needs (500 kcal) on the fast day. Our data also show that the ratio of energy consumed on the fast day versus the feed day did not differ between phases. However, it should be noted that there was a trend towards consuming less energy on the feed days, and more energy on the fast days over the course of the trial. It will therefore be of interest in long-term ADMF studies to examine whether restriction gradually diminishes on the fast day after several months of diet. The degree of hyperphagia that occurred on the feed day in response to the lack of food on the fast day was also assessed. Our data indicate that no hyperphagic response took place as subjects only consumed approximately 95% of their calculated energy needs on each feed day throughout the trial. These findings therefore suggest that obese subjects are able to drastically change their meal pattern in a way that conforms to the ADMF protocol. Nevertheless, there are several limitations to these data that must be discussed. First and foremost, it is well known that obese subjects underreport energy intake by 20 to 40% when completing food records [20
]. The extent to which these subjects underreported energy intake became apparent when we tried to relate reported energy intake to the weight loss achieved. From the food record data, we calculated that, on average, subjects were restricted by 37% of calculated needs every day. If the subjects were indeed restricted by this amount, this would have resulted in a rate if weight loss of 1.2 kg/week. In actuality, the rate of weight loss was 0.7 kg/week. This disparity between reported intake and weight loss can be observed when examining the limited amount of weight lost during the 4-week self-selected feeding phase (body weight reduction of 93.8 kg to 92.8 on the feed day, equivalent to 1 kg of weight loss). The incongruity between self-reported energy intake and rate of weight loss therefore suggests that subjects were underreporting energy intake. In view of this, it will be important for future ADMF trials to assess energy intake and energy restriction by more robust methods, such as the doubly labeled water technique [22
]. It should also be mentioned that assessing body weight changes by ADMF is difficult as weight measurements are drastically different from feed to fast day. This discrepancy in body weight is most likely due to the additional weight of food present in the gastrointestinal tract, and not changes in fat mass from day to day. As a potential solution, future trials of ADMF should average body weight measurements taken from consecutive feed and fast days to attain a more accurate assessment of weight.
In addition to energy intake, we also examined changes in dietary macronutrient composition throughout the course of the trial. We hypothesized that during the ADMF controlled feeding phase (weeks 3-6), when dietary counseling was not provided, subjects would instinctively choose higher fat/more energy dense foods on the feed day to make up for the lack of energy consumed on the fast day. Interestingly, fat intake did not increase from the baseline period (36% of kcal) to the ADMF-controlled feeding period (36% of kcal). These preliminary data suggest that subjects are not likely to consume higher fat diets on the feed day when partaking in an ADMF regimen. We also hypothesized that the dietary counseling provided during the self-selected feeding phase (weeks 7-10), would help subjects decrease total fat, saturated fat, and cholesterol intake, while increasing fiber intake. Results reveal that counseling assisted these individuals in lowering their total fat and saturated fat intakes to levels that conform with NCEP dietary recommendations [19
], and that these changes in fat intake were related to rate of weight loss. On the other hand, dietary counseling appeared to have no effect on cholesterol or fiber intake. This lack of effect of dietary counseling on the intakes of these nutrients has been reported previously [24
]. It should also be noted that fiber intake on the fast day was particularly low (7-10 g/d). Since quantity of food consumed on the fast day is limited, it would be difficult for individuals to meet fiber requirements [25
]. As such, it is recommended that future trials in the ADMF field provide a fiber supplement on the fast day to help individuals meet recommendations [19
Changes in perceived hunger, satisfaction with diet, and fullness were also evaluated on each fast day throughout the trial. This study is the first to show that obese subjects become habituated with ADMF after approximately 2 weeks of diet (i.e. feel very little hunger on the fast day). Our data also demonstrate that subjects become more satisfied with ADMF after approximately 4 weeks of diet. Feelings of fullness, however, remained low across the course of the trial suggesting that subjects never felt "full" at any point while undergoing 8-weeks of ADMF. These findings may have important implications for long-term adherence to ADMF by obese men and women [27
]. More specifically, since hunger virtually diminishes, and since satisfaction with diet considerably increases within a short amount of time (2-4 weeks), it is likely that obese participants would be able to follow the diet for longer periods of time. It is important to note, however, that the subjects only completed the VAS scales pre-bedtime. Thus, the data only reflects their feelings immediately before going to bed, and is not indicative of their feelings of hunger and satisfaction throughout the day. Future trials in this area should administer these VAS scales throughout the day to obtain a more complete data set for these variables. It should also be noted that hunger spiked at week 8. We speculate that this may have occurred because this study week corresponded to Memorial Day weekend, and subjects may have felt hungrier while attending food-related celebrations. Moreover, trials examining the ability of obese subjects to comply with ADMF for longer durations (i.e. 24 to 52 weeks), and in consequence, lose larger amounts of weight, will be an important focus of future research in this field.
The effects of ADMF on habitual physical activity was also assessed by having the subjects wear a pedometer on everyday of the study. We hypothesized that subjects would feel less energetic on the fast days, and would therefore be less physically active (i.e. take less steps/d) on fast days than feed days. Surprisingly, physical activity level did not differ between feed and fast days. Moreover, there was no difference in activity level when steps/d taken during the ADMF phase were compared to steps/d taken during the control phase. Similar results have also been reported in normal weight individuals undergoing ADMF [12
]. These data suggest that obese individuals are able to maintain their level of habitual physical activity despite decreases in energy intake on the fast day. This maintenance of physical activity while undergoing ADMF would thus allow obese individuals to lose weight consistently on feed and fast days as energy expenditure would stay constant.
A key limitation of this study is that there was no true control group. Having a control arm run parallel to the treatment (ADMF) arm would have strengthened the study by allowing us to: 1) compare changes in the ADMF group to that of a non-restricted control group at each time point, and 2) identify events (such as holidays) that may have resulted in deviations from the prescribed diet. Future studies aiming to test similar objectives should employ a control group where possible.
In summary, these findings indicate that obese subjects quickly adapt to ADMF, and that changes in energy/macronutrient intake, hunger level, and maintenance of physical activity play a role in influencing rate of weight loss by ADMF. These preliminarily data offer promise for the implementation of ADMF as a long-term weight loss strategy in obese populations.