The present study is the first to examine the impact of CR on changes in appetite markers in a nonobese population, as well as whether the method used to obtain CR [i.e. CR alone, CR + EX, or LCD of 3724 kJ day−1 (890 kcal day−1)] differentially impacts appetite over time. The CR and CR + EX groups underwent a 25% energy deficit, resulting in a weight loss of approximately 10% of initial weight, and the LCD group lost approximately 14% of their initial body weight during the first 3 months of the study and subsequently remained weight stable for the remaining 3 months. Despite these significant weight losses, appetite ratings of participants in the three CR groups generally resembled those of a weight stable, non-restricted control group. Of importance, previous studies examining the effects of energy restriction on appetite have not included a weight stable control group; thus, the present study is the first to investigate natural variation in appetite changes among weight stable individuals over time.
In animal studies, hunger levels are typically elevated after prolonged CR, even during periods of energy balance (Speakman & Hambly, 2007
). Moreover, physiological satiety signals (e.g. GLP-1, leptin) are reduced after energy restriction in humans (Keim et al., 1998
; Adam & Westerterp-Plantenga, 2005
), and higher levels of self-reported hunger have been associated with poor weight loss maintenance (McGuire et al., 1999
; Pasman et al., 1999
). Few studies, however, have directly tested the effect that energy restriction has on subjective hunger and satiety levels (Wadden et al., 1997
; Doucet et al., 2003
). Moreover, findings from studies performed to date have been mixed. The reason for these disparate findings is not clear, but factors such as study duration, participant characteristics and the method used to induce energy restriction may all play an important role. Thus, we were particularly interested in studying the effects of prolonged CR (achieved through three different methods) on self-reported hunger and fullness levels in overweight, but not obese, humans. For hunger, the three CR groups changed in a similar manner; additionally, the control group had a +8% change, which was similar to the change observed in the CR groups. Reported fullness levels changed in a similar manner to hunger levels. There were no significant treatment effects, but reported fullness levels were decreased in all groups (range = 12–26%).
Participants’ reported Desire to Eat
was significantly increased at month 6 in the CR and LCD groups only, but was also increased in the CR + EX and control groups, suggesting that all four dietary interventions may have increased Desire to Eat
to some degree (range for all groups = 12–23%). Similarly, reported Prospective Food Consumption
was significantly increased only in the CR + EX but was also higher than baseline values in the CR and control groups (range for all groups = 13–21%). These findings are in line with previous studies demonstrating that both Desire to Eat and Prospective Food Consumption
were increased following 15 weeks of energy restriction and to an even greater extent after an 18-week low-fat diet and exercise follow-up programme (Doucet et al., 2000
). Ratings on the Satisfaction of Appetite
marker were significantly lower than baseline values at month 6 for the CR group only, but were also decreased in the all other groups, including the weight stable control group (range for all groups = 12–30%). To our knowledge, the present study is the first to examine the effects of prolonged energy restriction on Satisfaction of Appetite
; thus, this represents a novel finding of our study.
In general, the subjective appetite ratings of participants in the healthy diet control group changed in a similar manner to participants in the three treatment groups. The reason for these subjective appetite changes is not clear, particularly because the participants in the healthy diet control group were weight stable throughout the study. This is the first study to report subjective appetite changes among weight stable individuals on a healthy diet. Thus, our findings may reflect the natural variation in appetite that occurs among individuals consuming a weight stable diet. It is also possible, however, that the demand characteristics of the present study contributed to the reported changes in subjective appetite ratings over time among participants in this group. For example, similar to participants in the three treatment groups, the participants in the healthy diet control group were provided with a study diet and were asked to report their hunger levels when in a fasted state in a clinic environment. Moreover, participants in all groups were aware that they were participating in a caloric restriction study.
Relative to previous studies, the present study had several strengths. First, no study to date has examined the impact of CR on appetite in a nonobese population. As such, the present study provides important information regarding the feasibility of CR as a strategy to improve health in nonobese populations. Second, the present study is the first to test whether three different methods of CR vary in terms of their impact on appetite. Third, the present study is the first to include a healthy diet weight stable control group when evaluating the impact of different methods of CR on appetite markers. Without inclusion of a control group, the effect that CR has on appetite markers cannot be fully determined. The additional strengths of the study include the frequent assessment of appetite ratings and close monitoring of energy intake and expenditure; adherence levels were found to be very good in all conditions, as demonstrated by the significant weight losses in the three CR conditions. Finally, retention rates were very high (96%), particularly given the demands of the present study.
The present study also had a number of potential limitations. First, the VAS data were collected during feeding periods at the research center only and were not collected during the period of self-selected food intake (weeks 12–21). However, this may also be viewed as a strength because adherence to energy intake recommendations were closely monitored during the feeding periods at the research center. Second, the sample size for the present study was small, which decreased statistical power to detect significant interaction effects (i.e. treatment by time); thus, our results should be interpreted with caution. Nevertheless, effect size calculations (generalised eta squared) indicated that no more than 5% of the variance in appetite rating change could be attributable to treatment. Additionally, our findings are limited to the first 6 months of CR, which corresponds to the weight reduction phase rather than weight maintenance phase in most obesity studies. However, participants in the LCD condition were weight stable for the last 3 months of the study (half of the study duration) and their appetite ratings were similar to the other two CR groups. Future studies are warranted to determine whether appetite markers change with long-term weight maintenance, as well as whether changes in appetite markers predict weight regain. Another potential limitation may have been the inherent demand characteristics of the study, which may have affected reported changes in appetite ratings among participants in all groups.
In summary, there were no significant treatment effects on subjective ratings of appetite, and the effect sizes were small for all measured appetite markers. Participants’ Hunger and Fullness ratings in the three intervention groups did not differ from baseline levels and were similar to participants in the healthy diet weight stable control group, despite significant weight losses. Future studies are needed to further explore the viability of CR, as well as different methods of CR, to promote healthy weight management, as well as a strategy to potentially achieve other health benefits.