The purpose of this study was two-fold: 1) to determine the impact of replacing dietary carbohydrate with dietary protein and 2) to determine the impact of the Curves fitness and weight loss program on weight loss, body composition, energy expenditure, psychosocial outcomes, and markers of health in sedentary, obese females. This study represents the first of a series of studies by our research group to examine the effectiveness of the Curves fitness and weight loss program which is currently being followed by millions of women worldwide. Although our study contains many strengths such as our overall sample size, supervised exercise, dietary control measures and inclusion of exercise-only and no exercise/no diet controls, the marked difference in sample size among groups presents some challenges with interpreting our findings. While the authors acknowledge increasing the sample size in these groups would have been helpful, the primary objective was to assess the impact of altering the macronutrient ratio of the dietary regimens while also examining the overall impact of the exercise and diet programs used by the Curves system. Furthermore, statistical power analysis of our primary (e.g. waist circumference) and secondary outcomes (e.g. body mass and DXA body composition variables) ranged from 0.821 – 0.998 with partial eta squared values of 0.062 – 0.115 suggesting that our statistical analysis were adequately powered for our a priori determined end points. Our initial hypothesis was that participation in the exercise program would promote weight loss, improve body composition and fitness along with reducing markers of cardiovascular disease and that following a diet which restricted caloric intake while replacing dietary carbohydrate with protein at controlled fat intake levels would result in greater weight loss and improvements in health. Results from this study show that the greatest changes did occur in those groups that restricted their caloric intake in combination with the exercise program while participation in just the exercise program appears to have little to no effect over weight loss and body composition changes, a finding previously reported .
Results from the dietary analysis revealed significant differences in energy intake and carbohydrate and protein intake between dietary phases (e.g. Phase I, Phase II, etc.) and dietary groups (e.g. VLCHP and LCMP vs. HCLP). Overall dietary compliance during the 14 week program was deemed successful in all diet groups with the exception of the HED group, where it appears this group struggled to consume their prescribed level of calories (actual mean intake of 1846.6 ± 366.6 kcals·d1 vs. prescribed intake of 2,600 kcals·d-1). Caloric intake in this group, however, was higher by an average of 193 kcals·d-1, when compared to VLCHP, LCMP and HCLP. This allowed for comparison to lower and caloric intakes while exercising, which was the main premise for this higher dietary prescription, albeit not to the degree we originally hoped. This is likely due to either under-reporting or the inherent error associated with using dietary food records as recent studies from our laboratory have suggested that the magnitude of error associated with self-reporting of food intake is greater when higher caloric intakes are being prescribed [37, 38]. An additional important note is that although women in the moderate and high protein diet groups consumed a high percentage of calories in the form of protein, the relative intakes of this macronutrient ranged from 0.9 to 1.1 grams PRO·kg-1·d-1, respectively, suggesting that overall protein intake was increased above RDA guidelines.
Waist circumference is a key predictor of diabetes and cardiovascular disease . Findings from the present study suggest that participation in a resistance-based circuit exercise program can significantly reduce waist circumference irrespective of what dietary regimen is being followed. Further when dietary carbohydrate is replaced with protein, greater decreases in waist circumference may result (VLCHP:
= -6.3: [-8.7, -3.8]; LCMP:
= -6.7: [-8.7, -4.8]; HCLP:
= -5.7: [-7.5, -3.9]), but these between-group differences were not significant (Figure 2 and Table 2). Regarding weight loss, significant reductions in body mass were found in all groups that followed a diet and participated in the exercise program in amounts similar to recently published investigations [40, 41]. As expected, the greatest reductions in body mass occurred in those groups that restricted their calories to the greatest extent and for the longest period of time (VLCHP, LCMP and HCLP), but there were no differences between the amount of lost in the HCLP when compared to VLCHP and LCMP. These findings are consistent with other investigations that investigated the impact of macronutrient content on weight loss in diabetic and non-diabetic populations with no exercise intervention [14, 41–43]. Further, findings from Layman  suggested that a higher intake of protein with or without exercise was responsible for greater total weight loss . In the present study, participants following the VLCHP and LCMP programs experienced greater but non-significant levels of body mass and fat mass loss when compared to the HCLP group. Similar reductions (all P < 0.001) occurred in fat-free mass for VLCHP, LCMP and HCLP, while significant reductions in percent body fat were still found in all three groups. These findings contrast with other studies that have suggested higher protein diets may preferentially help to preserve lean tissue [12, 15] in those individuals who have replaced dietary carbohydrate intake with protein. Many reasons exist as to why differences in these findings exist, most notably subtle differences in the diet and exercise interventions. Nevertheless, present findings demonstrate the ability to lose a large percentage of weight as fat while maintaining a large amount of fat free mass while dieting and exercising [7, 16] and to a greater extent than other control conditions.
Reductions in energy expenditure have been reported in conjunction with restriction of caloric intake [42, 44]. This down-regulation has subsequently been linked to difficulties with weight maintenance as well as regaining the lost body mass over time . Consistent with many prior reports, significant reductions in REE occurred after consuming 1,200 kcal·d-1 for the first two weeks of the present study, even in conjunction with exercise. Caloric intake was increased by 400 kcals·d1 for the next 8 weeks which increased REE back to baseline levels in all groups except HED as this group experienced further increases in REE (Figure 5). A slight increase in energy intake during the following 8 weeks to 1,600 kcals·d1 (which was close to mean daily REE) stimulated increases in REE despite participants losing a significant amount of weight and fat mass. In these groups, REE was further increased when participant's began ingesting a 2,600 kcal: 55:15:30 diet with intermittent dieting (6 days out of 31) at their previously followed energy and macronutrient levels similar to previous energy balance approaches used in the literature . Collectively, these findings suggest that relative levels of energy expenditure can be maintained while completing a weekly resistance-based circuit training program in conjunction with modest caloric restriction (phase II) or potentially even intermittent dieting (Phase III).
As hypothesized, all groups except CON experienced improvements in peak oxygen uptake and maximal strength levels (Table 3) in accordance with other studies which have employed a regular exercise program to overweight/obese and/or sedentary populations [7, 40]. Minimal studies have incorporated a circuit-style resistance mode of exercise to this population, thus providing evidence that this type of training promotes general improvements in fitness in this population. Additionally, the serum reductions in glucose (2.6%), total cholesterol (2.5%) and LDL cholesterol (2.8%) were significant and similar to other studies which have employed dietary regimens that replace dietary carbohydrate in favor of greater dietary protein (Table 4) . While no significant group × time effect was found for insulin (P = 0.08), individuals in the VLCHP and LCMP groups experienced significant reductions, providing continued support that higher intakes of protein may help to promote such changes . A weakness of our study design is the inherent nature of pneumatic resistance. Such exercise modes make it impossible to determine how much resistance is actually being applied and subsequently creates challenges for documenting changes in total work throughout the exercise program in all groups. In this regard, it is possible that the total work completed from workout to workout and week to week didn't change or even decreased, although the physiological adaptations which did occur make this unlikely. For this reason, heart rate and ratings of perceived exertion were monitored throughout every exercise session in an attempt to promote consistency from workout to workout, while the number of completed repetitions were recorded to promote adequate progression. It was suggested that as fitness parameters increased each individual's ability to complete work would also be subsequently increased. Although each workout was intermittent in nature, heart rate monitoring did successfully keep individual workouts in the same relative range of exercise intensity
The relationship of leptin to weight loss, energy expenditure and insulin sensitivity has been characterized [3, 46, 47]. In this regard, circulating levels of leptin have been shown to decrease in response to decreases in energy availability , however, the influence of exercise and alterations in the macronutrient ratio is still undetermined. Volek and colleagues  suggested that significant decreases in leptin occur as part of an 8-week weight loss program, which similarly occurred in the present study. Further, Sartorio et al.  used a combination of energy-restricted diets with a 5 d·wk-1 aerobic and anaerobic exercise program and reported an acute, significant reduction in leptin levels which closely mimicked changes in body mass. As observed in the Sartorio study, serum leptin changes in the present study mimicked body mass changes and significantly decreased in all groups (except CON) after 2 weeks (Table 4). No significant changes occurred in any diet or exercise groups for markers of kidney and liver function (e.g. AST, ALT, GGT, creatinine, etc.) and fat or protein breakdown (e.g. ketones, total protein, BUN, etc.). Present findings support contentions that higher proteins do not invoke negative alterations in any of these serum variables after 14 weeks in obese, but otherwise health populations [1, 3, 43, 45].
An additional area of importance relative to public health and exercise adherence were the psychosocial assessments. In this regard, individual subscales of the SF-36 (e.g. physical functioning, bodily pain, general health, vitality, and mental health) significantly improved throughout exercise. While it could be stated that the significant increase in the bodily pain subscale was a negative response, it is an expected response considering the participants were sedentary for an extended period of time prior to the start of the study and placed in a resistance-training program, likely invoking some stiffness and delayed-onset muscle soreness. Additionally, significant increases in several subscales (e.g. appearance evaluation, appearance orientation, body area satisfaction and overweight preoccupation) of body image evaluation also improved in those individuals who were following the exercise program. In addition to improvement in several of the psychosocial variables, changes in these variables were significantly (P < 0.05) correlated to changes in body composition (e.g. fat mass and body fat %). Considering statistical probability that one of every twenty correlations run will result in a significant finding, it is possible that significance was found merely due to chance. While not intended to be causal, these findings suggest that changes in body composition may operate in conjunction with changes in body image as a result of exercise participation.
As such our investigation has many strengths and weaknesses to consider when evaluating this data against other data in this research area. Our study design could have been more complete by including a diet only control group with restricted caloric intake as this would have allowed us to more closely evaluate the impact of the exercise program in each dietary group. In this respect, a recent study has illustrated a similar weight-loss effect when a dieting only approach or a combination of exercise and diet is used if the net shift in energy expenditure is the same, however, a greater fitness benefit was reported which could have implications for cardiovascular health . Additionally, while the sample size across groups varied widely, our primary interest was in evaluating the changes in different dietary regimens with the exercise program. Strengths in our study involve the use of a rather large sample over several weeks of the intervention in comparison to other related studies. Also, a strength in the present study is our examination of a commercial program that is available to women all over the world and is currently the program of choice for millions of women at over 10,000 locations across the globe.