This study is one of few randomized trials comparing the effects of consuming low energy dense preloads as part of a dietary weight loss intervention in free-living obese adults. The study is unique because: 1) we utilized solid and liquid forms of a fruit preload that were matched for weight, energy, water contents, and thus, energy density; 2) GF and GFJ preloads were compared to a water preload matched by weight (127 g) since the composition of GF and GFJ is ~91% water; and 3) preloads were ingested 20 minutes before meals to avoid potential confounding effects of orogastrointestinal satiety signaling . Thus, any differential responses to the preload strategy would result from the higher fiber content of GF or higher bioflavonoid content of GFJ.
Under these conditions, dietary energy density reduced 20-28% and total energy intakes decreased 21-29% after preloads were incorporated into the meal plan. Interestingly, reduced energy intakes were not associated with higher VAS ratings of hunger, indicating that subjects remained satiated . If the amount (grams) of food consumed is a determinant of hunger , the lack of perceived hunger may be explained by the consistent amount of food consumed throughout the study. It is intriguing that subjects not only adjusted the total amount of their food intakes to incorporate the amount of the preloads, but also compensated for the energy content of the preloads by decreasing energy intakes from meals and snacks to achieve an overall reduction in total energy intakes.
The reduction in dietary energy density and energy intakes achieved represents an 8.5-16.5% (~250-500 kcal/d) greater reduction in calories consumed during the caloric restriction + preload phase than the 12.5% reduction prescribed during the caloric restriction phase. This finding is consistent with other community-based interventions in which consumption of a low energy dense diet has led to substantial reductions in energy intakes and body weight [37–40]. In the present study, while the overall weight loss of 7.1% of initial body weight was not statistically different among groups, weight loss was clinically meaningful based on current consensus that 5-10% weight loss decreases cardiometabolic risk .
Notably, the additional 8.5-16.5% reduction in energy intakes during the caloric restriction + preload phase was physiologically consistent with the 13.3% increase in the rate of weight loss during that phase. The compensation observed contrasts with some basic science models of energy balance utilizing the concept of negative adiposity feedback signaling to the brain [42–44] and data suggesting that obese individuals would defend adiposity and compensate for weight loss by increasing intakes of energy dense foods or total calories [42, 44, 45]. Nevertheless, our findings are consistent with the ability of individuals at lower BMI to respond to the energy content of an ingested preload . That our obese subjects exhibited such a response in the setting of negative energy balance and weight loss suggests that utilization of a low energy dense preload may fundamentally influence mechanisms involved in energy homeostasis [37, 46].
The present data indicate that preload weight and low energy density, not form (solid vs liquid), fiber or bioflavonoid content promoted the greater reductions in dietary energy density, total energy intakes, and body weight. While this contrasts with laboratory-based experiments that show differential effects on energy intakes at a meal based on the physical form of food [7, 8, 12, 13, 47, 48], it is consistent with data outside of the lab setting where subjects who logged 24-hour food diaries showing no differences in total energy intakes when consuming solid and liquid preloads of several different food items .
It is also intriguing that the water preload was equally efficacious for reducing energy intakes and body weight. This finding also suggests that it was preload consumption that affected dietary energy density and total energy intakes during the caloric restriction + preload phase. Since water adds weight (and volume) without energy, increasing the amount of water in a food or beverage item is a common method for manipulating energy density [36, 49–51] and incorporating water into beverage, soup and casserole preloads has reduced subsequent lunch meal energy intakes by 7-20% [5, 6, 50]. In addition, subjects who drank water with breakfast reported less hunger and greater satiety , and when drinking water replaced caloric beverages energy intakes decreased and subjects lost weight .
Though improvements in insulin sensitivity and lipoprotein profile frequently occur during weight loss , we did not detect significant changes in glucose, insulin, or total and LDL-cholesterol. A plausible explanation is that the small changes observed reflect a low level of insulin resistance in these relatively healthy obese subjects. It is striking that HDL-C levels increased up to 8.2% from baseline in GF and GFJ groups, a significant change compared to decreased HDL-C in the water preload group. Since epidemiological evidence indicates that raising HDL-C by only 1 mg/dL reduces cardiovascular risk by 2-3%, this finding supports earlier evidence of potential anti-atherosclerotic effects of GF or GFJ consumption. A possible explanation for the rise in HDL-C is increased antioxidant activity from greater vitamin C and/or flavonoid (ie, naringin) intakes [55, 56], although oxidative stress was not directly measured in the present study.
While the present study was carefully designed to compare the effects of GF, GFJ and water preloads, limitations are worth considering. In contrast to laboratory-based feeding, it was not possible to blind study RDs and subjects to preload assignment in this community-based dietary intervention. Second, there is no food or beverage that functions as a completely inactive comparator as even water may have metabolic effects under certain conditions [57, 58]. Yet, the high (~91%) water content of GF and GFJ made the water preload an appropriate control for analytic comparisons. Moreover, including water allowed all groups to experience similar behaviors and orogastric sensations while preloading three times daily for 12 weeks. Third, while we acknowledge that obese adults usually underreport energy intakes , the potential for underreporting should be equivalent among subjects as there were no differences by group in baseline BMI . Even so, to compensate for potential bias, RDs conducted unannounced randomly scheduled 24-hr recalls by telephone using validated methods and standardized scripts [61, 62]. Further, subjects were trained to estimate portion sizes using visual aids designed to improve recall accuracy . Additionally, our 24-hr recall and food log data agreed with the expected changes in dietary energy, fiber and vitamin C intakes. These improvements in nutrient intake profiles indicate reliable reporting as well as evidence of high compliance with the dietary protocol.