Effect of Weight Loss via Severe vs Moderate Energy Restriction on Lean Mass and Body Composition Among Postmenopausal Women With Obesity

Key Points Question What are the long-term effects of severe vs moderate energy restriction on lean mass and other aspects of body composition? Findings This randomized clinical trial included 101 postmenopausal women with obesity. At 12 months, participants who had undergone severe energy restriction experienced approximately 2-fold greater weight and fat loss, approximately 1.5 times as much loss of whole-body lean mass (proportional to total weight lost), and approximately 2.5 times as much loss of total hip bone mineral density compared with participants who had undergone moderate energy restriction. Meaning Although severe energy restriction is an effective obesity treatment, caution is necessary when implementing it in postmenopausal women, especially those with osteopenia or osteoporosis.

these changes on body composition, it is vital to assess whether VLEDs have negative effects 48 that could ultimately increase the risk of conditions such as diabetes, cardiovascular disease, 49 sarcopenia and osteoporosis. This question is of particular importance given that individuals may 50 use VLEDs for several months at a time repeatedly over many years in order to manage their 51 weight.

53
The overall aim of this project is to compare the consequences of weight loss via VLED To this end, healthy obese postmenopausal women will be randomized to one of two weight loss 59 interventions; one consisting of at least 12 months of lifestyle management with moderate energy 60 restriction (~500 kcal deficit per day, 'conventional diet'), and the other consisting of 3-5 months 61 on a VLED followed by conventional diet or partial meal replacement for a total of at least 12 62 months.

64
Participants will be randomised to VLED or conventional diet. Both interventions will be 115 administered at the Metabolism & Obesity Services and overseen by Professor Ian Caterson and 116 Dr Tania Markovic and Dr Janet Franklin. By having the interventions administered by actual 117 clinical services rather than by our research laboratory, our findings will reflect effects of each 118 diet as administered in the real world. We already know that both VLED and conventional diet 119 result in average weight losses of approximately 20% and 12% of initial body weight after 12 120 months, respectively, and this study will demonstrate the effects of such real world clinical 121 interventions on important parameters of long-term health and wellbeing, notably body 122 composition and the factors that influence it.

124
The VLED consists of a 2,100 kJ (~500 kcal) per day program (Optifast®, Nestlé Nutritionals, 125 Vevey, Switzerland), a high quality VLED with appropriate micronutrients that has been used 126 extensively in the Metabolism & Obesity Services. Participants in the VLED arm will follow the 127 VLED continuously for at least 3 months and for up to 5 months, or until a BMI of < 25 kg.m 2 is 128 reached. As starting BMI is 30-40 kg.m 2 , some volunteers on VLED may reach their target 129 weight within 5 months. All volunteers on the VLED will then switch to either partial meal 130 replacement for further weight loss if required (~1,500 kcal per day), or lifestyle intervention for 131 weight maintenance. Meal replacements such as Optifast® are routinely used in this way, and 132 many patients favour such ongoing usage. Two to three months on VLED is known to induce 133 endocrine changes that have been shown in turn to adversely affect fat distribution, muscle 134 function or bone density within 3 months as outlined above, and so using VLED for up to 5 135 months will be sufficient to reveal any effects on body composition, if present. The conventional 136 diet is a lifestyle management program consisting of moderate kilojoule restriction (~500 kcal 137 deficit per day) that has also been used extensively at the Metabolism & Obesity Services and 138 published by Professor Ian Caterson 51 . Participants in both interventions will be advised to use 139 pedometers and, after determination of baseline daily step counts, will be encouraged to 140 gradually increase their physical activity in increments of 15 minutes per day of moderate 141 physical activity such as walking (~1,500 steps per day). very large (>1.1), indicating that a sample of only 5 per group will provide 90% power to detect 164 differences between the diets at P<0.05. At later time points, differences between diets with 165 respect to body weight and parameters of body composition such as fat free mass will be smaller, 166 but as described at the start of this section our sample of 50 participants on each diet provides 167 sufficient power to detect this. We have access to gold-standard techniques to measure each of 168 the outcome variables in this project 50,52-55 , thereby minimizing variability and further 169 contributing to our power to detect clinically meaningful differences between diets. Rationale. With the increasing use of VLEDs, it is important to compare the effects of ketogenic 175 VLED and conventional diet directly on the adaptive responses to weight loss, since these 176 parameters can significantly influence body composition and metabolic health. A suite of 177 parameters will be investigated to identify possible mechanisms by which VLED and 178 conventional diet may induce differential effects on body composition and muscle strength, 179 thereby highlighting possible pathways for amelioration of any such differences. Additionally, 180 these physiological responses will provide insights into reasons for lack of weight loss in some

194
Time points and matching between groups. At 0, 0.25, 1, 3, 6 and 12 months after in ketosis due to its low carbohydrate and kilojoule content), participants will switch to a 222 maintenance diet for 1 week followed by re-measurement at t = 3.25 months. Fasting serum 223 concentrations of ketones such as -hydroxybutyrate will also be determined before and after this 224 transition.

226
Appetite. Participants will be provided with a fixed (200 kcal) liquid meal (Optifast®, Nestlé 227 Nutritionals, Vevey, Switzerland) at 9:00 am. Subjective measures of motivation to eat will be 228 measured immediately before, after and at 1, 2 and 3 hours thereafter using the Electronic 229 Appetite Rating System as described by Professor Neil King 59 .

231
Physical activity. Intensity and duration of daily physical activity will be measured using the 232 Actigraph triaxial accelerometer worn on the waistband or on a belt. Accelerometry data will be 233 collected during waking hours in the 7 days preceding each time point as previously described by 234 Professor Nuala Byrne. Descriptive data will also be recorded in an activity diary, which will 235 enable understanding of reasons for taking off the accelerometer (e.g. to swim or to sleep).

236
Although all subjects will have the same standardized prescription to encourage physical 237 activity, accelerometry and activity diary data combined with information from questionnaires 238 (below) will provide insight as to possible effects of VLED or conventional diet on time spent 239 exercising, in non-exercise activities or inactivity, and subjective measures of sleepiness.

241
Resting metabolic rate will be measured in the fasted state via indirect calorimetry using the hormones that could adversely affect body composition, as well as the duration of any such 258 changes. A suite of hormones will be measured, as it will allow a more accurate assessment of 259 whether the activity of each axis has been differentially altered by either diet.

261
Bone. The following parameters will be assessed to determine possible mechanisms by which  Paired T-tests will be used to determine whether switching from the ketogenic VLED to the non-294 ketogenic maintenance diet alters the magnitude of any of the adaptive responses to weight loss.      Hospital routinely use VLEDs for periods of 4 months.

603
 Three additional inclusion criteria and a screening step were added. For safety, 604 participants must not have any contraindications for following a VLED, and they must 605 not have donated whole blood in the 3 months preceding the trial. As this trial is labour 606 intensive for both investigators and participants, we also needed to ensure that potential 607 participants would be capable of completing activities required for this trial, such as 608 keeping food, activity and sleep diaries at home, wearing accelerometers and temperature 609 trackers for 7 days at a time, and keeping appointments with our team. Thus, after signing 610 the consent form, potential participants would be required to complete these screening 611 activities prior to being randomised for the trial. Hospital, because an independent study has shown both to be the most nutritionally 623 complete VLED products available. Prima Health Solutions was able to offer their 624 product, KicStart TM , at a lower cost than Nestlé. in urine and whole blood are more common than serum measurements.