Weight Loss Strategies for Obese Adults: Personalized Weight Management Program vs. Standard Care

Authors


School of Human Movement Studies, Queensland University of Technology, Victoria Park Road, Kelvin Grove, Q4059 Brisbane, Queensland, Australia. E-mail: n.byrne@qut.edu.au

Abstract

Objective: The objective of this study was to evaluate the effect of a 32-week personalized Polar weight management program (PWMP) compared with standard care (SC) on body weight, body composition, waist circumference, and cardiorespiratory fitness in overweight or obese adults.

Research Methods and Procedures: Overweight or obese (29 ± 2 kg/m2) men and women (n = 74) 38 ± 5 years of age were randomly assigned into either PWMP (men = 20, women = 21) or SC (men = 15, women = 18). Both groups managed their own diet and exercise program after receiving the same standardized nutrition and physical activity advice. PWMP also received a weight management system with literature to enable the design of a personalized diet and exercise weight loss program. Body weight and body composition, waist circumference, and cardiorespiratory fitness were measured at weeks 0, 16, and 32.

Results: Eighty percent of participants completed the 32-week intervention, with a greater proportion of the dropouts being women (PWMP: 2 men vs. 7 women; SC: 2 men vs. 4 women). At 32 weeks, PWMP completers had significantly (p < 0.001) greater losses in body weight [6.2 ± 3.4 vs. 2.6 ± 3.6 (standard deviation) kg], fat mass (5.9 ± 3.4 vs. 2.2 ± 3.6 kg), and waist circumference (4.4 ± 4.5 vs. 1.0 ± 3.6 cm). Weight loss and fat loss were explained by the exercise energy expenditure completed and not by weekly exercise duration.

Discussion: More effective weight loss was achieved after treatment with the PWMP compared with SC. The results suggest that the PWMP enables effective weight loss through tools that support self-monitoring without the requirement of more costly approaches to program supervision.

Introduction

The epidemic of obesity coupled with poor success of cost-effective weight loss treatments has prompted the call for further research to determine viable treatment options (1). Physical activity is important but is not the sole determinant of successful weight loss (2). Data from the Diabetes Prevention Program revealed that both exercise and dietary restraint facilitate weight loss and its long-term maintenance (3). The frequency of monitoring dietary intake was related to success at achieving the physical activity goal (minimum of 150 minutes moderate exercise per week), suggesting that adherence to one aspect of the intervention is related to adherence in other aspects. Likewise, success at achieving the physical activity goal was related to success at achieving the weight loss goal (7% initial body weight) in ∼24 weeks and also in maintaining the reduced-weight for an average 3.2 years (3). These results support the notion that successful weight loss is enabled when behavior changes cluster; with participants who adhere to one aspect of the lifestyle regimen more likely to adhere to other aspects (4, 5). Such findings suggest the need for weight loss interventions that link strategies for increasing physical activity levels with monitoring dietary intake. Importantly, however, these studies included considerable support networks of research staff to educate, prescribe, and monitor the intervention for the participants. For instance, in obese children/adolescents, biweekly phone contact has been shown to mediate weight maintenance after an intensive live-in weight loss program (6). While effective, there is a significant economic burden in such approaches.

In contrast to research studies or professionally managed treatments, some individuals have successful weight loss through self-management strategies. The National Weight Control Registry surveyed behaviors of >1000 individuals who self-reported successful weight loss and maintenance over a number of years (7). Approximately 45% of registrants had lost weight on their own, and as shown in other studies (8, 9, 10), the majority of individuals reported using behavioral strategies such as lowering dietary fat, increasing physical activity or structured exercise programs, weekly weighing, and self-monitoring of food intake. Self-help strategies purportedly have the additional advantage of enabling individuals to obtain a sense of power and the inward resources that give them more control over themselves and their environment (11). However, despite these successes, some individuals who have lost weight and attempt to maintain the weight-reduced state on their own express the need for programs with ongoing support offered at low or no cost (12). For these individuals, there is the need for tools, in addition to currently available resources, to assist in learning how to self-manage weight loss and long-term weight maintenance.

This study was designed to evaluate whether a 32-week personalized weight management program (PWMP)1 would enhance weight loss compared with standard care (SC). It was hypothesized that, compared with individuals given standard diet and physical activity advice, overweight/obese individuals given exercise energy expenditure targets and heart rate (HR) and dietary intake monitoring are better able to achieve a self-defined target weight loss. A further purpose of the study was to determine whether body size, body composition, and cardiorespiratory fitness improvements gained over the first 16 weeks of intervention, which included weekly contact with research staff, could be maintained over the second 16 weeks when there was no contact with study participants. A final aim in the PWMP group, where objective exercise training data were available, was to determine whether adherence to the exercise prescription predicted degree of change in body composition and success in meeting a self-defined target weight loss.

Research Methods and Procedures

Participants

Seventy-four participants were recruited for the study from the Brisbane metropolitan area through the public media (e.g., local radio, newspapers) and within the University (Figure 1). Respondents were initially screened by telephone, and if eligible based on self-reported height, weight, and medical history, were invited to attend an information session. Eligibility criteria were 30 to 45 years of age, overweight/obese (BMI = 27 to 32 kg/m2), sedentary (defined as <30 minutes of intentional moderate physical activity per week over the past 12 months, including work-related physical activity), weight stable (±2 kg) for the last 6 months, and ambulatory. Exclusion criteria were medications known to affect HR or body composition, pregnancy or lactation, planning to get pregnant during the intervention period, post-menopausal, and smoking. Additional exclusion criteria included the inability to walk on a treadmill or to undertake a graded test to achieve maximum aerobic capacity. Only those respondents who identified that they were in the “ready to change” stage according to the Transtheoretical Model (13) were included. Furthermore, participants categorized as overweight or obese but otherwise healthy gained clearance from their medical practitioner before enrollment in the study. The study was approved by the University Human Research Ethics Committee, and all participants gave written informed consent.

Figure 1.

Flow of participants in the study.

Interventions

After baseline testing, participants were randomly assigned to either of two intervention groups: PWMP or SC. Participants were stratified according to sex (men and women), age (30 to 37 and 38 to 45 years), and BMI (27 to 29.9 and 30 to 32 kg/m2) and were randomized to enable the intervention groups to be matched as closely as possible. Unfortunately, there were not enough volunteers in the BMI range willing to be randomized into the standard care group to enable a balanced research design; time restrictions precluded further recruitment of participants. However, as shown in Table 1, although the groups were not equal in number, there was no difference in outcomes variables at baseline.

Table 1.  Baseline characteristics of participants by randomization group and sex
   PWMP (N = 41)SC (N = 33)
 PWMP (N = 41)SC (N = 33)Men (N = 20)Women (N = 21)Men (N = 15)Women (N = 18)
  1. PWMP, personalized weight management program; SC, standard care; V̇o2max, maximum oxygen uptake; FFM, fat-free mass. Data presented as mean (standard deviation).

Age (years)37.6 (4.4)38.6 (4.8)37.4 (4.5)37.9 (4.5)39.8 (5.3)37.5 (4.3)
Height (m)1.71 (0.10)1.69 (0.10)1.80 (0.10)1.63 (0.10)1.78 (0.10)1.62 (0.10)
Weight (kg)87.2 (12.6)85.7 (12.5)97.9 (8.1)79.2 (7.4)96.7 (11.1)78.8 (8.4)
BMI (kg/m2)29.3 (1.6)29.3 (1.8)29.3 (1.6)29.3 (1.8)29.3 (1.6)29.3 (1.8)
Body fat (%)35.5 (6.0)36.4 (2.1)30.7 (4.0)40.0 (3.4)30.6 (5.2)41.2 (4.2)
Waist (cm)90.4 (8.7)91.4 (9.2)97.0 (5.2)84.0 (6.3)98.6 (5.9)85.4 (6.8)
o2max (mL/kg/min)37.6 (7.1)35.4 (7.8)43.4 (4.8)32.4 (4.1)39.6 (7.6)31.8 (7.3)
o2max (mL/kg FFM/min)58.6 (7.0)55.5 (7.1)62.8 (7.0)54.3 (7.1)56.8 (7.0)54.2 (7.0)

SC

Participants received standard care advice on weight management from a health professional with dual qualifications in dietetics and exercise physiology. The single consultation included simple advice regarding ways to increase physical activity levels and reduce energy intake, with clear directions that the goal should be to lose no >1 kg per week. This maximum was important to parallel the average target rate of weight loss selected by the PWMP group. Guidelines for physical activity participation were consistent with the Australian National Physical Activity Guidelines (14) for safe exercise by relatively unfit but otherwise healthy individuals. Nutrition advice was consistent with an ad libitum low-fat diet. Each week during the first 16 weeks of the intervention, participants reported to the University for the measurement of body weight and waist circumference.

PWMP

In addition to the standard care treatment outlined above, these participants were provided with an HR transmitter belt, a receiver watch with built in weight management program, and weight management program education resources (Polar Electro Oy, Kemple, Finland). These resources included an explanatory booklet with an overview of the program, users manual, diet diary, tape measure, and “calorie counting” book (15) that provides a comprehensive analysis of the energy content of most of the foods and ingredients available in Australia. The program calculates a weight loss (or maintenance) program, including the recommendation for daily energy intake and weekly exercise energy expenditure (defined in kilocalories). These individualized goals are based on information provided by each subject (body weight, height, age, sex, target weight, daily occupational activity assessment) and also on the chosen exercise intensity (low: 110 to 130 beats/min; moderate: 130 to 160 beats/min). The program allows the user to follow his/her progress daily in an electronic diary by tracking body weight, energy intake, and exercise energy expenditure (whenever exercising with the transmitter belt). Additionally, the program provides weekly updates and makes adjustments according to the progress of the individual by giving updated daily energy intake and weekly exercise energy expenditure values. Because the intention of the study was to investigate the effectiveness of the PWMP in the form that it would be provided to the general public, no additional requirement to complete exercise or diet diaries was deemed appropriate.

Participants were introduced to the PWMP in a training session to ensure comprehension of all functions. In the first few weeks of the intervention, participants were encouraged to e-mail a member of the research team if they encountered any problems with the program or needed assistance to follow instructions. As did the SC group, PWMP participants reported to the University Clinic each week for the first 16 weeks of the study for the measurement of body weight and waist circumference. For the following 16 weeks, participants had no contact with the study team. All baseline measurements were repeated after Weeks 16 and 32 of the intervention.

Experimental Protocol

All participants attended two testing sessions at baseline and after Weeks 16 and 32 of the intervention. On both testing days, participants reported to the University laboratory a minimum of 3 hours after their last food or fluid intake, wearing light-weight, comfortable clothing, having abstained from physical exercise and consumption of caffeine or alcohol in the previous 12 hours. The first session involved anthropometric measurements (height, body weight, and waist circumference) and body composition assessment by DXA (DPX-L; Lunar Radiation Corp., Madison, WI). The scans were analyzed with the ADULT software, version 1.33 (Lunar Radiation Corp.), and body composition was determined as previously reported (16). Measurements of body height (stretch stature) were taken using a Harpenden stadiometer, body weight was recorded on a digital scale, and BMI (weight divided height squared) was calculated from height and weight values. Waist circumference was measured at the narrowest point below the inferior border of the last ribs and above the iliac crest. Subsequently, participants were familiarized with the maximal aerobic power test protocol, breathing apparatus, and treadmill. Participants were familiarized with walking at 5.6 kph, the optimal walking and running speeds for the maximal graded exercise test were identified, and participants were familiarized with the Borg 6–20 scale for the rating of perceived exertion (17, 18).

On the second testing day, participants undertook a graded exercise test to assess maximal aerobic capacity. After being fitted with a Hans-Rudolf headset (with two-way breathing valve and pneumotach), a nose clip, and a Polar Coded Transmitter (Polar Electro Oy), participants walked for 4 minutes at 0% grade and 5.6 kph and the speed increased to an individualized speed, after which grade increased by 2.5% per minute until volitional exhaustion. Respiratory gases were collected throughout the test using a Q-PLEX Gas Analysis System (Quinton Instrument Co., Seattle, WA). The O2 and CO2 analysers were calibrated before each test against known gas concentrations and the flowmeter calibrated against a 3.0-liter syringe. HR and respiratory gases were averaged for the last 30 seconds of each stage and the highest average value for 30 seconds (provided respiratory exchange ratio ≥1.10) was recorded as the peak value. Duplicate 0.5-mL samples of capillary blood obtained by the finger-prick method were collected immediately at the end of each test. Blood lactate concentrations were subsequently analyzed by an ultraviolet endpoint method using the spectrophotometric assay procedure (19). To confirm that the peak values obtained were representative of maximal capacity, a booster test was used after the participants had rested for 5 minutes. A “booster” test is when the individual undertaking a test to volitional exhaustion is required, after a short rest, to resume exercising at the mechanical work they stopped at, and continue the test with the usual mechanical work increases until again reaching a point of volitional exhaustion. Five commonly used markers of maximal cardiorespiratory capacity were measured, and the data were accepted provided three of the following were achieved in the main test: oxygen uptake (V̇o2) increase <0.8 mL/kg per minute per percent grade in last stage; respiratory exchange ratio ≥1.10; HR greater than predicted maximum (220 – age) −10 beats/min; blood lactate >8 mM; and rating of perceived exertion ≥18.

Statistical Analyses

To determine the sample size required to detect a difference between two interventions, the power calculations were based on prior observations of modest weight losses through diet and exercise interventions (6% and 7% initial weight, respectively) in obese women (20). The within-group changes in body weight derived from a dietary intervention and an exercise intervention over 14 weeks were 5.2 ± 1.2 and 6.1 ± 1.2 kg, respectively. To detect this magnitude of between-group difference as statistically significant with α set at 0.05 and at a power of 0.80, 24 participants per group were required.

All statistical calculations were performed using SAS (version 8.02) with p < 0.05 considered significant. Data are presented as mean ± standard deviation or mean ± standard error as specified, and 95% confidence intervals for the mean group differences were calculated. Differences between characteristics of completers vs. dropouts were tested using independent t tests. When comparing treatment effects, data were analyzed using both completers-only data and intention-to-treat (ITT) analysis approaches. For the completers analyses, only data from the participants (N = 59) who completed the 32-week intervention was included. ITT analysis included all randomized participants (N = 74) with missing values imputed by last-observation-carried-forward method. Comparisons of outcome variables before and after 16 and 32 weeks of the intervention were assessed after adjusting for sex, using 2 (group) × 3 (time) repeated-measures analysis of covariance. When significant F ratios were obtained, Bonferroni post hoc tests were performed to locate differences among means.

Results

The baseline characteristics of the study sample are outlined in Table 1. As expected, compared with women of the same BMI, men had significantly lower percentage body fat and larger waist circumference. There were no differences between groups for any of the primary outcome variables at baseline, and no within-sex differences in body size or composition were noted between groups. Of the starting cohort, 80% of the participants completed the study (Figure 1). The retention rate was comparable between groups (PWMP = 78%, SC = 81%), but greater for men (89%) than women (72%). After adjustment for sex, examination of the average baseline characteristics for the completers (N = 59) compared with dropouts (N = 15) showed no significant difference in body weight (87.0 vs. 89.5 kg; p = 0.36); waist circumference (91.6 vs. 89.7 cm; p = 0.33); fat mass (30.4 vs. 32.3 kg; p = 0.51); fat-free mass (56.4 vs. 57.0 kg; p = 0.73); relative body fat (35.5% vs. 36.3%; p = 0.54); or V̇o2max (37.1 vs. 36.0 mL/kg per minute; p = 0.53). Ninety percent of the participants attended at least 80% of the weekly weigh-in sessions during the first 16 weeks of the study. There was no difference between groups in the attendance rate and no relationship between weight loss and frequency of attendance. Reasons for dropping out of the study were similar between groups; including personal reasons, work commitments, and unexpected requirement to travel overseas. However, two of the participants dropped out in week 2 because of being disappointed with being randomized into SC.

Table 2 outlines group and sex comparisons of changes in body weight, waist circumference, body composition, and maximal aerobic power from ITT analysis. Table 3 shows the same analysis, but using only data from participants who completed the 32-week intervention. The results of the two analysis approaches are comparable, with the magnitude of intervention effects defined by the ITT analyses being more modest. In terms of the participants who completed the 32-week study, at 32 weeks, PWMP had lost 3.5 kg more weight, 3.7 kg more fat mass, and 3.3 cm more waist circumference than SC. However the improvement in cardiorespiratory fitness was not appreciably greater in PWMP than SC.

Table 2.  ITT analysis of changes [mean (standard error)] in weight, waist, fat mass, and V̇o2max
 PWMP (n = 41)SC (n = 33)Group effect (p value)Difference between means95% confidence limits for group difference
  • ITT, intention to treat; PWMP, personalized weight management program; SC, standard care; FFM, fat-free mass; V̇o2max, maximum oxygen uptake; NS, not significant.

  • *

    Significantly different (p < 0.05) from Week 0 to Week 16 within the same intervention group.

  • Significantly different (p < 0.05) from Week 0 to Week 32 within the same intervention group.

Weight (kg)      
 Week 0 to Week 16−4.46 (0.5)−2.35 (0.6)NS2.11−4.520.30
 Week 16 to Week 32−0.39 (0.5)*0.12 (0.6)NS0.51−2.921.90
 Week 0 to Week 32−4.84 (0.5)−2.19 (0.6)<0.052.65−5.06−0.24
Waist (cm)      
 Week 0 to Week 16−3.15 (0.6)−1.28 (0.7)NS1.86−4.620.90
 Week 16 to Week 32−0.31 (0.6)*0.37 (0.7)NS0.68−3.432.08
 Week 0 to Week 32−3.44 (0.6)−0.86 (0.7)0.092.58−5.340.18
Fat mass (kg)      
 Week 0 to Week 16−3.99 (0.5)−2.12 (0.6)NS1.87−4.210.46
 Week 16 to Week 32−0.65 (0.5)*0.21 (0.6)NS0.85−3.181.48
 Week 0 to Week 32−4.63 (0.5)−1.87 (0.6)<0.012.76−5.09−0.42
o2max (mL/kg/min)      
 Week 0 to Week 163.14 (0.6)2.68 (0.7)NS0.46−2.143.05
 Week 16 to Week 320.08 (0.6)*−0.74 (0.7)*NS0.82−1.773.42
 Week 0 to Week 323.21 (0.6)1.92 (0.7)NS1.29−1.313.88
o2max (mL/kg FFM/min)      
 Week 0 to Week 161.90 (0.7)2.92 (0.8)NS1.02−2.054.09
 Week 16 to Week 320.61 (0.7)−1.22 (0.8)*NS0.62−3.682.45
 Week 0 to Week 321.29 (0.7)1.70 (0.8)NS0.41−2.663.47
Table 3.  Completers analysis of changes [mean (standard error)] in weight, waist, fat mass, and V̇o2max
 PWMP (n = 32)SC (n = 27)Group effect (p value)Difference between means95% confidence limits for group difference
  • o2max, maximum oxygen uptake; PWMP, personalized weight management program; SC, standard care; FFM, fat-free mass; NS, not significant.

  • *

    Significantly different (p < 0.05) from Week 0 to Week 16 within the same intervention group.

  • Significantly different (p < 0.05) from Week 0 to Week 32 within the same intervention group.

Weight (kg)      
 Week 0 to Week 16−5.66 (0.6)−2.82 (0.7)<0.052.84−5.56−0.12
 Week 16 to Week 32−0.45 (0.6)*0.19 (0.7)*NS0.64−3.362.07
 Week 0 to Week 32−6.15 (0.6)−2.3 (0.7)<0.013.53−6.24−0.81
Waist (cm)      
 Week 0 to Week 16−3.91 (0.8)−1.52 (0.7)NS2.39−5.760.97
 Week 16 to Week 32−0.31 (0.8)*0.50 (0.7)NS0.81−4.172.56
 Week 0 to Week 32−4.29 (0.8)−1.00 (0.7)0.063.29−6.650.08
Fat mass (kg)      
 Week 0 to Week 16−5.08 (0.6)−2.55 (0.7)0.072.53−5.180.12
 Week 16 to Week 32−0.79 (0.6)*0.29 (0.7)*NS1.09−3.741.56
 Week 0 to Week 32−5.89 (0.6)−2.25 (0.7)<0.0013.65−6.30−1.00
o2max (mL/kg/min)      
 Week 0 to Week 164.03 (0.7)3.26 (0.8)NS0.77−2.333.87
 Week 16 to Week 320.10 (0.7)*−0.93 (0.8)*NS1.03−2.074.13
 Week 0 to Week 324.12 (0.7)2.33 (0.8)NS1.79−1.314.89
o2max (mL/kg FFM/min)      
 Week 0 to Week 162.41 (0.9)3.57 (0.9)NS1.16−4.952.64
 Week 16 to Week 320.69 (0.9)−1.49 (0.9)*NS0.81−2.994.60
 Week 0 to Week 321.69 (0.9)2.08 (0.9)NS0.39−4.193.40

As shown in Figure 2 and Table 3, the average weight loss from baseline was significantly greater for PWMP compared with SC both at Week 16 and Week 32 for the completers analysis; but only at Week 32 for the ITT analysis. The weight loss at Week 32 equated with a 7.1% and 3.1% decrease from baseline values for PWMP and SC, respectively. The self-defined weight loss “target” was comparable for both groups: 8.9 ± 1.2 and 8.6 ± 1.2 kg for PWMP and SC, respectively. PWMP achieved a greater proportion of the weight loss goal than SC both at Week 16 (60 ± 36% vs. 34 ± 41%; p = 0.01) and Week 32 (70 ± 51% vs. 30 ± 51%; p = 0.008). As shown in Figure 3, at Week 16, the 50th percentile for SC equated with 26% of target weight loss, whereas for PWMP, the median was 72%. At Week 32, the 50th percentile for SC had dropped to 18% of target weight loss, whereas the PWMP median remained at 72%. A weight loss of ≥5% initial body weight is often referenced as an important clinical outcome. More of those in PWMP than SC achieved this relative weight loss by Week 16 (63% and 30%, respectively); these proportions were comparable at Week 32 (59% PWMP and 33% SC).

Figure 2.

Mean weight and waist change by group (error bars indicate standard error). Between-group difference (* p < 0.05; ** p < 0.01; § p = 0.06).

Figure 3.

Cumulative frequency distribution of the target weight loss achieved by Week 16 (top) and Week 32 (bottom).

While fat loss was significant for both groups, the loss from baseline was significantly greater at both Week 16 and Week 32 for PWMP compared with SC (p < 0.01). Similarly, waist circumference decrease from baseline was greater on average for PWMP compared with SC at both Week 16 and Week 32 (Figure 2). The average increase in V̇o2max from baseline for PWMP did not differ significantly from SC at Week 16 or Week 32 (p > 0.1). For all measures, the change from baseline for both groups was significant at both Week 16 and Week 32, but changes between Week 16 and Week 32 were not significant.

Table 4 outlines the exercise prescription goals and the exercise completed by the PWMP group. Despite a wide range of prescribed exercise energy expenditure being completed (28% to 145%), the PWMP, on average, achieved two thirds of the energy expenditure target over an average 2.8 h/wk (range: 1.3 to 5.4 h/wk) and at a mean intensity of 72% maximum HR (HRmax; 61% to 88% HRmax). There was no difference between men and women in the average weekly exercise training completed. Pearson correlation analyses showed that duration of weekly exercise completed did not explain the variance in weight loss (r = 0.21; 95% confidence interval, −0.21 to 0.57; p = 0.32) or fat loss (r = 0.21; 95% confidence interval, −0.22 to 0.57; p = 0.34). In contrast, Figure 4 shows the relationship between weight change and average weekly exercise energy expenditure completed. Based on the resulting prediction equation, even with no reduction in energy intake, the average target weight loss (8.9 kg) would have been achieved with an average exercise energy expenditure of 2341 kcal/wk. Furthermore, after adjusting for sex, the proportion of target exercise energy expenditure completed by PWMP participants was positively related to weight loss (r = 0.55, p < 0.05) and fat loss (r = 0.49, p < 0.05).

Table 4.  PWMP completers exercise prescription and training data [mean (standard deviation)]
VariableTotal PWMP (n = 32)Men (n = 18)Women (n = 14)Sex difference95% confidence limits for sex difference
  • PWMP, personalized weight management program; %HRmax, percentage of maximum heart rate.

  • *

    p < 0.001.

Target weight loss (kg)8.9 (1.2)9.7 (0.8)7.8 (0.7)1.9*1.33.5
Actual weight loss (kg)6.2 (3.7)7.3 (3.2)5.7 (3.9)1.61.64.7
Proportion of target weight loss (%)73 (40)75 (31)71 (50)4−3239
Target exercise energy expenditure (kcal/wk)2341 (519)2796 (234)1846 (117)950*7871114
Target exercise energy expenditure (kcal/kg/wk)26.2 (3.3)28.9 (1.4)23.1 (1.5)5.7*4.57.0
Actual exercise energy expenditure (kcal/wk)1509 (554)1637 (562)1369 (536)268−209745
Actual exercise energy expenditure (kcal/kg/wk)16.9 (5.5)16.9 (5.4)17.0 (6.0)0.1−5.04.9
Proportion target exercise energy expenditure (%)66 (31)57 (21)74 (39)18−4518
Exercise time (h/wk)2.8 (1.0)2.5 (0.7)3.2 (1.3)0.7−1.50.2
Exercise heart rate (beats/min)132 (13)134 (15)131 (11)2.7−1014
Exercise heart rate (%HRmax)72 (7)72 (8)73 (6)0.7−76
Figure 4.

Weight change at 32 weeks relative to average weekly exercise energy expenditure.

Stepwise and best-subsets regression analyses were used to determine which variables best explained the variance in average weekly exercise energy expenditure. Average weekly exercise energy expenditure was best explained (r = 0.92, p < 0.0001) by a combination of weekly exercise duration and exercise intensity (HR or %HRmax); addition of sex explained a further 6% of the variance.

Discussion

These study results support the hypothesis that overweight and obese individuals given exercise energy expenditure targets, HR monitoring, and calorie counting monitors are better able to achieve a self-defined target weight loss than those given standard diet and physical activity advice. The PWMP group achieved twice the weight loss of SC in both relative and absolute terms. Furthermore, the PWMP group achieved, on average, twice the proportion of the individualized weight loss goal. As expected, participants in the PWMP group achieved a greater reduction in absolute and percentage body fat compared with SC and significantly greater reduction in waist circumference.

The proportion of weight lost by PWMP is clinically meaningful (21). Previous weight loss studies have shown that waist circumference changes are matched with changes in intra-abdominal fat. While Irwin et al. (22) found a 2-kg weight loss in women was accompanied by a 7% reduction in visceral fat, Slentz et al. (23) also noted a 2-kg weight loss was accompanied by a 7% reduction in visceral fat, and Ross et al. (20) reported a 6-kg exercise-induced weight loss was accompanied by a 6.5-cm reduction in waist circumference and a 30% reduction in visceral fat. Based on the dose–response relationships between weight loss, waist circumference reduction, and visceral fat loss shown in these previous studies, it can be speculated that the current 32-week intervention resulted in greater losses in visceral fat for PWMP compared with SC. Thus, the greater weight loss for PWMP was accompanied by greater fat loss and importantly greater reduction in abdominal girth than SC. Collectively, these findings show that weight loss over 32 weeks is possible with both SC and PWMP; however, the additional support provided by PWMP augments the favorable changes achieved with SC. These findings hold both in analysis of “completers” data and ITT analyses.

The second purpose of the study was to determine whether improvements gained over the first 16 weeks of the intervention, which included weekly contact with research staff, could be maintained over the second 16 weeks when no contact was made with study participants. The results showed that, although the changes elicited in the second phase of the intervention were not significantly different between groups, the pattern of change tended to reflect further reductions in body weight and body composition for the PWMP group. The SC group tended toward a reversal in the improvements gained during the first phase. For some individuals, the cessation of the weekly “conscience” meetings may have resulted in lowered motivation to adhere to diet and exercise goals. However, despite no weekly weigh-ins, the PWMP group, in using electronic monitoring of exercise energy expenditure and dietary intake, had continued (albeit non-significant) body size and body composition reductions in the second 16 weeks.

In a study of similar design, Heshka et al. (24) compared weight loss achieved and maintained through self-help strategies with a structured weight loss program. The self-help group was provided with two 20-minute consultations with a dietitian and was given publicly available printed dietary and exercise advice, web sites, and health organization contact details. The individuals randomized to the commercial weight loss group were provided with vouchers entitling them to attend Weight Watchers. The self-help group achieved an average peak weight loss of 1.6 kg (1.7%) after 6 months of the intervention period compared with commercial intervention, which, also peaking at 6 months, achieved an average loss of 5.5 kg (5.8%). Weight losses at 16 weeks were, on average, ∼1.5 (1.6%) and 4.5 kg (4.8%) for the self-help and commercial groups, respectively. These findings suggest that even with continued “conscience” meetings, weight may be slowly regained after 24 weeks.

Recidivism in the months after weight loss elicited through behavior modification interventions are commonplace, and, thus, strategies to assist self-management of weight loss and weight loss maintenance in the longer term are required. The results from this study suggest the PWMP may provide a less costly way to enable individuals to self-monitor their weight loss process using more objective targets. Self-help strategies also theoretically have the additional advantage of enabling individuals to obtain a sense of power and the inward resources that give them more control over themselves and their environment (11). Segal et al. (25) argued that empowerment may be a primary principle underlying self-help goals and that empowerment may, in turn, increase self-efficacy, self-esteem, and the sense that one's own efforts can effect positive change. A longer period of self-management without weekly weigh-ins using PWMP and SC would be required to determine whether the trends we observed in the second 16-week period would reach statistical significance. Further research to determine the impact of PWMP on self-efficacy and self-esteem in relation to capacity to successfully self-manage weight loss and long-term weight management would be valuable.

Treatment efficacy is a measure of the benefit resulting from an intervention for a given health outcome under ideal conditions of a study; as such, it refers to the degree to which the intervention is shown to scientifically accomplish the desired outcome in people who fully comply with the intervention. In contrast, effectiveness is a measure of the benefit resulting from an intervention under usual conditions of clinical care for a particular group; hence, it considers both efficacy of an intervention and its acceptance by those to whom it is offered (26). The data discussed thus far indicate that PWMP was more effective than SC in achieving body size and body composition improvements over 32 weeks. However, the efficacy of PWMP needs to be considered with reference to the extent to which participants adhered to the intervention as prescribed. The third aim of the study was to determine whether adherence to the exercise prescribed predicted magnitude of body composition changes and success in meeting weight loss targets in participants who finished the 32-week intervention.

Participants who finished the 32-week intervention completed, on average, two thirds of the exercise energy expenditure prescribed and achieved approximately three quarters of the target weight loss. It is of interest to note that if the average prescribed exercise energy expenditure had been completed, the average target weight loss would have been achieved (Figure 4). However, it is important to recognize the relatively large standard error of the estimate of this prediction. The average exercise energy expenditure was 1509 kcal/wk (17 kcal/kg per week), and the time committed to structured exercise was, on average, 168 min/wk (range, 78 to 324 min/wk). An important finding was that minutes spent exercising did not explain the variance in weight loss or fat loss. However, absolute and relative exercise energy expenditure completed was positively related with weight and fat mass losses. Exercise energy expenditure was best explained by the combination of exercise duration and exercise intensity. This is an important finding, given that in many clinical and public health settings, exercise recommendations are based around time, with vague exercise intensity prescriptions.

The Centers for Disease Control and Prevention promote population-level recommendations of 150 minutes of exercise of moderate or greater intensity per week (27). These are comparable with the commonly cited 1000 kcal/wk (4184 kJ/wk) additional energy expenditure through purposeful physical activity or exercise recommended for weight reduction (28). On average, participants in the PWMP met both of these recommendations. With no concomitant change to dietary intake, completing 1000 kcal/wk of exercise should result in a weight loss of ∼2 kg over 16 weeks. This rate of weight loss is supported by the studies of Ross et al. (29), in which, for men, a daily exercise energy expenditure of 700 kcal and isocaloric diet over 12 weeks (55,800 kcal) elicited the predicted 7.2-kg weight loss. Similarly, for women, a daily exercise energy expenditure of 500 kcal and isocaloric diet over 14 weeks (49,000 kcal) elicited the predicted 6.3-kg weight loss (20). Based on the relationship between energy deficit and weight loss (3500 kcal energy deficit = 0.45 kg weight loss), the predicted average weight loss of the PWMP in the current study would be 6.3 kg. The actual average weight loss was 6.2 kg.

Another important finding of this study was that, despite men having a higher exercise energy expenditure target, women and men completed the same weekly exercise energy expenditure, and this was undertaken at the same relative intensity. Unlike many other weight loss studies, the weight loss (absolute and relative) did not differ significantly between men and women. Doucet et al. (30) noted a difference in fat mass losses between men and women when participants were on a moderate dietary restriction plus exercise program and reported that the effect could be entirely explained by the difference in the net energy cost of the exercise. The authors suggested that this sex difference was attributable to an exercise intensity effect (31, 32), reporting that, despite the same exercise intensity being prescribed, women exercised at a relative exercise intensity of, on average, 19%, which is lower than for men. In this study, the monitored exercise intensity revealed that men and women exercised as the same physiological intensity, and the resultant weight loss did not differ between men and women.

According to the American College of Sports Medicine, depending on the quantity and quality of training, improvement in V̇o2max ranges from 5% to 30% (33). In this study, the average relative change in V̇o2max (mL/kg per minute) was ∼10%; however, only 66% of the prescribed exercise was completed (PWMP). Thus, although the relative improvements are at the lower end of the expected range, the completed exercise was much less than prescribed in the majority of participants. While genetics explains some of the variance in the magnitude and rate of change in cardiorespiratory fitness with aerobic training, initial V̇o2max may also be a factor (34). The average baseline V̇o2max values for the PWMP men was between the 50th and 60th percentile and for SC men was in the 40th percentile; for women in both groups, the average was between the 30th and 40th percentile (33). However, the range for both sexes was from the 10th to the 90th percentile. Therefore, there was considerable variance within the cohort; and no reason to anticipate that improvement in cardiorespiratory fitness would be limited in either group. Furthermore, it is important to note that the change in V̇o2max was not correlated with baseline V̇o2max (r = 0.12; 95% confidence interval, −0.14 to 0.37; p = 0.37). Similar findings were reported by Skinner et al. (35) in a study in which participants were trained on cycle ergometers for 20 weeks (3 times/wk) for 30 to 50 minutes per session at the HR associated with 55% to 75% V̇o2max measured at baseline.

Data from 633 participants who finished 95% of the required training sessions revealed that ΔV̇o2max was not related to baseline V̇o2max. However, when the baseline values were correlated with relative (%) change in V̇o2max, there was a significant (p < 0.01) correlation coefficient of −0.37. In the current study, no relationship was found between relative change in V̇o2max and baseline values (r = −0.11; 95% confidence interval, −0.36 to 0.15; p = 0.41). Therefore, the changes in V̇o2max are not lower than expected given the training completed; and the magnitude of change was not related to initial fitness level. We would speculate that the modest improvements in cardiorespiratory fitness are attributed to modest training rather than evidence of a ceiling effect.

In conclusion, this study was designed to determine the efficacy and effectiveness of a personalized weight management strategy using biofeedback of exercise intensity and exercise energy expenditure and dietary energy targets compared with generalized standard care. Our ability to more definitively attribute causality of the exercise dose on the weight (or fat) loss is limited by not having objective data on changes in dietary patterns or energy intake. However, despite this limitation, the data, analyzed both by completers and ITT analysis, revealed a strong intervention effect on the magnitude of change in outcome variables. Compared with SC, PWMP showed greater reduction in body weight, waist circumference, and body fat at Week 16 and after a further 16-week period without weekly weigh-in meetings. The study showed that provision of PWMP more than doubled the changes in body weight and body composition achieved with standard nutrition and physical activity advice. However, individuals given the opportunity to self-manage behavior with the aid of PWMP varied in the extent to which they met the exercise energy expenditure and dietary guidelines. While successful weight loss and weight maintenance can be achieved with varying strategies, the PWMP is a tool that more effectively assists overweight and obese individuals to self-manage this process. It remains to be determined whether such self-management strategies will prove effective for longer-term weight management.

Acknowledgments

The authors thank Connie Wishart and Nigel Smith for laboratory assistance during the study. Funding for this study was provided by the Queensland University of Technology Industry Collaborative Grant Scheme and Polar Electro Oy.

Footnotes

  • 1

    Nonstandard abbreviations: PWMP, personalized weight management program; SC, standard care; HR, heart rate; ITT, intention to treat; V̇o2, oxygen uptake; V̇o2max, maximum oxygen uptake; HRmax, maximum HR.

  • The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Ancillary