Differences in Daily Energy Expenditure in Lean and Obese Women: The Role of Posture Allocation

Authors


(Darcy.Johannsen@pbrc.edu)

Abstract

Objective: A low resting metabolic rate (RMR) is considered a risk factor for weight gain and obesity; however, due to the greater fat-free mass (FFM) found in obesity, detecting an impairment in RMR is difficult. The purposes of this study were to determine the RMR in lean and obese women controlling for FFM and investigate activity energy expenditure (AEE) and daily activity patterns in the two groups.

Methods and Procedures: Twenty healthy, non-smoking, pre-menopausal women (10 lean and 10 obese) participated in this 14-day observational study on free-living energy balance. RMR was measured by indirect calorimetry; AEE and total energy expenditure (TEE) were calculated using doubly labeled water (DLW), and activity patterns were investigated using monitors. Body composition including FFM and fat mass (FM) was measured by dual energy X-ray absorptiometry (DXA).

Results: RMR was similar in the obese vs. lean women (1601 ± 109 vs. 1505 ± 109 kcal/day, respectively, P = 0.12, adjusting for FFM and FM). Obese women sat 2.5 h more each day (12.7 ± 3.2 h vs. 10.1 ± 2.0 h, P < 0.05), stood 2 h less (2.7 ± 1.0 h vs. 4.7 ± 2.2 h, P = 0.02) and spent half as much time in activity than lean women (2.6 ± 1.5 h vs. 5.4 ± 1.9 h, P = 0.002).

Discussion: RMR was not lower in the obese women; however, they were more sedentary and expended less energy in activity than the lean women. If the obese women adopted the activity patterns of the lean women, including a modification of posture allocation, an additional 300 kcal could be expended every day.

Introduction

Obesity occurs when there is a chronic disruption in energy balance, i.e., when energy intake exceeds energy expenditure over an extended period of time (1). Low resting metabolic rate (RMR) is considered a risk factor for weight gain leading to obesity (2). For most individuals, RMR comprises 60–80% of total energy expenditure (TEE) (3), thus, low RMR may result in low TEE favoring an imbalance in intake vs. expenditure. Several studies suggest that there is a strong genetic component to RMR (4,5); e.g., a study with Pima Indians showed that low RMR predicted subsequent weight gain (6). Additionally, a meta-analysis on RMR in post-obese subjects showed that the mean metabolic rate was 3–5% lower than that of never-obese persons, suggesting a propensity to gain weight (7).

Detecting impairment in RMR that may predispose an individual to weight gain is difficult once that person is obese. Obese persons have a greater amount of fat-free mass (FFM) as well as fat mass (FM) than normal weight persons, and since 60–85% of RMR can be attributed to FFM (8), RMR is typically higher in obesity (9). Thus, the primary purpose of this study was to examine differences in components of energy expenditure (EE) in lean and obese women controlling for FFM. Specifically, the primary aim was to determine differences in RMR, activity energy expenditure (AEE), and TEE in lean (BMI = 18.0–24.9 kg/m2) and obese (BMI ≥ 30.0 kg/m2) women after accounting for FFM. Self-reported diet records and doubly labeled water (DLW) estimates of TEE were used to examine energy intake and reporting error in relation to expenditure.

The following were hypothesized: (i) RMR would be similar between the lean and obese women, (ii) TEE would be greater in the obese women due to the higher energy cost of daily movements, and (iii) TEE would be over-compensated in the obese women by a greater daily caloric intake and/or greater degree of energy under-reporting.

Methods and procedures

Subjects

Healthy women aged 18 years and older were recruited to participate in the study. One hundred and eighty-nine women were screened for participation, which included measures of height to the nearest 1.0 cm using a fixed stadiometer and weight to the nearest 0.1 kg using a calibrated balance beam scale. The original intent of the study protocol was to enroll lean and obese women who had similar FFM; thus all potential participants had their body composition measured using bioelectrical impedance analysis (BIA; Quantum X, RJL Systems, Clinton TWP., MI) as part of the screening procedure. All measurements were taken in light clothing without shoes.

Exclusion criteria included smoking, pregnancy, lactation, presence of chronic disease, use of medications known to affect energy metabolism and/or water balance, limitations in mobility, weight change ≥4.5 kg in the past 6 months, anticipated change in physical activity level (PAL), irregular menstrual cycles, history of eating disorder, and a BMI of 25–29 kg/m2. Women reporting high levels of physical activity (PA) (e.g., training for competition or active participation in sporting events) were also excluded. Twenty women (10 lean and 10 obese) were chosen to participate based on their FFM from BIA and their height and age. The study was approved by the Iowa State University Institutional Review Board and all subjects signed an informed consent document prior to participating in both the screening and study procedures.

Study design

This study was observational by design and consisted of 14 continuous days for each subject during the months of August and September. Subjects were asked to not do any traveling or participate in activities or events outside of their normal, daily routine during the 2-week period. Subjects reported to the Human Metabolic Unit of the Center for Designing Foods to Improve Nutrition at Iowa State University between 5 and 7 am on the first day of the study (day 1) and on days 2, 7, and 14. They were instructed to fast for 10 h before each visit and were asked to do minimal activity prior to arrival. Each subject began the study ∼5 days after starting her menstrual period, ensuring that all subjects completed the study during the follicular phase.

Anthropometry and body composition

On day 1, weight was measured as previously described, and circumferences of the waist (smallest circumference between the lower rib and iliac crest), abdomen (umbilicus) and hip (largest protrusion) were measured in duplicate to the nearest 0.1 cm. An average was recorded for each site. Waist-to-hip ratio was calculated as waist circumference divided by hip circumference. Weight was also measured on days 2, 7, and 14.

Also on day 1, a second measure of body composition was obtained using dual energy X-ray absorptiometry (DXA; QDR Delphi, Hologic Inc., Bedford, MA) to determine total and regional FFM and FM. Percent body fat was calculated from the total body mass and FM measurements. One trained operator was responsible for conducting and analyzing scans for all subjects.

Resting metabolic rate

RMR was measured using a hand-held indirect calorimetry device (MedGem Indirect Calorimeter, Microlife Inc. USA, Dunedin, FL). Subjects reclined on a soft, padded chair for 20 min in a dimly lit room, after which they were asked to breathe into the mouthpiece of the device for a total of 10 min. RMR was calculated from the steady state VO2 using an assumed respiratory quotient of 0.85. Separate values were obtained on all 4 days of testing and were averaged to obtain an individual RMR in kilocalories per day. The MedGem has been shown to have excellent agreement with a traditional metabolic cart for measuring RMR via indirect calorimetry (10,11,12).

Total energy expenditure

TEE was determined by the DLW method (13). Subjects consumed a mixed, weight-specific dose of DLW consisting of 1.5 g H218O/kg body weight (10 atom % excess) and 0.06 g 2H2O/kg body weight (99.9 atom % excess) (Cambridge Isotope Laboratories) followed by a 100 ml tap water rinse before leaving the Human Metabolic Unit on day 1. Urine samples were collected in non-acidified plastic containers at baseline, 4 h after dosing and on days 2, 7, and 14 and were stored in 15 ml cryovials at 20 °C until analysis. Samples were used to determine elimination rates of 2H and 18O over the study period by isotope ratio mass spectrometry (Metabolic Solutions Inc., Nashua, NH) using Europa instrumentation (Europa Scientific, Crew, UK). The hydrogen equilibration method of Scrimgeour and colleagues (14) was used for deuterium analysis, and an H2O–CO2 equilibration system was used for measuring 18° (15).

Deuterium and 18° zero-time intercepts and elimination rates were calculated using least-squares linear regression on the natural logarithm of the isotope concentration as a function of elapsed time from dose administration. The zero-time intercepts were used to determine the isotope pool sizes at the time of the dose, and the 2H and 18O pool sizes were used to estimate total body water. The rCO2 production was calculated from the isotope elimination rates and total body water, and CO2 production was used to determine average daily TEE in kilocalories per day during the study period using an estimated respiratory quotient of 0.85 (16).

Activity energy expenditure

AEE was calculated using a standard procedure (AEE = TEE − RMR − (0.1 × TEE)) which assumes that diet-induced thermogenesis represents ∼10% of TEE (17). Another indicator of activity was a standardized PAL which expresses TEE relative to RMR using a simple ratio (18).

PA and posture

Subjects wore two activity monitors simultaneously for 48 continuous hours two times during the study (days 1 and 2 and days 7 and 8). Monitors were worn on weekdays only. The first monitor (Intelligent Device for Energy Expenditure and Activity (IDEEA), MiniSun LLC, Fresno, CA) uses an array of five integrated sensors that are connected to a portable data collection/storage unit. Through the five sensors, the IDEEA monitors body and limb positions continuously on a second-by-second basis and integrates the information to provide the predominant activity type as well as the duration, intensity and EE associated with each activity. The IDEEA monitor has been shown to detect the type, onset, duration and intensity of most fundamental movements with 98% accuracy (19) and to determine EE with 95–99% accuracy (20). Estimations are not affected by body weight, height, BMI, or age.

The second monitor (SenseWear Pro 2 Armband, BodyMedia Inc., Pittsburgh, PA) is another pattern recognition device that is worn over the triceps of the right arm. The monitor uses a two-axis accelerometer, heat flux sensor, skin temperature sensor, near-body ambient temperature sensor, and galvanic skin response sensor to differentiate between different types of movement patterns. The SenseWear Pro 2 Armband provides an estimate of EE on a minute by minute basis. This monitor has been shown to yield more accurate estimates of EE than traditional accelerometers (21).

Subjects were instructed to wear the monitors continuously for both 48-h monitoring periods, removing them only when bathing. They were told to go about their normal daily routine and engage in typical activities while wearing the monitors. Total output from each monitor was averaged to obtain single estimates of daily activity (SenseWear Pro 2 Armband) and posture (IDEEA). The average time the monitors were worn each day was 1370 ± 77 min, except for two individuals (one lean, one obese) who removed the monitors at night due to discomfort (average time worn, 816 ± 81 min). Missing minutes of data were corrected by adding back minutes of RMR so that daily activity estimates reflected a full 24-h period. One obese woman did not wear the IDEEA monitor at any time due to irritation from the tape used to secure the sensors to the skin. Thus, all data reporting posture allocation has n = 9 for obese women and n = 10 for lean women.

Dietary Intake

Subjects recorded all foods and beverages consumed each day for the 14 study days. They were given detailed instructions on how to record items and were provided with materials to assist them in determining portion sizes consumed. They were instructed to eat and drink as they would normally, i.e., they were asked to not deviate from their usual, typical daily eating pattern. Self-reported intake was entered into Nutritionist Pro™ (Axxya Systems LLC, Stafford, TX) by trained study personnel and was analyzed using Nutrition Analysis Software Version 1.3 to obtain estimates for all 14 days. All days were averaged to obtain one value of daily kilocalorie, carbohydrate, protein, and fat intake.

Error in self-reported dietary intake in kilocalories per day was calculated as:

image

where EI represents average daily self-reported energy (kcal) intake, TEE represents average daily EE from DLW, act wt chg × 2.2 is actual body weight change (lb) over the study period, and 3,500 represents the assumption of 3,500 kcal/lb of body weight.

On the last day of the study, subjects were asked to report and explain any deviations from their usual activity or eating patterns during the past 14 days and describe their occupation over the time period. They also filled out a PA questionnaire pertaining to their activities over the previous 2 weeks.

Statistical analysis

Analyses were performed using JMP 5.1 and SAS version 9.0 statistical software packages (SAS Institute, Cary, NC). Two-tailed independent t-tests were used to compare anthropometric and body composition variables, components of EE, daily activity patterns, and self-reported dietary intake and reporting error between lean and obese women. Although FFM was not significantly different between the lean and obese women using BIA (53.4 ± 3.4 vs. 55.9 ± 3.6 kg, respectively, P = 0.12), obese women had greater FFM as determined by DXA (P ≤ 0.001). Since FFM and FM were significantly greater in the obese women, simple linear regression with Pearson's correlation coefficients was used to determine associations between FFM, FM, and TEE and RMR, due to the possible influence of body mass on EE estimates. FFM and FM were significantly related to RMR (r = 0.48, P = 0.03 and r = 0.52, P = 0.02, respectively), thus, RMR was further analyzed using analysis of covariance controlling for FFM and FM. FFM, but not FM, was related to TEE (r = 0.44, P = 0.05), thus, TEE was further analyzed using analysis of covariance controlling for FFM. Data are reported as mean ± s.d. and significance was declared if P ≤ 0.05.

Results

Anthropometric and body composition data are presented in Table 1. All subjects were white and there were no differences in age or height between lean and obese women. Weight, BMI, FFM, FM, percentage body fat, waist-to-hip ratio, and all circumference measurements were greater in the obese women. Although the subjects were specifically instructed to not alter their eating behaviors during the 2-week study period, both groups lost an average of 0.6 kg over the 14 days. All subjects except two reported having occupations that required them to be at a desk for 6–8 h a day (e.g., secretary or other office worker, professor, accountant, manager, administrator). One lean woman was a stay-at-home mom and one obese woman worked from home during the study period. No subjects reported participating in non-typical activities. Lean women reported an average of 40 min of PA a day while obese women reported an average of 21 min of PA. Typical PAs included housework, gardening, walking, and biking.

Table 1. . Anthropometric and body composition characteristics (mean ± s.d.)
 Lean women (n = 10)Obese women (n = 10)P
  1. DXA, dual energy X-ray absorptiometry.

  2. Independent t-tests were used to determine differences (P) between lean and obese women.

Age (y)39.6 ± 5.938.5 ± 6.10.68
Height (cm)169 ± 5167 ± 50.42
Day 1 wt (kg)65.7 ± 4.791.2 ± 9.4<0.001
Wt change (day 1–14, kg)−0.6 ± 1.3−0.6 ± 1.10.96
BMI (kg/m2)23.0 ± 1.632.7 ± 3.1<0.001
Waist circumference (cm)74.1 ± 5.097.3 ± 7.0<0.001
Abdomen circumference (cm)83.9 ± 6.3111.1 ± 7.1<0.001
Hip circumference (cm)98.5 ± 4.1118.3 ± 5.1<0.001
Waist/hip ratio0.75 ± 0.050.82 ± 0.050.008
Fat-free mass (DXA, kg)44.9 ± 3.551.0 ± 3.0<0.001
Fat mass (DXA, kg)17.1 ± 3.536.2 ± 6.7<0.001
Percent (%) fat26.5 ± 4.440.1 ± 3.7<0.001

Self-reported energy intake and reporting error are presented in Table 2. Lean and obese women reported consuming similar amounts of carbohydrate (241 ± 78 g/day vs. 231 ± 58 g/day respectively, P = 0.74), protein (73 ± 12 g/day vs. 80 ± 27 g/day respectively, P = 0.49, fat (73 ± 20 g/day vs. 75 ± 22 g/day respectively, P = 0.87), and total kilocalories per day. Both groups reported similar deficits in daily energy balance (605 ± 629 kcal/day vs. 658 ± 291 kcal/day, lean and obese respectively, P = 0.81) and after factoring in actual weight change over the study period, both groups were found to under-report their caloric intake by 275–312 kcal/day. Seven out of the ten obese women under-reported their caloric intake (48 to 1586 kcal/day), as did 8 out of the 10 lean women (74 to 845 kcal/day). There were no significant differences in degree of under-reporting between groups.

Table 2. . Energy intake and energy expenditure (mean ± s.d.)
 Lean (n = 10)Obese (n = 10)P
  • The average of 14 days of self-reported EI was used to determine daily EI. TEE was determined using doubly labeled water, RMR was measured by indirect calorimetry, and AEE was calculated as AEE = TEE - RMR - (0.1 × TEE), where 0.1 × TEE represents diet-induced thermogenesis. Independent t-tests were used to determine differences (P) between lean and obese women.

  • a

    Analysis of covariance was used to determine differences in TEE and AEE (P) between groups after controlling for fat-free mass (data presented as mean ± s.e.).

  • b

    Amount of error in energy (kcal) reporting after factoring in actual weight change over the 14-day study period (EI-TEE)-(act wt chg × 2.2 × 3,500/14).

  • c

    Analysis of covariance was used to determine differences in RMR (P) between groups after controlling for fat-free and fat mass (data presented as mean ± s.e.).

Self-reported energy intake (EI, kcal/day)1,914 ± 4181,935 ± 4810.92
Total energy expenditure (TEE, kcal/day)2,519 ± 4182,593 ± 3190.66
    Adjusted TEEb2,698 ± 1262,414 ± 1260.02
EI reporting errorb (kcal/day)−275 ± 378−312 ± 6690.88
Resting metabolic rate (RMR, kcal/day)1,440 ± 1041,666 ± 2600.03
    Adjusted RMRc1,505 ± 1091,601 ± 1090.12
Activity energy expenditure (AEE, kcal/day)820 ± 411673 ± 3040.38
    Adjusted AEEa943 ± 133550 ± 1330.09

Table 2 also demonstrates daily energy expenditure. TEE was not different between lean and obese women; however, after adjusting for FFM, TEE was significantly lower in the obese women. Absolute RMR was higher in obese women and although the difference became non-significant after adjusting for FFM and FM, a moderate effect size for a difference remained (0.48). AEE was not significantly different between lean and obese women; however, the calculated effect size for the difference was moderate (0.41) and after adjusting for FFM, obese women expended almost 400 kcal less in activity per day. Likewise, PAL was not significantly different between groups (1.75 ± 0.34 vs. 1.59 ± 0.25, lean and obese, respectively, P = 0.23) but a moderate effect size was noted (0.55), and a trend for obese women to have lower PAL was evident after adjusting for FFM (1.51 ± 0.11 vs. 1.83 ± 0.11, P = 0.09).

Time spent at different activity intensities (SenseWear Pro 2 Armband monitor) was investigated using metabolic equivalents (1 metabolic equivalent = 1 kcal/kg/h), which allowed for examination of activity data after factoring in body weight (Figure 1). Obese women spent significantly more time each day at rest or in sedentary behaviors than lean women. Accordingly, obese women spent less time being active than lean women, including less time in light, moderate, and vigorous activities. Total minutes of daily activity were lower in obese women by more than half (158 ± 88 min/day vs. 323 ± 113 min/day, P = 0.002).

Figure 1.

: Average minutes per day spent in different activity intensities (reported as metabolic equivalents or METs, 1 MET = 1 kcal/kg/h) in lean (n = 10) and obese (n = 10) women as determined by SenseWear Pro 2 Armband monitor over 96 h of monitoring. Two-tailed independent t -tests were used to determine differences between groups. *Obese women spent more time at rest (≤1.5 METs) than lean women at P = 0.002 (1,282 ± 88 vs. 1,117 ± 113 min/day). †Obese women spent less time in light activity (>1.5 to ≤3 METs) than lean women at P = 0.003 (103 ± 70 vs. 221 ± 83 min/day). ‡There was a trend for obese women to spend less time in moderate activity (>3 to ≤6 METs) than lean women at P = 0.08 (51 ± 28 vs. 79 ± 39 min/day). §Obese women spent less time in vigorous activity (≥6 METs) than lean women at P = 0.04 (4 ± 7 vs. 23 ± 26 min/day).

Daily posture allocation (IDEEA monitor) is illustrated in Figure 2. Obese women spent more time each day sitting than lean women and less time standing. Time spent lying down was not different between groups nor was total locomotion including walking, stair-stepping, running, and jumping. When calculated as a percentage of an average day, obese women sat 53% and lean women sat 42% of the time. Likewise, obese women stood 11% and lean women stood 20% of the time. Obese women also took fewer steps per day on average; 6,970 ± 2,351 compared to 11,393 ± 3,384 steps/day in the lean women (P = 0.003).

Figure 2.

: Average minutes per day spent in different posture allocations in lean (n = 10) and obese (n = 9) women as determined by Intelligent Device for Energy Expenditure and Activity monitor over 96 h of monitoring. One obese woman did not wear a monitor due to irritation from the tape used to secure sensors to the skin. Two-tailed independent t -tests were used to determine differences between groups. *There was a trend for obese women to spend more time sitting than lean women at P = 0.06 (760 ± 193 vs. 604 ± 118 min/day). †Obese women spent less time standing than lean women at P = 0.02 (163 ± 58 vs. 284 ± 134 min/day). Time spent lying down was not different between lean and obese women (489 ± 96 vs. 467 ± 173 min/day, respectively, P = 0.73) nor was total locomotion (60 ± 29 vs. 48 ± 16 min/day, respectively, P = 0.27).

Discussion

The objectives of this study were to investigate differences in daily energy expenditure between lean and obese women controlling for FFM and to examine differences in daily activity patterns between the two groups. Specifically, we sought to determine whether obese women had lower RMR than lean women, which may have pre-disposed them to weight gain leading to their obese state.

We did not find differences between the groups in either energy intake or reporting error, although we cannot be certain of the absolute values associated with reporting error since the assumption that one pound of body weight change = 3,500 kcal corresponds to the change consisting entirely of fat tissue. Our subjects most likely lost lean tissue as well as fat tissue and also experienced shifts in fluid balance over the 14-day period. However, we are confident that there were no significant differences in the magnitude of reporting error between the two groups.

As stated previously, the original intent of the study protocol was to select lean and obese women who had similar FFM. Although values were similar based on BIA measurements, we found that our obese women had significantly more FFM than did our lean women according to DXA. Previous studies have shown BIA to overestimate FFM compared to DXA (22,23) and our results indicated similar overestimation, which was more pronounced in the lean women. Thus, to minimize the impact of the differences in FFM on our results, we investigated differences in EE using actual values and after adjusting for FFM statistically.

Given that these obese women had greater FFM and FM than the lean women, we were not surprised to find that absolute RMR was higher by ≥220 kcal/day, since both components were found to be significantly related to RMR. Following adjustment for FFM and FM, our finding of similar RMR supports our hypothesis of no impairment pre-disposing obese women to weight gain. In fact, RMR remained higher in the obese women by ∼100 kcal/day even after adjusting for body composition. Due to the small number of subjects in each group, we cannot speculate as to why RMR remained higher in the obese group of women, since our sample size does not allow for the accurate determination of major contributors to RMR within groups. Our results are supported by previous work showing that once RMR is normalized for FFM, there are no differences between obese and non-obese persons (24,25). Although some research suggests a strong genetic component to RMR and implicates low RMR as a predictor of weight gain (4,5,6), other studies have shown no contribution of RMR to future obesity risk (26,27,28). It may be possible that RMR is impaired in certain homogeneous populations, and in our small sample of heterogeneous women, we were not able to detect such impairment.

Our finding of similar absolute TEE is contrary to our hypothesis of greater TEE in the obese women, which was based on previous research showing that obese persons expend more total kilocalories per day than do non-obese persons (9). After adjusting TEE for FFM, we found that obese women expended 285–375 fewer kilocalories per day than lean women. This finding is contrary to other studies that reported similar TEE in obese and non-obese persons after accounting for components of body size (18,29).

When we examined EE related to activity, we found that absolute AEE was lower in obese women by ∼150 kcal/day, although this difference was not statistically significant (P > 0.05). However, since the energetic cost of conducting weight-bearing movements is greater in obese persons than equivalent movements in non-obese persons (30), our results suggested that levels of activity in our obese women were actually lower than what DLW estimates indicated. Several studies have demonstrated that estimates of AEE from DLW may inflate the amount of physical activity actually done by an obese person (31,32). Supporting this concept, we found AEE to be lower in our obese women by ∼400 kcal/day after adjusting for FFM, and although the significance did not change, the difference increased to 455 kcal/day when controlling for total body mass.

Investigating differences in PAL further supported the discrepancies between lean and obese women in energy expended being active. We found a PAL of 1.75 in the lean women compared to 1.59 in the obese women, and after adjusting for FFM, the difference in PAL increased. Previous studies have found that subjects who were active enough to raise their PAL to >1.7 were better able to maintain their weight than those below that threshold (33,34). From our results, we concluded that energy expended in activity was lower in the obese women. An interesting observation further supporting our conclusion was that on average the lean women expended 12.5 kcal/kg body weight in activity, whereas the obese women expended only 7.4 kcal/kg body weight. There is some evidence to suggest that the minimal amount of EE by PA required for protection against body fat gain is ∼12 kcal/kg body weight per day (35).

The major difference in activity patterns between the lean and obese women was the time spent being sedentary versus being lightly active, i.e., sitting vs. standing. We found that our obese women were seated 2.70 h more each day and this was reflected in the 2.75 h of additional sedentary behavior observed. Accordingly, the obese women stood 2 h less each day and this was reflected in the 2 h less of light-intensity activity observed. Overall, the obese women were active for ∼2.5 h each day whereas the lean women were active for over 5 h daily. Although the lean women in our study claimed to be quite sedentary with minimal planned PA, it appeared that they were engaging in amounts of activity recommended for avoidance of weight gain, e.g., ≥10,000 steps and ∼80 min of moderate-intensity activity per day with a PAL of ≥1.7 (32), whereas the obese women fell well below these recommendations.

Supporting our results, Levine et al. (36) recently examined posture and movement over 10 days in a group of 10 lean and 10 mildly obese sedentary subjects and found that the obese individuals were seated 2 h longer per day than the lean individuals and were upright (standing) 2.5 h less. The authors estimated that if the obese individuals adopted the activity behaviors of the lean individuals, they could expend an additional 350 kcal/day. This estimate is similar to our value of ∼315 kcal/day (2.75 h light/moderate activity at 2.5 metabolic equivalents 2.75 h sedentary activity at 1.5 metabolic equivalents = (2.5 kcal × 91.0 kg × 2.75 h) (1.5 kcal × 91.0 kg × 2.75 h) = 313 kcal). One could argue that differences in activity behavior are a result of the obese state and not a cause, and we are unable to draw any conclusions from our cross-sectional study. However, in the study mentioned previously (36), when the lean individuals gained weight and the obese individuals lost weight over a 2-month period, their posture allocation remained unchanged. This maintenance of daily posture despite changes in body weight suggests that lower activity precedes weight gain and obesity, although the mechanism(s) driving this activity behavior remains unclear.

A unique aspect of the present study is the utilization of state-of-the-art technologies to investigate differences in daily EE and activity patterns in a group of heterogeneous, healthy, vibrant, lean, and obese women. In our sample of women, we found that those who were obese spent more time each day being sedentary, e.g., sitting, and less time each day being active, e.g., standing. These differences in posture allocation were reflected in daily EE; obese women expended less energy in activity than lean women and instead relied on a greater body mass to burn similar amounts of total kilocalories. Supporting this conclusion, once adjusted for body mass characteristics, RMR was similar and AEE and TEE were lower in the obese women. We did not find higher energy intake or a greater degree of energy under-reporting in the obese women; however even a minimal energy excess of a few kilocalories per day combined with less AEE would promote continued weight gain over time. Additional research is needed to better understand the impact of differences in posture allocation and activity patterns on individual risks for overweight and obesity.

DISCLOSURE

The authors declared no conflict of interest.

Acknowledgments

The authors would like to acknowledge the participants for their time and commitment to this project. This article is dedicated to the late Paul Flakoll, who made this work possible. Dr Flakoll was a wonderful mentor, teacher, and friend who is sorely missed and whose memory lives on.

Ancillary