ABSTRACT
Background
Individuals with overweight or obesity commonly underreport energy intake (EI), but it is unknown if the tendency to underreport persists in formerly obese individuals who lose significant weight and maintain their weight loss over long periods of time.
Objective
Assess the accuracy of self-reported EI in successful weight loss maintainers (WLM) compared with controls of normal body weight (NC) and controls with overweight/obesity (OC).
Methods
Participants for this case-controlled study were recruited in 3 groups: WLM [n = 26, BMI (in kg/m2) 24.1 ± 2.3; maintaining ≥13.6 kg weight loss for ≥1 y], NC (n = 33, BMI 22.7 ± 1.9; similar to current BMI of WLM), and OC (n = 32, BMI 34.0 ± 4.6; similar to pre–weight loss BMI of WLM). Total daily energy expenditure (TDEE) was measured over 7 d using the doubly labeled water (DLW) method, and self-reported EI was concurrently measured from 3-d diet diaries. DLW TDEE and self-reported EI were compared to determine accuracy of self-reported EI.
Results
WLM underreported EI (median, interquartile range) (–605, –915 to –314 kcal/d) to a greater degree than NC (–308, –471 to –68 kcal/d; P < 0.01) but not more than OC (–310, –970 to 18 kcal/d; P = 0.21). WLM also showed a greater degree of relative underreporting (–25.3%, –32.9% to –12.5%) compared with NC (–14.3%, –19.6% to –3.1%; P = 0.02) but not OC (–11.2%, –34.1% to –0.7%; P = 0.16). A greater proportion of WLM was classified as underreporters (30.8%) than NC (9.1%; P = 0.05) but not OC (28.1%; P = 1.00).
Conclusions
WLM underreported EI in both absolute and relative terms to a greater extent than NC but not OC. These findings call into question the accuracy of self-reported EI in WLM published in previous studies and align with recent data suggesting that WLM rely less on chronic EI restriction and more on high levels of physical activity to maintain weight loss. This trial was registered at clinicaltrials.gov as NCT03422380.
Keywords: weight loss maintainers, self-reported energy intake, 3-d diet diaries, doubly labeled water, food log accuracy, cognitive restraint, Three-Fold Eating Inventory
Introduction
Assessment of energy intake (EI) comprises a critical component of weight management and has been heavily relied on to shape national health promotion and disease prevention guidelines (1, 2). However, self-reported EI is problematic as an assessment tool because it suffers from inaccuracy and bias (3–5). Underreporting of EI is a common bias and has been documented in numerous populations (3, 6). Adults with obesity [BMI (in kg/m2) ≥30] seem especially prone to underreport EI compared with those of normal body weight (7, 8), but it is less clear whether this tendency to underreport persists in formerly obese individuals who successfully lose weight and maintain weight loss over long periods of time (9, 10). Characterizing the accuracy of self-reported EI in successful weight loss maintainers may have important clinical implications as this population has been frequently investigated to identify strategies for preventing weight regain (11–14).
Previous studies that analyzed self-reported data from weight loss maintainers have suggested they achieve long-term energy balance by decreasing EI (15–18) and increasing physical activity (PA) (19–22). The doubly labeled water (DLW) method for estimating total daily energy expenditure (TDEE) (23), along with increasingly sophisticated accelerometer technology, has provided compelling objective evidence to verify the high levels of self-reported PA in weight loss maintainers (24–26). However, far less attention has been given to verifying their self-reported EI (10, 20). In a study that included 11 “successful slimmers” (maintaining 27.7 ± 12.5-kg weight loss), Black et al. (10, 27) compared DLW TDEE with food records and found evidence of underreporting in 55% of participants, but 2 subjects had maintained weight loss for only 1 mo and the sample size was small. In a previous study using DLW, we showed that TDEE in weight-loss maintainers was similar to TDEE in individuals with overweight and obesity of a substantially higher body weight (25). These results suggest that successful maintainers use high levels of PA (rather than chronic restriction of EI) to achieve long-term energy balance at a reduced body weight and raise significant concerns about the reliability of self-reported EI data in this population. Further investigation using DLW to evaluate the accuracy of self-reported EI of weight-loss maintainers is still needed.
The primary aim of the present study was to use DLW-derived measures of TDEE to assess the accuracy of self-reported EI in successful weight loss maintainers compared with controls of normal body weight and controls with overweight/obesity. Given the similar TDEE in weight loss maintainers and individuals with overweight/obesity in our prior study (25), we hypothesized that weight loss maintainers would more closely resemble individuals with overweight/obesity than individuals of normal weight in both the degree and prevalence of underreporting of EI. As a secondary aim, we sought to identify potential drivers of underreporting by examining the correlation between underreporting of EI and other behavioral parameters, including self-reported macronutrient intake, eating behavior (cognitive restraint, disinhibition, and hunger), and device-measured PA and sedentary behavior.
Methods
This secondary data analysis was conducted on a case-control study carried out at the University of Colorado Anschutz Medical Campus between October 2009 and August 2012. The study was approved by the Colorado Multiple Institutional Review Board. Participants were recruited through campus flyers and email announcements, and interested individuals underwent a preliminary telephone screening to determine eligibility for 1 of 3 subject groups: 1) weight loss maintainers [WLM; maintaining ≥13.6 kg (30-lb) weight loss for ≥1 y], 2) normal-weight controls (NC; current BMI similar to the current BMI of WLM and no history of obesity), and 3) controls with overweight/obesity (OC; current BMI similar to the pre–weight loss maximum BMI of WLM and not maintaining weight loss ≥13.6 kg). Weight history was obtained from self-report. A nested subject selection procedure was used to obtain similar distributions for age (categories <36, 36–49, and ≥50 y) and sex (male compared with female) across all three groups. It was also designed to ensure similar distribution for BMI between NCs' and WLMs' current BMI (kg/m2; categories <22, 22 to <25, and 25–30) and similar distribution for BMI between OCs’ and WLMs’ pre–weight-loss maximum BMI (categories 25 to <30, 30 to <35, 35 to <40, and ≥40).
Individuals were excluded from the study if they had any physical or medical condition that restricted PA (including diabetes, cardiovascular disease, cancer, and significant musculoskeletal, neurologic, or psychiatric disorders); had undergone weight loss surgery; were taking weight loss medications or followed extreme diets such as prolonged fasting, ketogenic diets, or very low-calorie diets (<800 kcal/d); were smokers; were not weight stable (>5-kg fluctuation in body weight over past 6 mo); or were taking medications known to affect appetite or metabolism. Women who were pregnant or lactating were also excluded.
Body weight and composition
Body weight was measured at the screening visit and on days 1 and 8 of the study with a calibrated digital scale (to the nearest 0.2 lb; BWB-800; Tanita Corp.). Height was measured at screening with a wall-mounted stadiometer (to the nearest 0.1 cm). Fat mass (FM) and fat-free mass (FFM) were measured with DXA (Delphi-W version 13.1.0; Hologic) at screening. One OC participant's supine body width exceeded scan window dimensions, so FM and FFM were determined from bioelectrical impedance analysis (TBF-105; Tanita).
Resting energy expenditure
Resting energy expenditure (REE) was measured using standard indirect calorimetry (Truemax 2400; Parvo Medics) with the ventilated hood technique. Before each test, the gas analyzers and flow meter were calibrated per manufacturer recommendations. Participants were instructed to fast for 12 h overnight, which was confirmed by study staff. Upon arrival (∼07:00), participants rested supine, awake, and lightly clothed with access to a blanket in a thermoneutral (20–23°C), dimly lit, quiet room for 30 min (28). Respiratory gas exchange was measured for 15 min, and the last 10 min was used to average REE. REE was calculated using the Weir equation. Criteria employed to determine whether the REE measurement was acceptable included stability (coefficient of variance of the final 10 min <5%) and average metabolic equivalents (METs) <1.10, as previously described (29). REE was measured on days 1 and 8 and averaged to produce a single value for REE. In some instances, participants had only 1 valid REE (WLM: n = 2; NC: n = 5; OC: n = 3), and the single measure was used in place of an average REE. Intraclass correlation coefficient for day 1 and day 8 within-subject REE measures was high (0.96).
Self-reported energy intake, Three-Factor Eating Inventory
Three-day diet diaries were used to assess self-reported EI. Participants were trained by a registered dietitian on how to complete the diet diaries and then instructed to record their intake to include 2 weekdays and 1 weekend day. Assessments were performed during a “typical” week, and participants who were traveling, ill, engaged in atypical levels of physical activity, or celebrating holidays were rescheduled for a different week. The 3-d diet diaries were reviewed for completeness by a registered dietitian and analyzed by the Colorado Clinical Translational Research Institute Nutrition Core staff, blinded to participant group assignment, using Nutrition Data Systems for Research software (30). The Three-Factor Eating Inventory (TFEI) questionnaire was completed on day 8. The TFEI is a self-assessment scale used widely in studies of eating behavior (31, 32), and it includes 3 subscales: cognitive restraint, disinhibition, and hunger. Cognitive restraint is associated with the conscious intent and ability to restrain eating (33). Disinhibition is a measure of the tendency toward episodic overeating (34). The hunger scale measures participants’ perceived levels of hunger (31). Participant responses to each question were coded as either 0 or 1 point. Each subscale contained a different number of possible points (cognitive restraint: 21 points, disinhibition: 16 points, hunger: 14 points). Points for each subscale were added together, with higher cumulative scores representing stronger characteristics in the associated domain. Scores in the 3 domains were averaged across each group for subsequent comparison.
Total daily energy expenditure, physical activity energy expenditure, and patterns of physical activity
During the same week in which the 3-d diet diary was completed, TDEE was measured using the DLW method (23). On the dosing day (day 1), participants arrived at the testing center following a 12-h overnight fast. Upon arrival, participants voided their bladder and provided a baseline urine sample for determination of background levels of 2H and 18O. Participants then consumed an oral dose of DLW containing 2.05 g/kg total body water (TBW; estimated as 73% of FFM determined from DXA) of 10 atom percent excess 18O and 0.14 g/kg TBW of 99.8 atom percent excess 2H (ISOTEC; Sigma Aldrich). The dosing cup was rinsed twice with 30 mL of water, and the rinsing dose was consumed. Exact time of dosing was recorded. Participants were instructed to void their bladder 1 to 3 h after dosing, and additional urine samples were collected 4 and 5 h after the dosing. On collection day (day 8), participants returned to the testing center. They were instructed to void upon waking, and the second and third voids were obtained at the same times as the 4- and 5-h postdose samples on the dosing day. The urine samples were analyzed for 18O enrichment by isotope ratio mass spectrometry after equilibration with carbon dioxide. 2H was reduced with a platinum catalyst, and the deuterium enrichment was determined by isotope ratio mass spectrometry (DELTA V Advantage; Thermo Electron North America LLC). Each sample was analyzed in duplicate. If the difference between duplicate runs exceeded 2 δ ‰ for 2H:1H or 1 δ ‰ for 18O:16O for a given sample, the sample was run again, and only duplicate values that fell within this range were used. TBW was calculated as the average of the dilution spaces of 2H and 18O (35), and the rate of carbon dioxide production was calculated using equation A6 from Schoeller et al. (23).
To be included in the analysis, data had to meet the following quality control (QC) criteria: 1) 1.0–1.07 dilution space ratio, 2) TBW estimates from 2H and 18O within 10% of TBW calculated from DXA FFM, 3) slopes of elimination using the 4- and 5-h urine collections on day 1 and day 8 within 10% of each other, and 4) an abundance of 18O at least 15 δ ‰ above background at day 8. TDEE was then estimated using the Weir equation, assuming a respiratory quotient of 0.86 (28). Because participants were weight stable during the assessment period [median weight change (kg), IQR; WLM: –0.3, –0.8 to 0.1 kg, NC: –0.1, –0.3 to 0.3 kg, OC: –0.2, –1.1 to 0.6 kg; P = 0.39], DLW TDEE was used as an objective measure of EI for comparison with self-reported diet diary data, as previously described (36).
Physical activity energy expenditure (PAEE) was calculated as [TDEE − (0.1 × TDEE) − REE], which assumes the thermic effect of feeding is 10% of TDEE. Because the energy cost of PA is proportional to body weight for a given intensity and duration (37), PAEE was calculated relative to body weight (in kilograms) and used to test for correlations with reporting accuracy.
Physical activity was assessed using the ActivPAL activity monitor (PAL Technologies). Participants were asked to wear the device continuously for 7 d. Data were considered valid and used for analysis if the device was worn for >10 h/d on ≥4 d (including ≥2 weekdays and ≥1 weekend day). Stepping events were categorized as light-intensity PA (LPA, 1.50–2.99 METs) and moderate to vigorous intensity PA (MVPA, ≥3.00 METs) by using 75 steps/min as a threshold (75 steps/min = 3.00 METs). Total MVPA (min/d) was computed as the sum of time spent in MVPA (stepping events ≥3.00 METs). To control for differences in sleep time and express data in terms of the percentage of awake time, the event data file was visually inspected to estimate time into bed at night and time out of bed in the morning by using methods described previously (38). Metrics of sedentary behavior were estimated during time spent awake and included total sedentary time (total time spent in sitting/lying events), total breaks in sedentary time (number of times a sitting/lying event was followed by a standing or stepping event), sedentary break rate (total number of breaks per total sedentary time in hours), time (min/d) in bouts ≥30 and ≥60 min, and number of discrete sedentary bouts ≥30 and ≥60 min/d.
Comparison of self-reported energy intake to total daily energy expenditure
To assess the accuracy of self-reported EI, TDEE from DLW was compared with energy intake from self-reported 3-d diet diaries. Both absolute (self-reported EI − DLW TDEE) and relative [(self-reported EI − DLW TDEE)/DLW TDEE] * 100 reporting accuracy were assessed. Participants were classified as underreporters on the basis of the lower boundary of the 95% confidence limit of the expected self-reported EI/DLW TDEE ratio of 1.0, given the following equation:
 |
(1) |
where CVEI denotes within-participant coefficient of variation for self-reported EI, CVTDEE denotes the within-participant coefficient of variation for TDEE as measured by DLW, and D denotes the number of days of dietary assessment (39). Root square of the average squared individual CV over the 3 days of diet diary reporting was used to calculate a within-subject CVEI of 24.4% in our data (40). The previously published value of 8.2% was used for within-participant CVTDEE (39). Based on these values, the lower boundary of the 95% confidence limit of the expected self-reported EI/DLW TDEE ratio of 1.0 was 0.674, and participants with a self-reported EI/DLW TDEE ratio <0.674 were classified as underreporters.
Statistical analysis
Statistical analyses were performed using Stata version 16.1 (StataCorp), and the type I error rate was set at 0.05. Categorical characteristics were compared across groups using χ2 and Fisher exact tests. Normality of outcome measures and equal variance assumptions were verified with the Shapiro–Wilk test and Bartlett test, respectively. In cases where data were not normally distributed (P < 0.05), nonparametric tests were used, with the Kruskal–Wallis test comparing overall differences and the Wilcoxon rank-sum test used for pairwise comparisons. One-way ANOVA with Bonferroni correction for multiple comparisons was used to test between-group differences in macronutrient composition of diets as recorded in the 3-d diet diaries. Within-group differences in self-reported EI compared with DLW TDEE were measured with a paired samples t test. Spearman's ρ was used to test the correlation between relative underreporting and PA data (ActivPAL), reported macronutrient intake, PAEE, and TFEI scores. All correlations were performed using aggregated data from the 3 groups. Bland–Altman analysis was conducted to assess agreement between DLW TDEE and self-reported EI, in accordance with published methods (41). Power was estimated for the previously published (25) primary outcome of the study (difference in DLW TDEE between groups) using power and sample size software (NCSS). Based on the most conservative assumptions, it was estimated that 35 participants per group would be needed to have 80% power to detect a 130-kcal/d difference in TDEE. Although <35 subjects per group were included in the final analysis, sufficient statistical power was retained to observe a between-group difference of 130 kcal/d, as evidenced by the 95% CI of the between-group difference in TDEE (WLM/NC 60−538 kcal/d, WLM/OC −316 to 158 kcal/d) (25). There was no a priori power analysis for the outcomes of the secondary data analysis.
Results
The study Consolidated Standards of Reporting Trials diagram is shown in Figure 1. Of the 106 participants who were enrolled in the study, TDEE data from 12 participants (WLM: n = 9; NC: n = 2; OC: n = 1) were excluded from the analysis based on the DLW QC criteria outlined in the Methods. Three OC participants were later excluded because their self-reported weight history was incorrect, and they no longer met the criteria for inclusion in any of the 3 study groups. Of the remaining participants, 4 did not have valid REE (NC: n = 1; OC: n = 3), resulting in a sample size of 87 participants used to obtain correlations between self-reported EI accuracy and PAEE (WLM: n = 26; NC: n = 32; OC: n = 29). Fifteen participants did not have valid ActivPAL data (WLM: n = 5; NC: n = 2; OC: n = 8), resulting in a sample size of 76 participants to obtain correlations between self-reported EI accuracy and measures of PA (WLM: n = 21; NC: n = 31; OC: n = 24).
FIGURE 1.
Study Consolidated Standards of Reporting Trials diagram. DLW, doubly labeled water; EI, energy intake; PA, physical activity; PAEE, physical activity energy expenditure; QC, quality control; REE, resting energy expenditure.
Participant characteristics are shown in Table 1. The nested subject selection procedures successfully achieved similar group means for age and sex, but there was a difference in the proportion of nonwhite participants between groups (P = 0.02). The current BMI of WLM (mean ± SD; 24.1 ± 2.3) was not different from the current BMI of NC (22.7 ± 1.9; P = 0.11). The pre–weight loss maximum BMI of WLM (median, IQR; 32.7, 30.2–34.4) was not different from the current BMI of OC (34.0 ± 4.6; P = 0.46) (statistical comparison not shown in Table 1). WLM were maintaining a weight loss of approximately 22 kg for 4.5 y.
TABLE 1.
Characteristics of study participants1
| Characteristic | WLM (n = 26) | NC (n = 33) | OC (n = 32) | Overall P value | P value, WLM/NC | P value, WLM/OC | P value, NC/OC |
|---|---|---|---|---|---|---|---|
| Age, y | 44 (36–54) | 49 (32–57) | 46 (38–56) | 0.77 | 0.53 | 0.55 | 0.81 |
| Sex, male, n (%) | 5 (19) | 8 (24) | 7 (22) | 0.95 | 0.76 | 1.00 | 1.00 |
| Ethnicity,2n (%) | 0.58 | 0.37 | 0.62 | 0.37 | |||
| Hispanic/Latino | 1 (4) | 4 (12) | 3 (9) | ||||
| Not Hispanic/Latino | 25 (98) | 29 (88) | 29 (91) | ||||
| Race,2n (%) | 0.02* | 0.19 | 0.03* | 0.10 | |||
| White | 26 (100) | 27 (82) | 26 (81) | ||||
| Black/African American | 0 (0) | 2 (6) | 6 (17) | ||||
| Asian | 0 (0) | 3 (9) | 0 (0) | ||||
| Not reported | 0 (0) | 1 (3) | 0 (0) | ||||
| Anthropometric measures | |||||||
| Current weight, kg | 68.0 ± 9.3 | 62.8 ± 10.4 | 96.6 ± 18.8 | <0.01* | 0.16 | <0.01* | <0.01* |
| Current BMI, kg/m2 | 24.1 ± 2.3 | 22.7 ± 1.9 | 34.0 ± 4.6 | <0.01* | 0.11 | <0.01* | <0.01* |
| Maximum weight,2,3 kg | 89.8 (81.6–98.9) | 65.8 (60.3–70.3) | 99.4 (83.2–114.1) | <0.01* | <0.01* | 0.17 | <0.01* |
| Maximum BMI,2,3 kg/m2 | 32.7 (30.2–34.4) | 24.3 (22.8–25.4) | 35.0 (31.6–39.2) | <0.01* | <0.01* | 0.03* | <0.01* |
| Weight loss currently maintained,2,4 kg | 22.3 (17.3–26.5) | 4.4 (2.4–5.1) | 2.8 (0.85–6.1) | <0.01* | <0.01* | <0.01* | 0.36 |
| Weight loss maintenance duration,2 y | 4.5 (3.0–11.0) | 5.5 (1.0–12.5) | 6 (2.0–15.0) | 0.82 | 0.60 | 0.95 | 0.59 |
Values are reported as median (IQR) for nonnormally distributed data and mean ± SD reported for normal data unless otherwise indicated. Kruskal–Wallis test used for overall significance and Wilcoxon rank-sum test used for individual comparisons of nonparametric data. Fisher exact test used for categorical variables; 1-factor ANOVA used to compare normally distributed data. *Significant P values (α < 0.05) indicated by asterisk. NC, normal-weight controls; OC, controls with overweight/obesity; WLM, weight loss maintainers.
Self-reported.
After age 18 y and excluding pregnancy.
After age 18 y and excluding illness. Calculated as highest adult weight – current weight.
TDEE and self-reported EI data (median kcal/d, IQR) are presented in Table 2. DLW TDEE was significantly different between WLM (2399, 2235–2609 kcal/d) and NC (2029, 1909–2469 kcal/d; P < 0.01) but not OC (2631, 2191–2913 kcal/d; P = 0.34). Self-reported EI was significantly lower in all 3 groups compared with DLW TDEE (P < 0.01) (Figure 2). WLM underreported EI to a greater degree in absolute terms (–605, –915 to –314 kcal/d) compared with NC (–308, –471 to –68 kcal/d; P < 0.01) but not compared with OC (–310, –970 to 18 kcal/d; P = 0.21) (Figure 3). Based on bias and limits of agreement, the Bland–Altman analysis showed better agreement between DLW TDEE and self-reported EI for NC than WLM or OC (Figure 4). WLM also showed a greater degree of relative underreporting (–25.3%, –32.9% to –12.5%) compared with NC (–14.3%, –19.6% to –3.1%; P = 0.02) but not OC (–11.2%, –34.1% to –0.7%; P = 0.16) (Figure 3). A greater proportion of WLM was classified as underreporters (30.8%) compared with NC (9.1%; P = 0.05) but not OC (28.1%; P = 1.00) (Figure 5).
TABLE 2.
Comparison of weight change, DLW TDEE, self-reported EI, self-reporting accuracy, macronutrient intake, and TFEI scores across groups1
| Characteristic | WLM (n = 26) | NC(n = 33) | OC (n = 32) | Overall P value | P value, WLM/NC | P value, WLM/OC | P value, NC/OC |
|---|---|---|---|---|---|---|---|
| Weight change, DLW TDEE, SREI, self-reporting accuracy, median (IQR) | |||||||
| Weight change (day 8 – day 1), kg | –0.3 (–0.8 to 0.1) | –0.1 (–0.3 to 0.3) | –0.2 (–1.1 to 0.6) | 0.39 | 0.12 | 0.77 | 0.53 |
| DLW TDEE, kcal/d | 2399 (2235–2609) | 2029 (1909–2469) | 2631 (2191–2913) | <0.01* | <0.01* | 0.34 | <0.01* |
| Self-reported EI, kcal/d | 1873 (1622–2141) | 1878 (1624–2161) | 2046 (1761–2568) | 0.18 | 0.94 | 0.14 | 0.10 |
| Absolute SREI accuracy,2 kcal/d | –605 (–915 to –314) | –308 (–471 to –68) | –310 (–970 to 18) | 0.06 | <0.01* | 0.21 | 0.76 |
| Relative SREI accuracy,3 % | –25.3 (–32.9 to –12.5) | –14.3 (–19.6 to –3.1) | –11.2 (–34.1 to –0.7) | 0.07 | 0.02* | 0.16 | 0.91 |
| Proportion of underreporters,4 % | 30.8 | 9.1 | 28.1 | 0.07 | 0.05* | 1.00 | 0.06 |
| Self-reported macronutrient intake, mean ± SD | |||||||
| % kcal from fat | 30.8 ± 7.3 | 33.3 ± 6.1 | 36.4 ± 5.2 | <0.01* | 0.12 | <0.01* | 0.05* |
| % kcal from carbohydrate | 49.5 ± 7.2 | 47.7 ± 7.2 | 44.1 ± 6.8 | 0.01* | 0.33 | <0.01* | 0.04* |
| % kcal from protein | 18.6 ± 5.5 | 16.4 ± 4.3 | 17.8 ± 4.8 | 0.23 | 0.09 | 0.54 | 0.26 |
| TFEI measures, median (IQR) | |||||||
| Cognitive restraint | 16 (12–18) | 11 (8–13) | 7 (5–10) | <0.01* | <0.01* | <0.01* | 0.02* |
| Disinhibition | 6 (5–9) | 4 (2–6) | 10 (7–13) | <0.01* | 0.01* | <0.01* | <0.01* |
| Hunger | 5 (3–7) | 3 (2–5) | 5 (4–7) | <0.01* | 0.04* | 0.56 | <0.01* |
Median (IQR) reported for nonnormally distributed data and mean ± SD reported for normal data. Nonparametric analysis used for energy intake, energy expenditure, reporting accuracy, and TFEI scores, with the Kruskal–Wallis test used for overall significance and Wilcoxon rank-sum test used for individual comparisons. One-factor ANOVA used to compare self-reported macronutrient intakes across groups with a Bonferroni correction for multiple comparisons. Fisher exact test used to compare proportion of underreporters. *Significant P values (α < 0.05) indicated by asterisk. DLW TDEE, total daily energy expenditure as determined by doubly labeled water; EI, energy intake; NC, normal-weight controls; OC, controls with overweight/obesity; SREI, self-reported energy intake; TFEI, Three-Factor Eating Inventory; WLM, weight loss maintainers.
Calculated as self-reported EI – DLW TDEE. Negative numbers represent underreporting in kcal/d.
Calculated as [(self-reported EI – DLW TDEE)/DLW TDEE] * 100. Negative numbers represent underreporting as a percentage of TDEE.
Individuals with a self-reported EI/DLW TDEE ratio of <0.674 were classified as underreporters.
FIGURE 2.
Within-group differences between measured and self-reported EI. Results are from paired-samples t test with significance set at P < 0.05. Boxes represent IQR (25–75% of datapoints), and lines within boxes represent median values. WLM (n = 26); NC (n = 33); OC (n = 32). DLW TDEE, total daily energy expenditure measured by doubly labeled water; EI, energy intake; NC, normal-weight controls; OC, controls with overweight/obesity; WLM, weight loss maintainers.
FIGURE 3.
Comparison of absolute and relative self-reported energy intake (EI) accuracy across groups. Nonparametric Wilcoxon rank-sum test used for between-group comparisons. Significance set at P < 0.05. Boxes represent IQR (25–75% of datapoints), and lines within boxes represent median values. The dashed line at 0 kcal/d would be the expected median if participants accurately reported EI. WLM (n = 26); NC (n = 33); OC (n = 32). Absolute self-reported EI accuracy calculated as self-reported EI – DLW TDEE. Relative self-reported EI accuracy calculated as [(self-reported EI – DLW TDEE)/DLW TDEE] * 100. NC, normal-weight controls; OC, controls with overweight/obesity; WLM, weight loss maintainers.
FIGURE 4.
Bland–Altman plot of absolute difference between self-reported EI and DLW TDEE plotted against mean DLW TDEE + self-reported EI. Dashed line at 0 kcal/d would be the expected difference if participants accurately reported EI. Dotted lines represent limits of agreement (mean bias ± 2 SD). Bias calculated as the mean difference between self-reported EI and DLW TDEE. DLW, doubly labeled water; EI, energy intake; LoA, limits of agreement; NC, normal-weight controls; OC, controls with overweight/obesity; TDEE, total daily energy expenditure; WLM, weight loss maintainers.
FIGURE 5.
Proportion of underreporters across groups. Participants with a self-reported energy intake/doubly labeled water total daily energy expenditure ratio <0.674 were classified as underreporters. WLM (n = 26); NC (n = 33); OC (n = 32). NC, normal-weight controls; OC, controls with overweight/obesity; WLM, weight loss maintainers.
Macronutrient intake data (mean ± SD) are presented in Table 2. WLM reported consuming less fat as a percentage of total calories (30.8 ± 7.3%) compared with OC (36.4 ± 5.2%; P < 0.01) but not NC (33.3 ± 6.1%; P = 0.12). WLM reported a higher intake of carbohydrate (49.5 ± 7.2%) compared with OC (44.1 ± 6.8%; P < 0.01) but not NC (47.7 ± 7.2%; P = 0.33). There was no significant difference in protein intake across groups (P = 0.23).
TFEI scores (median, IQR) are presented in Table 2. WLM scored higher in cognitive restraint (16, 12–18) compared with NC (11, 8–13; P < 0.01) and OC (7, 5–10; P < 0.01). WLM scored higher in disinhibition (6, 5–9) than NC (4, 2–6; P = 0.01) but lower than OC (10, 7–14; P < 0.01). WLM experienced greater hunger (5, 3–7) than NC (3, 2–5; P = 0.04) but not OC (5, 4–7; P = 0.56).
Correlations between self-reported EI accuracy and age, sex, BMI, macronutrient consumption, PA measures, PAEE, and TFEI scores are presented in Table 3. There was no significant correlation between self-reported EI accuracy and age, sex, or BMI. There was a small but significant correlation between self-reported EI accuracy and self-reported dietary fat, such that lower reported fat intake was associated with greater underreported EI (ρ = 0.21; P = 0.05). There was a modest negative correlation between self-reported EI accuracy and self-reported protein intake (ρ = –0.45; P < 0.01), such that a higher reported protein intake was associated with greater underreporting of EI. There was no correlation between self-reported EI accuracy and PAEE or any ActivPAL measure of PA. Underreporting of EI was negatively correlated with cognitive restraint (ρ = –0.32; P < 0.01) such that higher scores in that domain were associated with greater underreporting. There was no significant correlation between underreporting of EI and disinhibition (ρ = –0.05; P = 0.62) or hunger (ρ = 0.12; P = 0.25).
TABLE 3.
Correlation between relative self-reported EI accuracy and participant characteristics, PA measures, relative PAEE, macronutrient intake, and TFEI scores1
| Participant characteristics (n = 91) | Spearman's correlation (ρ) | P value |
|---|---|---|
| Age | –0.01 | 0.95 |
| Sex | –0.05 | 0.63 |
| BMI | 0.01 | 0.94 |
| Physical activity measures (n = 76) | ||
| Average step count, steps/d | –0.13 | 0.23 |
| Average sedentary, min/d | 0.01 | 0.91 |
| Average light activity, min/d | 0.00 | 0.97 |
| Average MVPA, min/d | –0.14 | 0.23 |
| Physical activity energy expenditure (n = 87) | ||
| Relative PAEE2 | –0.16 | 0.13 |
| Reported macronutrient intake (n = 91) | ||
| % kcal from fat | 0.21 | 0.05* |
| % kcal from carbohydrate | 0.03 | 0.78 |
| % kcal from protein | –0.45 | <0.01* |
| TFEI scores (n = 91) | ||
| Cognitive restraint of eating score | –0.32 | <0.01* |
| Disinhibition score | –0.05 | 0.62 |
| Hunger score | 0.12 | 0.25 |
Relative self-reported EI accuracy calculated as [(self-reported EI – DLW TDEE)/DLW TDEE] * 100. Spearman's correlation performed using aggregated data from all groups. *Significant P values (α < 0.05) indicated by asterisk. EI, energy intake; MVPA, moderate to vigorous physical activity; PA, physical activity; PAEE, physical activity energy expenditure; TFEI, Three-Factor Eating Inventory.
Relative PAEE measured as [TDEE – (TDEE × 0.1) – REE]/wt (kg).
Discussion
To our knowledge, this is the first study using DLW to compare self-reported EI accuracy in successful weight loss maintainers to normal-weight controls and controls with overweight/obesity. Our primary finding is that WLM underreported EI (absolute and relative) to a greater extent than normal-weight adults of a similar BMI not maintaining weight loss (NC) but not more than adults with overweight or obesity (OC). We also found the proportion of underreporters among WLM was 3 times higher than NC but almost identical to OC. Together, these data suggest that in terms of self-reporting EI accuracy, successful maintainers resemble individuals with obesity more than they resemble normal-weight adults who have never been obese. Our findings call into question the accuracy of previous studies that have relied on self-reported EI data in successful weight loss maintainers.
Numerous studies have shown that WLM perceive dieting to be an integral part of their weight maintenance routine, and they report regulating EI in various ways. Klem et al. (16) found that among a large sample of successful WLM in the National Weight Control Registry (n = 784), 92% self-reported limiting certain foods they ate, 49% self-reported limiting the quantity of food they ate, 38% self-reported limiting the energy they consumed from fat, 35% self-reported counting calories, and 30% self-reported counting fat grams. Self-reported EI (block FFQ) in the National Weight Control Registry has been reported to be ∼1400 kcal/d, further adding to the common assumption that successful maintainers chronically restrict EI (16). However, our finding that DLW TDEE in WLM, which can be interpreted to represent EI in weight-stable individuals (36), was similar to that of OC but significantly higher than NC extends our previous work in this area and further calls into question the accuracy of self-reported calorie restriction frequently cited in this population. It also underscores the importance of obtaining a thorough weight history in both clinical and research settings and suggests that without a weight history, self-reported EI should not be assumed to be modest in individuals with a BMI in the healthy range.
Our finding of underreported EI in WLM also calls into question the accuracy of self-reported macronutrient intake. Evidence suggests that individuals are more likely to underreport carbohydrate and fat (42, 43), whereas nitrogen balance studies show that protein is likely to be correctly reported or even overreported (42). Although our study was not designed to assess the accuracy of macronutrient intake, it is possible that WLM disproportionately underreported foods they perceived as unhealthy (e.g., that were higher in fat). Our correlation analysis supports this hypothesis in that lower self-reported fat intake was associated with greater underreporting of EI (Table 3). Surprisingly, higher self-reported protein intake was also associated with greater underreporting, and further investigation is needed to give context to this novel finding. Although new technologies, such as digital photography, are emerging (44), it remains challenging to objectively measure macronutrient intake in free-living conditions, and further research is needed before the self-reported macronutrient intakes of WLM should be used as a guide for clinical weight management.
The results of our study also align with a growing body of evidence that WLM rely on high levels of PA as their primary means for maintaining long-term energy balance at a reduced body weight. In a case-control study using a similar cohort of participants as the present study, Ostendorf et al. (24) compared device-measured PA and found that WLM engaged in ∼30 min/d more MVPA, ∼56 min/d more LPA, and ∼58 min/d less time sedentary compared to OC. Kerns et al. (45) reported similar results in participants from the televised series The Biggest Loser who were followed for 6 y after completion of the competition. In that study, successful maintainers had higher PAEE (12.2 kcal/kg · d) compared with weight regainers (8.0 kcal/kg · d, P = 0.04), and importantly, using the DLW intake balance method to objectively estimate EI, there was no significant difference in changes in EI over the 6 y between the 2 groups (−8.7 ± 5.6% compared with −7.4 ± 2.7%, respectively; P = 0.83). Our finding that WLM underreport EI while consuming significantly more kcal than NC highlights the importance of relying on objective measures of EI to understand how this population achieves long-term energy balance.
Although most WLM underreported EI, some reported greater intake compared with DLW TDEE, in one instance by nearly 500 kcal/d (Figure 4). This heterogeneity suggests that at the individual level, WLM may employ different strategies to achieve energy balance. One possibility is that individuals who are most likely to underreport EI (and therefore potentially consume more calories) might also be more likely to engage in high levels of PA as a strategy to offset the greater EI. However, we found no correlation between underreporting EI and any measure of PA or PAEE (Table 3).
Although several underlying causes for biased reporting have been proposed (46–50), it remains unclear the degree to which psychosocial factors may influence this phenomenon. Previous studies have found that cognitive restraint tends to be higher in populations that underreport food intake (10, 46, 49). Our results concur with these previous data in that WLM scored higher in cognitive restraint than NC and showed a greater tendency to underreport food intake. Furthermore, using aggregated TFEI data, we found a significant association between higher cognitive restraint and an individual propensity to underreport EI that existed independent of group assignment. More research is needed to identify the possible mechanisms that link increased cognitive restraint to underreporting of EI.
Our study has several limitations. The case-control design limits our ability to assess longitudinal changes that may have occurred in TDEE and self-reported EI. We also relied on previously published CVTDEE to estimate underreporting, which may have introduced error if that value differed from the true CVTDEE in our sample. Food logs were not reviewed with each participant at the time of submission, which might have led us to miss obvious errors in recording accuracy. Therefore, intentional misreporting of EI is unlikely to be the sole reason for underreporting in this study, and other root causes, such as mismeasured portion sizes and forgotten snack/beverage consumption, are likely contributors that our study design was unable to detect. Finally, our WLM sample size was relatively small and homogeneous, which limits the generalizability of our findings to the larger population of successful maintainers.
In conclusion, although all 3 groups underreported EI, WLM underreported to a greater extent than NC, but not OC, and had a greater proportion of underreporters compared with NC, but not OC. These findings call into question the accuracy of previously published self-reported EI data in WLM. They also align with recent data suggesting that WLM rely less on chronic restriction of EI and more on higher levels of PA to achieve long-term energy balance. WLM reported lower fat intake than OC and higher cognitive restraint than either NC or OC, which were both significantly correlated with an increased propensity to underreport EI. Our results may have practical implications for individuals who have successfully lost weight and are transitioning into the weight maintenance phase.
Acknowledgements
The authors’ contributions were as follows—VAC: designed the study, conducted the research, and acquired the data; ZP and JHD: performed the statistical analysis; JHD: wrote the manuscript and had primary responsibility for final content; DMO, AZ, EM, and VAC: provided significant edits; and all authors: have read and approved the manuscript.
Author disclosure: The authors report no conflicts of interest.
Notes
This work was supported by grants from the NIH (K23 DK078913, P30 DK048520, UL1 TR002535) and National Institute of Diabetes and Digestive and Kidney Diseases (F32 DK122652). The contents do not represent the views of the US Department of Veterans Affairs or the US government.
Abbreviations used: CVEI, within-participant coefficient of variation for self-reported energy intake; CVTDEE, within-participant coefficient of variation for total daily energy expenditure; DLW, doubly labeled water; EI, energy intake; FFM, fat-free mass; FM, fat mass; LPA, light-intensity physical activity; MET, metabolic equivalent; MVPA, moderate to vigorous physical activity; NC, normal weight control; OC, control with overweight/obesity; PA, physical activity; PAEE, physical activity energy expenditure; QC, quality control; REE, resting energy expenditure; TBW, total body water; TDEE, total daily energy expenditure; TFEI, Three-Factor Eating Inventory; WLM, weight loss maintainers.
Contributor Information
Jared H Dahle, Graduate School, Integrated Physiology Program, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
Danielle M Ostendorf, Department of Medicine, Anschutz Health and Wellness Center, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
Adnin Zaman, Department of Medicine, Division of Endocrinology, Metabolism, and Diabetes, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
Zhaoxing Pan, Department of Biostatistics and Informatics, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
Edward L Melanson, Department of Medicine, Division of Endocrinology, Metabolism, and Diabetes, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
Victoria A Catenacci, Department of Medicine, Anschutz Health and Wellness Center, University of Colorado Anschutz Medical Campus, Aurora, CO, USA; Department of Medicine, Division of Endocrinology, Metabolism, and Diabetes, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
Data Availability
Data described in the manuscript will be made available upon request pending application to Vicki.Catenacci@cuanschutz.edu.
References
- 1. Centers for Disease Control and Prevention . National Health and Nutrition Examination Survey Questionnaire. Hyattsville (MD): US Department of Health and Human Services, Centers for Disease Control and Prevention; 2015–2016. [Google Scholar]
- 2. Naska A, Lagiou A, Lagiou P. Dietary assessment methods in epidemiological research: current state of the art and future prospects. F1000Res. 2017;6:926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hill RJ, Davies PSW. The validity of self-reported energy intake as determined using the doubly labelled water technique. Br J Nutr. 2001;85(4):415–30. [DOI] [PubMed] [Google Scholar]
- 4. Walker JL, Ardouin S, Burrows T. The validity of dietary assessment methods to accurately measure energy intake in children and adolescents who are overweight or obese: a systematic review. Eur J Clin Nutr. 2018;72(2):185–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dhurandhar NV, Schoeller D, Brown AW, Heymsfield SB, Thomas D, Sørensen TIA, Speakman JR, Jeansonne M, Allison DB; Energy Balance Measurement Working Group. Energy balance measurement: when something is not better than nothing. Int J Obes. 2015;39(7):1109–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Burrows TL, Ho YY, Rollo ME, Collins CE. Validity of dietary assessment methods when compared to the method of doubly labeled water: a systematic review in adults. Front Endocrinol. 2019;10(850):850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Lichtman SW, Pisarska K, Berman ER, Pestone M, Dowling H, Offenbacher E, Weisel H, Heshka S, Matthews DE, Heymsfield SB. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N Engl J Med. 1992;327(27):1893–8. [DOI] [PubMed] [Google Scholar]
- 8. Prentice AM, Black AE, Coward WA, Davies HL, Goldberg GR, Murgatroyd PR, Ashford J, Sawyer M, Whitehead RG. High levels of energy expenditure in obese women. BMJ. 1986;292(6526):983–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Catenacci VA, Odgen L, Phelan S, Thomas JG, Hill J, Wing RR, Wyatt H. Dietary habits and weight maintenance success in high versus low exercisers in the National Weight Control Registry. J Phys Act Health. 2014;11(8):1540–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Black AE, Jebb SA, Bingham SA, Runswick SA, Poppitt SD. The validation of energy and protein intakes by doubly labelled water and 24-hour urinary nitrogen excretion in post-obese subjects. J Hum Nutr Diet. 1995;8(1):51–64. [Google Scholar]
- 11. Hill JO, Wyatt H, Phelan S, Wing R. The National Weight Control Registry: is it useful in helping deal with our obesity epidemic?. J Nutr Educ Behav. 2005;37(4):206–10. [DOI] [PubMed] [Google Scholar]
- 12. Phelan S, Wyatt H, Nassery S, DiBello J, Fava JL, Hill JO, Wing RR. Three-year weight change in successful weight losers who lost weight on a low-carbohydrate diet. Obesity. 2007;15(10):2470–7. [DOI] [PubMed] [Google Scholar]
- 13. Shariaty S, Bahonaran A, Ayremlou P, Hoseinalizdeh L, Alizadeh M. Comparison of dietary patterns, food groups, nutrients intake, cardio-metabolic biomarkers, and liver enzymes in successful and unsuccessful weight loss maintainers. Obesity Medicine. 2019;16:100147. [Google Scholar]
- 14. Wyatt HR, Grunwald GK, Mosca CL, Klem ML, Wing RR, Hill JO. Long-term weight loss and breakfast in subjects in the national weight control registry. Obes Res. 2002;10(2):78–82. [DOI] [PubMed] [Google Scholar]
- 15. Colvin RH, Olson SB. A descriptive analysis of men and women who have lost significant weight and are highly successful at maintaining the loss. Addict Behav. 1983;8(3):287–95. [DOI] [PubMed] [Google Scholar]
- 16. Klem ML, Wing RR, McGuire MT, Seagle HM, Hill JO. A descriptive study of individuals successful at long-term maintenance of substantial weight loss. Am J Clin Nutr. 1997;66(2):239–46. [DOI] [PubMed] [Google Scholar]
- 17. Shick SM, Wing RR, Klem ML, McGuire MT, Hill JO, Seagle H. Persons successful at long-term weight loss and maintenance continue to consume a low-energy, low-fat diet. J Am Diet Assoc. 1998;98(4):408–13. [DOI] [PubMed] [Google Scholar]
- 18. Greene LF, Malpede CZ, Henson CS, Hubbert KA, Heimburger DC, Ard JD. Weight maintenance 2 years after participation in a weight loss program promoting low-energy density foods. Obesity. 2006;14(10):1795–801. [DOI] [PubMed] [Google Scholar]
- 19. Thomas JG, Bond DS, Phelan S, Hill JO, Wing RR. Weight-loss maintenance for 10 years in the National Weight Control Registry. Am J Prev Med. 2014;46(1):17–23. [DOI] [PubMed] [Google Scholar]
- 20. Catenacci VA, Ogden LG, Stuht J, Phelan S, Wing RR, Hill JO, Wyatt HR. Physical activity patterns in the National Weight Control Registry. Obesity (Silver Spring). 2008;16(1):153–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Jakicic JM, Marcus BH, Lang W, Janney C. Effect of exercise on 24-month weight loss maintenance in overweight women. Arch Intern Med. 2008;168(14):1550–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Tate DF, Jeffery RW, Sherwood NE, Wing RR. Long-term weight losses associated with prescription of higher physical activity goals: are higher levels of physical activity protective against weight regain?. Am J Clin Nutr. 2007;85(4):954–9. [DOI] [PubMed] [Google Scholar]
- 23. Schoeller DA, Ravussin E, Schutz Y, Acheson KJ, Baertschi P, Jequier E. Energy expenditure by doubly labeled water: validation in humans and proposed calculation. Am J Physiol. 1986;250(5):R823–R30. [DOI] [PubMed] [Google Scholar]
- 24. Ostendorf DM, Lyden K, Pan Z, Wyatt HR, Hill JO, Melanson EL, Catenacci VA. Objectively measured physical activity and sedentary behavior in successful weight loss maintainers. Obesity (Silver Spring). 2018;26(1):53–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Ostendorf DM, Caldwell AE, Creasy SA, Pan Z, Lyden K, Bergouignan A, MacLean PS, Wyatt HR, Hill JO, Melanson ELet al. Physical activity energy expenditure and total daily energy expenditure in successful weight loss maintainers. Obesity (Silver Spring). 2019;27(3):496–504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Drenowatz C, Hill JO, Peters JC, Soriano-Maldonado A, Blair SN. The association of change in physical activity and body weight in the regulation of total energy expenditure. Eur J Clin Nutr. 2017;71(3):377–82. [DOI] [PubMed] [Google Scholar]
- 27. Black AE, Bingham SA, Johansson G, Coward WA. Validation of dietary intakes of protein and energy against 24 hour urinary N and DLW energy expenditure in middle-aged women, retired men and post-obese subjects: comparisons with validation against presumed energy requirements. Eur J Clin Nutr. 1997;51(6):405–13. [DOI] [PubMed] [Google Scholar]
- 28. Weir J. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol. 1949;109(1–2):1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Ostendorf DM, Melanson EL, Caldwell AE, Creasy SA, Pan Z, MacLean PS, Wyatt HR, Hill JO, Catenacci VA. No consistent evidence of a disproportionately low resting energy expenditure in long-term successful weight-loss maintainers. Am J Clin Nutr. 2018;108(4):658–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Feskanich D, Sielaff BH, Chong K, Buzzard IM. Computerized collection and analysis of dietary intake information. Comput Methods Programs Biomed. 1989;30(1):47–57. [DOI] [PubMed] [Google Scholar]
- 31. Stunkard AJ, Messick S. The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosom Res. 1985;29(1):71–83. [DOI] [PubMed] [Google Scholar]
- 32. Cappelleri JC, Bushmakin AG, Gerber RA, Leidy NK, Sexton CC, Lowe MR, Karlsson J. Psychometric analysis of the Three-Factor Eating Questionnaire-R21: results from a large diverse sample of obese and non-obese participants. Int J Obes. 2009;33(6):611–20. [DOI] [PubMed] [Google Scholar]
- 33. Laessle RG, Tuschl RJ, Kotthaus BC, Pirke KM. A comparison of the validity of three scales for the assessment of dietary restraint. J Abnorm Psychol. 1989;98(4):504–7. [DOI] [PubMed] [Google Scholar]
- 34. Lawson OJ, Williamson DA, Champagne CM, DeLany JP, Brooks ER, Howat PM, Wozniak PJ, Bray GA, Ryan DH. The association of body weight, dietary intake, and energy expenditure with dietary restraint and disinhibition. Obes Res. 1995;3(2):153–61. [DOI] [PubMed] [Google Scholar]
- 35. Racette SB, Schoeller DA, Luke AH, Shay K, Hnilicka J, Kushner RF. Relative dilution spaces of 2H- and 18O-labeled water in humans. Am J Physiol Endocrinol Metab. 1994;267(4):E585–E90. [DOI] [PubMed] [Google Scholar]
- 36. Trabulsi J, Schoeller DA. Evaluation of dietary assessment instruments against doubly labeled water, a biomarker of habitual energy intake. Am J Physiol Endocrinol Metab. 2001;281(5):E891–9. [DOI] [PubMed] [Google Scholar]
- 37. Schoeller DA, Jefford G. Determinants of the energy costs of light activities: inferences for interpreting doubly labeled water data. Int J Obes. 2002;26(1):97–101. [DOI] [PubMed] [Google Scholar]
- 38. van der Berg JD, Willems PJ, van der Velde JH, Savelberg HH, Schaper NC, Schram MT, Sep SJ, Dagnelie PC, Bosma H, Stehouwer CDet al. Identifying waking time in 24-h accelerometry data in adults using an automated algorithm. J Sports Sci. 2016;34(19):1867–73. [DOI] [PubMed] [Google Scholar]
- 39. Black AE, Cole TJ. Within- and between-subject variation in energy expenditure measured by the doubly-labelled water technique: implications for validating reported dietary energy intake. Eur J Clin Nutr. 2000;54(5):386–94. [DOI] [PubMed] [Google Scholar]
- 40. Scagliusi FB, Ferriolli E, Pfrimer K, Laureano C, Cunha CS, Gualano B, Lourenço BH, Lancha AH Jr. Characteristics of women who frequently under report their energy intake: a doubly labelled water study. Eur J Clin Nutr. 2009;63(10):1192–9. [DOI] [PubMed] [Google Scholar]
- 41. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet North Am Ed. 1986;327(8476):307–10. [PubMed] [Google Scholar]
- 42. Macdiarmid J, Blundell J. Assessing dietary intake: who, what and why of under-reporting. Nutr Res Rev. 1998;11:231–53. [DOI] [PubMed] [Google Scholar]
- 43. Poppitt SD, Swann D, Black AE, Prentice AM. Assessment of selective under-reporting of food intake by both obese and non-obese women in a metabolic facility. Int J Obes. 1998;22(4):303–11. [DOI] [PubMed] [Google Scholar]
- 44. Martin CK, Correa JB, Han H, Allen HR, Rood JC, Champagne CM, Gunturk BK, Bray GA. Validity of the Remote Food Photography Method (RFPM) for estimating energy and nutrient intake in near real-time. Obesity (Silver Spring). 2012;20(4):891–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Kerns JC, Guo J, Fothergill E, Howard L, Knuth ND, Brychta R, Chen KY, Skarulis MC, Walter PJ, Hall KD. Increased physical activity associated with less weight regain six years after “The Biggest Loser” competition. Obesity. 2017;25(11):1838–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Bathalon GP, Tucker KL, Hays NP, Vinken AG, Greenberg AS, McCrory MA, Roberts SB. Psychological measures of eating behavior and the accuracy of 3 common dietary assessment methods in healthy postmenopausal women. Am J Clin Nutr. 2000;71(3):739–45. [DOI] [PubMed] [Google Scholar]
- 47. Becker W, Foley S, Shelley E, Gibney M. Energy under-reporting in Swedish and Irish dietary surveys: implications for food-based dietary guidelines. Br J Nutr. 1999;81(Suppl 1):S127–S31. [DOI] [PubMed] [Google Scholar]
- 48. Bellisle F, Blundell JE, Dye L, Fantino M, Fern E, Fletcher RJ, Lambed J, Roberfroid M, Specter S, Westenhöfer Jet al. Functional food science and behaviour and psychological functions. Br J Nutr. 1998;80(Suppl 1):S173–S93. [DOI] [PubMed] [Google Scholar]
- 49. Lafay L, Basdevant A, Charles MA, Vray M, Balkau B, Borys JM, Eschwège E, Romon M. Determinants and nature of dietary underreporting in a free-living population: the Fleurbaix Laventie Ville Santé (FLVS) study. Int J Obes. 1997;21(7):567–73. [DOI] [PubMed] [Google Scholar]
- 50. Price GM, Paul AA, Cole TJ, Wadsworth MEJ. Characteristics of the low-energy reporters in a longitudinal national dietary survey. Br J Nutr. 1997;77(6):833–51. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data described in the manuscript will be made available upon request pending application to Vicki.Catenacci@cuanschutz.edu.