Under-Reporting in Food Diaries: What the Doubly Labeled Water Literature Shows

Keywords: doubly labeled water; dietary under-reporting; food diaries; energy intake; total energy expenditure; dietary assessment; 24-hour recall; food frequency questionnaire

Last updated: October 2026

1. What is doubly labeled water, and why is it used to check food diaries?

Doubly labeled water (DLW) measures how much energy a person uses in everyday life without asking them to report anything. The person drinks water enriched with deuterium and oxygen-18; the isotopes leave the body at different rates, the difference reflects carbon dioxide production, and total energy expenditure is calculated from it. Urine, saliva or blood samples are collected at the start and end of an observation period of one to three weeks [1]. First published in 1955 and applied to people from 1982, it has been called the gold standard for human energy requirement under daily living conditions [1]. In a 1982 validation in four adults eating a measured diet for 13 days, it differed from dietary intake plus change in body composition by 2% on average (coefficient of variation 6%) [2].

It works as a yardstick for diaries because a person whose weight is stable eats about as much energy as they use. At group level, weight can be treated as stable over a dietary assessment, so mean reported intake can be compared with mean DLW expenditure [3]. DLW measures energy used, not food eaten, so it cannot show which foods went unreported.

2. How much do food diaries and recalls under-report?

By a meaningful margin on average, though it varies by study. A 2019 systematic review of 59 studies comparing self-reported intake with DLW expenditure in 6,298 free-living adults found that most reported significant under-reporting [4]. For food records, 19 studies found significant under-reporting of 11% to 41%, three found no significant difference and one found 8% over-reporting [4]. The pattern was already visible in studies from the 1980s. In 1986, self-recorded intakes were accurate in lean women but 3.5 MJ/day (837 kcal/day; 1 MJ is about 239 kcal) below measured expenditure in women with obesity [5]. In 1990, seven-day weighed records from 31 adults averaged 9.66 MJ/day against a concurrent expenditure of 12.15 MJ/day [6]. Larger studies give figures by instrument:

• Food records and diaries. Intake was underestimated by 18–21% on unweighed four-day paper records in IDATA (the Interactive Diet and Activity Tracking in AARP study) of adults aged 50–74 [7]. A web-based seven-day diary averaged 9.5 MJ/day against expenditure of 11.5 MJ/day in 120 Danish volunteers [8]. In 221 UK adults, four-day written diaries averaged about 1,830 kcal/day against about 2,760 kcal/day of expenditure (Table 1 of that study) [9].
• 24-hour recalls. Under-reporting was 12–14% in men and 16–20% in women in the Observing Protein and Energy Nutrition (OPEN) study of 484 adults [10], 11% overall in a USDA study of 524 adults [11], and 15–17% for automated self-administered recalls (ASA24) in IDATA [7]. Two interviewer-administered recalls matched expenditure on average in the Danish study [8].
• Food frequency questionnaires. About 28% in a pooled analysis of five US validation studies [12], 29–34% in IDATA [7] and 31–38% in OPEN [10].

3. Does the shortfall depend on body weight, sex or age?

Body weight is the factor most often linked to a larger shortfall; sex and age matter less consistently. In the USDA study, normal-weight participants (body mass index, BMI, under 25) under-reported by less than 3%, while the shortfall and the share of low reporters were highest in people with obesity (BMI over 30) [11]. The pooled analysis found BMI, education and age related to under-reporting, with BMI a strong predictor [12]. A 2025 equation built from 6,497 DLW measurements to predict expenditure from easily measured variables such as weight, age and sex was applied to UK and US national surveys; the gap between reported intake and predicted expenditure widened with BMI, was absent at a BMI of about 15–20, and changed little with age in adults [13]. Women under-reported more than men on recalls in OPEN [10], and the 2019 review found misreporting more frequent in women for recall-based methods [4].

Not every analysis agrees. In the 221 UK adults who kept four-day diaries, absolute under-reporting was greater with obesity, but the effect of body mass disappeared when intake and expenditure were scaled allometrically; the authors concluded that in weight-stable adults people with obesity do not under-report more [9].

4. Is the shortfall under-eating, forgetting or something else?

Several mechanisms operate together. In 30 men with obesity (mean BMI 34) who kept a food diary for a week while expenditure was measured with DLW, reported intake averaged 37% below expenditure; the authors attributed about 26% to under-eating, judged from weight loss during the recording week, and about 12% to under-recording, judged from water balance [14]. In a six-person focus group after the Danish study, participants said they did not eat foods that were cumbersome to record, forgot to log second portions and became more aware of what they ate [8]. The 2025 analysis lists difficulty estimating amounts, fallible memory, deliberate falsification in some cases and changed intake while recording among the sources of misreporting [13].

Which foods go missing is less settled. In a 24-hour metabolic-facility study of 33 women whose intake was measured covertly, meals were reported accurately but between-meal snacks were significantly under-reported; reported carbohydrate and added sugar were significantly lower than measured, while fat and protein were not [15]. That study did not use DLW. Among DLW studies, OPEN found little under-reporting of the percentage of energy from protein [10], whereas the study of men with obesity found fat selectively under-reported [14].

5. Is the gap the same for everyone?

No. In OPEN, 9% of men and 7% of women were classed as under-reporters of both energy and protein on recalls, against 35% of men and 23% of women on questionnaires [10]. In the USDA study about 78% of men and 74% of women were acceptable energy reporters [11]. In the Danish study 34% of participants were classed as under-reporters on the seven-day diary against 4% on the two recalls [8]. In the 1990 weighed-record study, intake-to-expenditure ratios were close to 1.0 in the highest third of recorded intakes and between 0.61 and 0.89 in the lower thirds [6].

6. What do researchers do about under-reporting?

Researchers screen reports for plausibility and adjust study design and analysis. In 1991, Goldberg and colleagues used whole-body calorimeter and DLW measurements to derive cut-offs, expressed as multiples of basal metabolic rate, below which a reported intake is incompatible with long-term energy balance [16]. The 2025 analysis argues that such cut-offs catch only very low reports: the Goldberg cut-off indicated 16.7% under-reporting and a modified version 23.4%, both far less than its DLW-based equation did [13].

Design and analysis help too. Repeated recalls raised correlations with true protein intake from 0.40 with one recall to 0.49 with three in the pooled analysis, where calibration equations that included personal characteristics also improved prediction [12]. Averaging several days did not reduce implausible reports in either national survey, which the 2025 authors read as consistent under-reporting rather than day-to-day variation [13]. Comparing nutrient densities rather than absolute intakes can also help: in IDATA, mean protein and sodium densities were similar to biomarker values on every instrument, though potassium density from questionnaires ran 26–40% high [7].

7. What do these studies not show?

DLW cannot show which foods were missed, and these studies do not test whether keeping a diary helps people manage weight or health. DLW gives average expenditure over one to three weeks [1], equals intake only when weight is stable [3] and carries its own error [2]. Many cohorts were volunteers: IDATA recruited AARP members around one US city [7], the Danish sample answered advertisements [8], and the UK diary sample volunteered for a dietary survey, may have been health-conscious and kept diaries two to three weeks before the DLW measurement [9]. Anyone using a food diary for weight management, a medical condition or recovery from disordered eating should discuss it with a doctor or registered dietitian.

For related evidence in this journal, see the 2026 systematic review of weight-loss intervention categories, the evidence hierarchy for physical activity, nutrition and weight loss and the 2024 meta-analysis of tracking accuracy and weight-management outcomes.

Frequently asked questions

How accurate is the doubly labeled water method?

In a 1982 validation in four adults, DLW expenditure differed from measured intake plus change in body composition by 2% on average (coefficient of variation 6%) [2]. Individual results remain estimates.

Are food diaries accurate?

Not for absolute energy intake, on average: in a 2019 systematic review, 19 studies of food records found significant under-reporting of 11% to 41% against DLW [4].

Do people lie on food diaries?

These studies measure the size of the gap, not intent. The 2025 analysis lists estimation difficulty, fallible memory, deliberate falsification in some cases and changed intake while recording [13].

Do people with obesity under-report more?

In most large studies, on average, yes [11, 12]; one analysis of 221 UK adults found no difference after allometric scaling for body mass [9].

Is a 24-hour recall more accurate than a food diary?

For group-average energy intake, often closer: in the Danish study two recalls matched measured expenditure while a seven-day diary fell about 2 MJ/day short [8]; in IDATA the difference was small [7].

References

1. [1]Westerterp KR. Doubly labelled water assessment of energy expenditure: principle, practice, and promise. Eur J Appl Physiol. 2017;117(7):1277–1285. doi:10.1007/s00421-017-3641-x
2. [2]Schoeller DA, van Santen E. Measurement of energy expenditure in humans by doubly labeled water method. J Appl Physiol Respir Environ Exerc Physiol. 1982;53(4):955–959. doi:10.1152/jappl.1982.53.4.955
3. [3]Livingstone MB, Black AE. Markers of the validity of reported energy intake. J Nutr. 2003;133(Suppl 3):895S–920S. doi:10.1093/jn/133.3.895S
4. [4]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. doi:10.3389/fendo.2019.00850
5. [5]Prentice AM, Black AE, Coward WA, et al. High levels of energy expenditure in obese women. Br Med J (Clin Res Ed). 1986;292(6526):983–987. doi:10.1136/bmj.292.6526.983
6. [6]Livingstone MB, Prentice AM, Strain JJ, et al. Accuracy of weighed dietary records in studies of diet and health. BMJ. 1990;300(6726):708–712. doi:10.1136/bmj.300.6726.708
7. [7]Park Y, Dodd KW, Kipnis V, et al. Comparison of self-reported dietary intakes from the Automated Self-Administered 24-h recall, 4-d food records, and food-frequency questionnaires against recovery biomarkers. Am J Clin Nutr. 2018;107(1):80–93. doi:10.1093/ajcn/nqx002
8. [8]Biltoft-Jensen A, Ygil KH, Knudsen L, et al. Validation of the 2 × 24 h recall method and a 7-d web-based food diary against doubly labelled water in Danish adults. Br J Nutr. 2023;130(8):1444–1457. doi:10.1017/S0007114523000454
9. [9]Waterworth SP, Kerr CJ, McManus CJ, et al. Obese individuals do not underreport dietary intake to a greater extent than nonobese individuals when data are allometrically-scaled. Am J Hum Biol. 2022;34(7):e23743. doi:10.1002/ajhb.23743
10. [10]Subar AF, Kipnis V, Troiano RP, et al. Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study. Am J Epidemiol. 2003;158(1):1–13. doi:10.1093/aje/kwg092
11. [11]Moshfegh AJ, Rhodes DG, Baer DJ, et al. The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. Am J Clin Nutr. 2008;88(2):324–332. doi:10.1093/ajcn/88.2.324
12. [12]Freedman LS, Commins JM, Moler JE, et al. Pooled results from 5 validation studies of dietary self-report instruments using recovery biomarkers for energy and protein intake. Am J Epidemiol. 2014;180(2):172–188. doi:10.1093/aje/kwu116
13. [13]Bajunaid R, Niu C, Hambly C, et al. Predictive equation derived from 6,497 doubly labelled water measurements enables the detection of erroneous self-reported energy intake. Nat Food. 2025;6(1):58–71. doi:10.1038/s43016-024-01089-5
14. [14]Goris AH, Westerterp-Plantenga MS, Westerterp KR. Undereating and underrecording of habitual food intake in obese men: selective underreporting of fat intake. Am J Clin Nutr. 2000;71(1):130–134. doi:10.1093/ajcn/71.1.130
15. [15]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 Relat Metab Disord. 1998;22(4):303–311. doi:10.1038/sj.ijo.0800584
16. [16]Goldberg GR, Black AE, Jebb SA, et al. Critical evaluation of energy intake data using fundamental principles of energy physiology: 1. Derivation of cut-off limits to identify under-recording. Eur J Clin Nutr. 1991;45(12):569–581. PMID: 1810719