Misreporting of dietary intake is a universal phenomenon that appears to occur both randomly and non-randomly( Reference Livingstone and Black 1 – Reference Mattisson, Wirfalt and Aronsson 3 ). Furthermore, it may be selective for different kinds of foods and nutrients( Reference Poppitt, Swann and Black 4 – Reference Freedman, Commins and Moler 6 ), although without biomarkers for each food or nutrient of interest this is hard to articulate with absolute certainty, and may differ by population. Biases inherent in the use of self-reported dietary data make it complicated to interpret studies on diet and health, which may distort or obscure the associations between diet and health or even create spurious ones( Reference Livingstone and Black 1 , Reference Mattisson, Wirfalt and Aronsson 3 , Reference Rosell, Hellenius and De Faire 5 ). To better understand this issue, it is essential to identify the characteristics associated with misreporting (under-reporting and over-reporting) of dietary intake.
As all nutrients must be provided within the quantity of food needed to fulfil the energy requirement, energy intake (EI) is the foundation of the diet( Reference Livingstone and Black 1 ). Unfortunately, under-reporting of EI has long been a serious problem in almost all dietary surveys( Reference Livingstone and Black 1 , Reference Freedman, Commins and Moler 6 ). In particular, overweight and obese subjects tend to under-report EI to a greater extent than normal-weight subjects( Reference Livingstone and Black 1 – Reference Freedman, Commins and Moler 6 ). Moreover, recent studies have shown that, in addition to under-reporting, over-reporting of EI also needs to be taken into account, in some populations at least, such as those with low BMI( Reference Mattisson, Wirfalt and Aronsson 3 , Reference Lutomski, van den Broeck and Harrington 7 , Reference Johansson, Solvoll and Bjorneboe 8 ). Investigation of dietary misreporting should be conducted in each country, as it is conceivable that the way in which survey participants comply with dietary assessment procedures may differ from one country to another. Nevertheless, information on the whole picture of characteristics associated with dietary misreporting in a representative sample in each country is still limited( Reference Lutomski, van den Broeck and Harrington 7 – Reference Archer, Hand and Blair 14 ).
In the continuous National Health and Nutrition Examination Survey (NHANES), the US Department of Agriculture (USDA) Automated Multiple-Pass Method is used for collecting 24-h dietary recall information. Although this method has been validated against total energy expenditure measured by doubly labelled water( Reference Moshfegh, Rhodes and Baer 15 , Reference Blanton, Moshfegh and Baer 16 ) and against observed actual intake( Reference Conway, Ingwersen and Vinyard 17 , Reference Conway, Ingwersen and Moshfegh 18 ) in highly selected populations, the validity in a representative sample of US adults remains largely unknown. In the present study, the prevalence and characteristics of under-reporting and over-reporting of EI among US adults were evaluated using data from the NHANES.
Methods
Survey design
The present cross-sectional analysis was based on public domain data from NHANES, a continuing population-based survey that uses a complex, stratified multi-stage probability sample design to create a representative sample of the non-institutionalised civilian US population( Reference Zipf, Chiappa and Porter 19 , Reference Johnson, Paulose-Ram and Ogden 20 ). Initiated in 1999, the survey examines about 5000 persons each year, and the data are released every 2 years. Each survey consists of questionnaires administered at home, followed by a standardised health examination, including an in-person 24-h dietary recall interview, in a mobile examination centre. Since 2002, a second 24-h dietary recall was also obtained by telephone; two 24-h dietary recall data are publicly available since 2003. The unweighted response rates for the examined persons for NHANES 1999–2000, 2001–2002, 2003–2004, 2005–2006, 2007–2008, 2009–2010 and 2011–2012 were 76, 80, 76, 77, 75, 77 and 70 %, respectively( 21 ). The documentation and data for each of these surveys can be downloaded from the NHANES website( 22 ). The NHANES was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving human subjects were approved by National Center for Health Statistics Research Ethics Review Board. Written informed consent was obtained from each subject.
Analytical sample
The analytical sample was limited to adults aged ≥20 years with two complete and reliable, self-reported, 24-h dietary recall data (n 21 921). After excluding pregnant (n 618) and lactating (n 153) respondents, as well as those with missing information on the variables of interest (n 1754), the final analytical sample included 19 396 respondents from NHANES 2003–2012. An additional analysis was also conducted using only the first dietary recall data in 28 794 respondents from NHANES 1999–2012.
Assessment of energy intake
All surveys collected dietary information using a 24-h dietary recall administered by a trained interviewer in the mobile examination centre. Beginning with 2002, a second 24-h dietary recall was also obtained via telephone 3–10 d after the first recall. The dietary recalls collected for the NHANES 1999–2000 and 2001 survey years used a computer-assisted interview that included a 4-step multiple pass approach. Since 2002, the dietary data were collected using an automated 5-step multiple pass approach – namely, the USDA Automated Multiple-Pass Method( Reference Moshfegh, Rhodes and Baer 15 – Reference Conway, Ingwersen and Moshfegh 18 , 22 ). This method consists of (1) a quick list pass, in which the respondent is asked to list everything eaten or drunk the previous day; (2) a forgotten foods list pass, in which a standard list of foods or beverages, often forgotten, is read to prompt recall; (3) a time and occasion pass, in which the time of and the name for the eating occasion are collected; (4) a detail and review pass, in which detailed descriptions and portion sizes are collected and the time interval between meals is reviewed to check for additional food intake; and (5) the final probe pass, one last opportunity to remember foods consumed. Estimates of EI from all reported foods and beverages were calculated by using the USDA food composition databases. In 1999–2000, the USDA 1994–1998 Survey Nutrient Database was the food composition database used; in subsequent surveys, the USDA Food and Nutrient Database for Dietary Studies was used( 22 ). The average of EI over the 2 d for each participant was used for the present analysis.
Assessment of non-dietary variables
Consistent with NHANES sample-selection methods, age was categorised as 20–39, 40–59 and ≥60 years. Race/ethnicity was categorised as non-Hispanic white, non-Hispanic black, Mexican American and others. As indicators of socio-economic status, we considered family income as a percentage of the federal poverty threshold and years of education. The family poverty income ratio was categorised as <130, 130–349 and ≥350 %. The educational level was categorised as <12 years, 12 years, some college and college degree or more. Information on smoking status (never, former or current) and perceived weight status (underweight, about the right weight or overweight) was also collected. Based on self-report of either any moderate or vigorous activities lasting ≥10 min in the past 30 d (NHANES 1999–2006) or without a specified period (NHANES 2007–2012), any recreational physical activity (yes or no) was assessed. Body weight and height were measured by trained interviewers using standardised procedures with calibrated equipments. BMI (kg/m2) was calculated as weight (kg) divided by height (m) squared. Weight status was defined based on BMI according to World Health Organization( 23 ) recommendations as follows: underweight (<18·5 kg/m2), normal (≥18·5 to <25 kg/m2), overweight (≥25 to <30 kg/m2) and obese (≥30 kg/m2).
Evaluation of the accuracy of energy intake reporting
Misreporting of EI was evaluated based on the ratio of EI to BMR (the Goldberg cut-off)( Reference Black 24 ) and the ratio of EI to estimated energy requirement (EER) – namely, the procedure proposed by Huang et al.( Reference Huang, Roberts and Howarth 25 ). Subjects were identified as acceptable reporters, under-reporters and over-reporters of EI according to whether the individual’s ratio was within, below or above the 95 % confidence limits for agreement between EI:BMR and the respective physical activity level (PAL) or of the expected EI:EER of 1·0. For the principles of the Goldberg cut-off, the PAL for sedentary lifestyle (i.e. 1·55)( Reference Black 24 ) was applied for all subjects, because of a lack of an objective measure of physical activity in the present study. BMR was estimated using Schofield sex- and age-specific equations based on body height and weight( Reference Schofield 26 ). The 95 % confidence limits for agreement (upper and lower cut-off values) between EI:BMR and the PAL were calculated, taking into account CV in intakes and other components of energy balance (i.e. the within-subject variation in EI: 23 %; the precision of the estimated BMR relative to the measured BMR: 8·5 %; and the between-subject variation in PAL: 15 %)( Reference Black 24 ). Consequently, under-reporters, acceptable reporters and over-reporters were defined as having EI:BMR<0·96, 0·96–2·49 and >2·49 for 2-d data and <0·87, 0·87–2·75 and >2·75 for 1-d data, respectively.
EER was calculated using sex- and age-specific equations for use in populations with a range of weight statuses, published from the US Dietary Reference Intakes, based on sex, age, body height and weight and physical activity( 27 ). Because of a lack of an objective measure of physical activity as mentioned above, we assumed ‘low active’ level of physical activity (i.e. PAL≥1·4 to <1·6)( 27 ) for all subjects during this calculation. The 95 % confidence limits of the expected EI:EER ratio of 0 on the natural log scale were calculated, taking into account CV in intakes and other components of energy balance (i.e. the within-subject variation in EI: 23 %; the error in the EER equations: 11 %; and the day-to-day variation in total energy expenditure: 8·2 %)( Reference Black 24 , Reference Huang, Roberts and Howarth 25 , 27 ). Consequently, under-reporters, acceptable reporters and over-reporters were defined as having EI:EER<0·65, 0·65–1·53 and >1·53 for 2-d data and <0·59, 0·59–1·71 and >1·71 for 1-d data, respectively.
Statistical analysis
Statistical analyses were performed using SAS statistical software (version 9.2, SAS Institute). All reported P values are two-tailed, and P<0·05 was considered to be statistically significant. All the analyses used the NHANES-provided sampling weights that were calculated to take into account unequal probabilities of selection, resulting from the sample design, non-response and planned over-sampling of selected sub-groups, so that the results are representative of the US community-dwelling population( Reference Johnson, Paulose-Ram and Ogden 20 , 28 ). For EI, BMR, EER, EI:BMR and EI:EER, sample-weighted means (with their se) were generated using the PROC SURVYMEANS procedure. Differences in these variables across categories of each of the characteristics were examined by Wald’s F test using the PROC SURVEYREG procedure. Proportions (with their se) of under-reporters, acceptable reporters and over-reporters of EI were calculated using the PROC SURVEYFREQ procedure. Differences in proportions of under-reporters, acceptable reporters and over-reporters across categories of each of the characteristics were examined by the χ 2 test using the PROC SURVEYFREQ procedure.
The risk of being classified as an under-reporter of EI, compared with being an acceptable reporter, or as an over-reporter, compared with being an acceptable reporter, was estimated using logistic regression. First, using the PROC SURVEYLOGISTIC procedure, crude OR and 95 % CI for the risk of being classified as an under-reporter or over-reporter were calculated for each category of factors, which are possibly associated with EI misreporting – namely, sex (reference: men), age group (reference: 20–39 years), race/ethnicity (reference: non-Hispanic white), years of education (reference: <12 years), family poverty income ratio (reference: <130 %), weight status (reference: normal), perceived weight status (reference: about the right weight), smoking status (reference: never), any recreational physical activity (reference: yes) and survey cycle (reference: 2003–2004). Multivariate-adjusted OR and 95 % CI were then calculated by entering all the variables simultaneously into the regression model in order to assess the independent associations.
These analyses were conducted separately for men and women. The results on the association between EI reporting and the variables examined were essentially the same in men and women, although the percentage of under-reporters was higher in women but that of over-reporters was higher in men, as shown below. The present report, therefore, presents the results for men and women combined.
Results
Among 19 396 subjects with 2-d dietary data, the sample-weighted mean EI:BMR was 1·28, whereas the corresponding value for EI:EER was 0·85 (Table 1). Men had a higher mean EI:BMR than women. Mean EI:BMR differed significantly among age groups, with the highest in the youngest group (20–39 years) and the lowest in the oldest group (≥60 years); among race/ethnicity groups, with the highest in non-Hispanic whites and Mexican Americans and the lowest in non-Hispanic blacks; among smoking status groups, with the highest in current smokers; and among survey cycles, with the highest in 2003–2004 and the lowest in 2007–2008. Years of education and family poverty income ratio were positively associated with EI:BMR. Mean EI:BMR in obese and overweight subjects was lower compared with normal-weight and underweight subjects. Mean EI:BMR similarly differed according to perceived weight status, with the highest in those who considered themselves underweight and the lowest in those who considered themselves overweight. Subjects with any recreational physical activity had a higher mean EI:BMR than those without any activity. Similar associations of these characteristics with EI:EER were also observed.
Table 1 Characteristics of the subjects: National Health and Nutrition Examination Survey (NHANES) 2003–2012 (n 19 396)Footnote * (Mean values with their standard errors)
EI, energy intake; EER, estimated energy requirement.
* All % and mean values are weighted to reflect the survey design characteristics. Analyses are based on subjects with complete data on two 24-h dietary recalls as well as complete information on the variables of interest.
† Based on average values of the two 24-h dietary recalls.
‡ Estimated using Schofield’s sex- and age-specific equations based on body height and weightReference Schofield (26) .
§ Calculated using sex- and age-specific equations for use in populations with a range of weight statuses published from the US Dietary Reference Intakes based on sex, age and body height and weight assuming ‘low active’ level of physical activity for all subjects (27) .
|| Based on Wald’s F test.
¶ Defined based on BMI (kg/m2) according to World Health Organization( 23 ) recommendations: <18·5 for underweight, ≥25 to <30 for normal, ≥25 to <30 for overweight and ≥30 for obese subjects.
The sample-weighted percentages of under-reporters, acceptable reporters and over-reporters of EI were 25·1, 73·5 and 1·4 %, respectively, based on EI:BMR, and 25·7, 71·8 and 2·5 %, respectively, based on EI:EER (Table 2). Using EI:BMR, the percentage of under-reporters was higher in women but that of over-reporters was higher in men. With regard to age, there were more under-reporters among the oldest group, whereas there were more over-reporters among the youngest group. For race/ethnicity, there were more under-reporters among non-Hispanic blacks. Years of education and family poverty income ratio were inversely associated with the percentages of both under-reporters and over-reporters. There were more under-reporters and fewer over-reporters among overweight and obese subjects. For perceived weight status, there were more under-reporters among those who considered themselves overweight and more over-reporters among those who considered themselves underweight. Current smokers had a higher percentage of over-reporters, whereas those with any recreational physical activity had a lower percentage of under-reporters. The proportion of under-reporters and over-reporters differed among survey cycles, with more under-reporters in 2007–2008 and more over-reporters in 2005–2006. The results were similar based on using EI:EER to estimate misreporters, except for no difference according to the survey cycle.
Table 2 Numbers and percentages of under-reporters, acceptable reporters and over-reporters of energy intake (EI): National Health and Nutrition Examination Survey (NHANES) 2003–2012 (n 19 396)Footnote * (Percentages with their standard errors)
EER, estimated energy requirement.
* All % values are weighted to reflect the survey design characteristics. Analyses are based on subjects with complete data on two 24-h dietary recalls as well as complete information on the variables of interest. Average EI values of the two 24-h dietary recalls were used.
† Under-reporters were defined as subjects with EI:BMR<0·96; acceptable reporters as subjects with EI:BMR 0·96–2·49; and over-reporters as subjects with EI:BMR>2·49. BMR was estimated using Schofield’s sex- and age-specific equations based on body height and weightReference Schofield (26) .
‡ Under-reporters were defined as subjects with EI:EER<0·65; acceptable reporters as subjects with EI:EER 0·65–1·53; and over-reporters as subjects with EI:EER>1·53. EER was calculated using sex- and age-specific equations for use in populations with a range of weight statuses published from the US Dietary Reference Intakes based on sex, age and body height and weight assuming ‘low active’ level of physical activity for all subjects (27) .
§ Based on χ 2 test.
|| Defined based on BMI (kg/m2) according to World Health Organization( 23 ) recommendations: <18·5 for underweight, ≥25 to <30 for normal, ≥25 to <30 for overweight and ≥30 for obese subjects.
Odds ratios and 95 % confidence intervals for the risk of being an under-reporter compared with an acceptable reporter are shown in Table 3. The results for the crude and multivariate-adjusted models were generally similar except for any recreational physical activity. In the multivariate analyses, based on EI:BMR and EI:EER, a higher risk of being an under-reporter was associated with the female sex, age ≥60 years (EI:BMR only) (compared with age 20–39 years), non-Hispanic blacks (compared with non-Hispanic white), overweight and obesity (compared with normal weight), perceived overweight (EI:EER only) (compared with about the right weight) and survey cycle 2007–2008 (compared with 2003–2004). A lower risk of being an under-reporter was associated with higher years of education (compared with the lowest), higher family poverty income ratio (compared with the lowest), Mexican Americans and former smoking (EI:EER only) (compared with never smoking).
Table 3 Risk of being an under-reporter of energy intake (EI) compared with being an acceptable reporter of EI: National Health and Nutrition Examination Survey (NHANES) 2003–2012Footnote * (Odds ratios and 95 % confidence intervals)
EER, estimated energy requirement.
* Analyses are based on subjects with complete data on two 24-h dietary recalls as well as complete information on the variables of interest. Average EI values of the two 24-h dietary recalls were used.
† Under-reporters were defined as subjects with EI:BMR<0·96; acceptable reporters as subjects with EI:BMR 0·96–2·49. Over-reporters (subjects with EI:BMR>2·49; n 273) were excluded from the analysis. BMR was estimated using Schofield’s sex- and age-specific equations based on body height and weightReference Schofield (26) .
‡ Under-reporters were defined as subjects with EI:EER<0·65; acceptable reporters as subjects with EI:EER 0·65–1·53. Over-reporters (subjects with EI:EER>1·53; n 478) were excluded from the analysis. EER was calculated using sex- and age-specific equations for use in populations with a range of weight statuses published from the US Dietary Reference Intakes based on sex, age and body height and weight assuming ‘low active’ level of physical activity for all subjects (27) .
§ Each of the variables listed was entered into the model separately.
|| All the variables listed were entered into the model simultaneously.
¶ Defined based on BMI (kg/m2) according to World Health Organization( 23 ) recommendations: <18·5 for underweight, ≥25 to <30 for normal, ≥25 to <30 for overweight and ≥30 for obese subjects.
Table 4 lists the OR and 95 % CI for the risk of being an over-reporter compared with an acceptable reporter. The results for the crude and multivariate-adjusted models were again generally similar except for years of education. In the multivariate analyses, a lower risk of being an over-reporter was associated with the female sex (EI:EER only), age ≥60 years, higher family poverty income ratio, overweight and obese, perceived overweight and survey cycle 2011–2012 (EI:BMR only). A higher risk of being an over-reporter was associated with underweight (EI:EER only) and current smoking.
Table 4 Risk of being an over-reporter of energy intake (EI) compared with being an acceptable reporter of EI: National Health and Nutrition Examination Survey (NHANES) 2003–2012Footnote * (Odds ratios and 95 % confidence intervals)
EER, estimated energy requirement.
* Analyses are based on subjects with complete data on two 24-h dietary recalls as well as complete information on the variables of interest. Average EI values of the two 24-h dietary recalls were used.
† Over-reporters were defined as subjects with EI:BMR>2·49; acceptable reporters as subjects with EI:BMR 0·96–2·49. Under-reporters (subjects with EI:BMR<0·96; n 5633) were excluded from the analysis. BMR was estimated using Schofield’s sex- and age-specific equations based on body height and weightReference Schofield 26 .
‡ Over-reporters were defined as subjects with EI:EER>1·53; acceptable reporters as subjects with EI:EER 0·65–1·53. Under-reporters (subjects with EI:EER<0·65; n 5560) were excluded from the analysis. EER was calculated using sex- and age-specific equations for use in populations with a range of weight statuses published from the US Dietary Reference Intakes based on sex, age and body height and weight assuming ‘low active’ level of physical activity for all subjects 27 .
§ Each of the variables listed was entered into the model separately.
|| All the variables listed were entered into the model simultaneously.
¶ Defined based on BMI (kg/m2) according to World Health Organization( 23 ) recommendations: <18·5 for underweight, ≥25 to <30 for normal, ≥25 to <30 for overweight and ≥30 for obese subjects.
We repeated all the analyses using 28 794 subjects with the first dietary recall data. The sample-weighted mean EI:BMR was 1·31, whereas the corresponding value for EI:EER was 0·87 (online Supplementary Table S1). The sample-weighted percentages of under-reporters, acceptable reporters and over-reporters of EI were 20·5, 77·5 and 2·0 %, respectively, based on EI:BMR, and 21·0, 76·4 and 2·6 %, respectively, based on EI:EER (online Supplementary Table S2). Factors significantly associated with the risk of being an under-reporter or being an over-reporter compared with being an acceptable reporter were generally similar (online Supplementary Tables S3 and S4, respectively), except for no association of survey year with both under-reporting and over-reporting and an inverse association between years of education and over-reporting.
Discussion
Using two 24-h dietary recalls from NHANES 2003–2012, we found that misreporting, particularly under-reporting, of EI remains prevalent and differential in US adults aged ≥20 years. Percentages of under-reporters and over-reporters of EI were 25·1 and 1·4 %, respectively, based on EI:BMR, and 25·7 and 2·5 %, respectively, based on EI:EER. A higher risk of being an under-reporter of EI compared with being an acceptable reporter was associated with female sex, older age, non-Hispanic blacks (compared with non-Hispanic whites), lower education, lower family poverty income ratio and overweight and obesity. A higher risk of being an over-reporter compared with being an acceptable reporter was associated with male sex, younger age, lower family poverty income ratio, current smoking (compared with never smoking) and underweight. Similar findings were observed when analysing based on the first 24-h dietary recall only (NHANES 1999–2012). To our knowledge, this is the first study to examine the prevalence and characteristics of misreporting of EI in a representative sample of US adults from the continuous NHANES.
Only a few recent national studies have examined misreporting of EI among adults. Among 1487 adults in Britain, EI assessed by a 7-d weighed dietary record was evaluated according to EI:EER( Reference Murakami, McCaffrey and Livingstone 9 ). The prevalence of under-reporters and over-reporters was 63 and 0·4 %, respectively, for men, and 55 and 0 %, respectively, for women. A French study evaluated EI assessed by a 7-d diet record among 1567 adults based on the Goldberg principles( Reference Vanrullen, Volatier and Bertaut 10 ). The prevalence of under-reporters was 24 % in men and 21 % in women (over-reporters not defined). EI estimated by a 24-h dietary recall was similarly evaluated in 3919 adults in New Zealand, and the prevalence of under-reporters was 21 % for men and 25 % for women (over-reporters not defined)( Reference Gemming, Jiang and Swinburn 11 ). Similar prevalence of under-reporting of EI (obtained from a 24-h dietary recall) was also observed in Korean adults: 14 % for men and 23 % for women (over-reporters not defined)( Reference Kye, Kwon and Lee 12 ). A study in Ireland investigated EI estimated by a FFQ using the Goldberg principles (n 7521), and the prevalence of under-reporting and over-reporting was 33 and 12 %, respectively( Reference Lutomski, van den Broeck and Harrington 7 ). Similar findings have been observed in a study among Norwegians where EI was assessed by a FFQ; prevalence of under-reporting was 20 % for men and 25 % for women, with prevalence of over-reporting being 7 % for men and 5 % for women( Reference Johansson, Solvoll and Bjorneboe 8 ). In the previous NHANES III (1988–1991), 18 % of the men and 28 % of the women were classified as under-reporters( Reference Briefel, Sempos and McDowell 13 ). In the present analysis, the prevalence of under-reporting both using two 24-h recall data (25·1 % based on EI:BMR and 25·7 % based on EI:EER in NHANES 2003–2012) and using one 24-h recall data (20·5 % based on EI:BMR and 21·0 % based on EI:EER in NHANES 1999–2012) was relatively similar to those observed in other countries. Although it is difficult to determine whether the difference in the prevalence among countries reflects the true difference in the accuracy of reporting or is merely due to differences in the criteria used to identify misreporters, dietary assessment instruments, food composition databases and population characteristics, these national studies clearly show that misreporting of EI remains a serious problem in dietary surveys among adults.
In this study, overweight and obese subjects were more likely to under-report EI, which has been consistently observed in many studies( Reference Livingstone and Black 1 – Reference Archer, Hand and Blair 14 ). The association between weight status and EI under-reporting should be carefully considered in any relevant analysis based on continuous NHANES, given that there has been an increase in the prevalence of obesity and extreme obesity (BMI≥40 kg/m2) since previous NHANES III( Reference Fryar, Carroll and Ogden 29 ). In addition, female sex and older age were associated with under-reporting of EI, although the associations of sex and age with under-reporting are not consistent in the literature( Reference Livingstone and Black 1 , Reference Freedman, Commins and Moler 6 , Reference Lutomski, van den Broeck and Harrington 7 , Reference Vanrullen, Volatier and Bertaut 10 ). For other correlates of misreporting, research is limited or the results are generally inconsistent( Reference Livingstone and Black 1 ). For race/ethnicity, we found that a higher risk of under-reporting was associated with non-Hispanic blacks (compared with non-Hispanic whites), which has also been observed among US adults from previous NHANES 1988–1991( Reference Briefel, Sempos and McDowell 13 ). Lower education and lower family poverty income ratio were associated with a higher risk of under-reporting. Both low( Reference Mattisson, Wirfalt and Aronsson 3 , Reference Vanrullen, Volatier and Bertaut 10 , Reference Briefel, Sempos and McDowell 13 ) and high( Reference Tooze, Subar and Thompson 2 , Reference Freedman, Commins and Moler 6 , Reference Lutomski, van den Broeck and Harrington 7 ) socio-economic statuses have been shown to be associated with under-reporting. Characteristics associated with over-reporting of EI are less understood. We found that over-reporting was associated with male sex, younger age, lower family poverty income ratio, current smoking and underweight. In an analysis of Irish adults, younger age, lower social class and underweight were associated with a higher risk of over-reporting( Reference Lutomski, van den Broeck and Harrington 7 ). Underweight has also been associated with over-reporting in other studies( Reference Mattisson, Wirfalt and Aronsson 3 , Reference Johansson, Solvoll and Bjorneboe 8 ). Although these variables may not always be associated with EI misreporting, and the association should be dependent on the population characteristics, dietary assessment methods and the procedure for identifying misreporters, accumulating literature clearly indicates that misreporting occurs non-randomly in adult populations. Specific to NHANES, we found that survey cycle was associated with both under-reporting and over-reporting of EI at least in some analyses based on two 24-h dietary recalls, which has also been indicated in a previous univariate analysis( Reference Archer, Hand and Blair 14 ). This differential reporting may severely distort the validity of trend analyses using dietary intake data. Thus, previously reported trend analyses should be cautiously interpreted in this regard, and future analyses should properly take into account misreporting of EI. Nonetheless, it should also be pointed out that survey cycle was not associated with either under-reporting or over-reporting when only the first 24-h dietary recall was analysed.
Several limitations of the present study are acknowledged. At present, the only way to obtain unbiased information on energy requirements in free-living settings is to use doubly labelled water as a biomarker( Reference Livingstone and Black 1 ). This technique is expensive and impractical for application to large-scale epidemiological studies, and thus alternative procedures are used( Reference Mattisson, Wirfalt and Aronsson 3 , Reference Rosell, Hellenius and De Faire 5 , Reference Murakami, McCaffrey and Livingstone 9 , Reference Huang, Roberts and Howarth 25 ). In the present study, EER was calculated with the use of equations from the US Dietary Reference Intakes, which have been developed based on a large number of measurements of total energy expenditure by the doubly labelled water method and are highly accurate (R 2 0·82 for men and 0·79 for women)( 27 ). In the absence of actual, measured total energy expenditure, these equations should serve as the best proxy. Owing to constraints within the data set, we did not have a validated and individualised measure of physical activity. Instead, we assumed ‘low active’ level of physical activity for all subjects during the calculation of EER (as well as using the PAL for sedentary lifestyle for all subjects when using the Goldberg principles). This seems adequate for most US adults, based on the accelerometer-based data in NHANES 2003–2006( Reference Troiano, Berrigan and Dodd 30 , Reference Luke, Dugas and Durazo-Arvizu 31 ). Nevertheless, in some very active individuals, EER would be underestimated, having the effect of overestimating EI:EER, thus tending to retain those individuals as acceptable reporters or over-reporters. Further, we do not know the sensitivity and specificity of the procedures for identifying under-reporters and over-reporters of EI used; in addition, there is currently not enough information on relative merits of the different methods (i.e. EI:BMR and EI:EER) for detecting misreporters. Thus, we are unable to determine whether the associations found between misreporting of EI and several characteristics are true, or were artifacts caused by the procedure used to identify misreporters, as well as errors associated with food composition databases used. Finally, the cross-sectional nature of the study does not permit the assessment of causality, owing to the uncertain temporality of the association.
In conclusion, in this comprehensive analysis based on data from NHANES 2003–2012, we found that misreporting of EI assessed by two 24-h dietary recalls was too prevalent to ignore in US adults aged ≥20 years: 26·5 % based on EI:BMR and 28·2 % based on EI:EER. Unfortunately, such EI misreporting was differential among populations. Under-reporting was associated with female sex, older age, non-Hispanic blacks (compared with non-Hispanic whites), lower education, lower family poverty income ratio and overweight and obesity, whereas over-reporting was associated with male sex, younger age, lower family poverty income ratio, current smoking and underweight. The results were similar when only the first 24-h dietary recall was assessed based on data from NHANES 1999–2012. Thus, it is essential to consider this differential misreporting of EI when using dietary data from NHANES.
Acknowledgements
This work was supported in part by the Grants-in-Aid for Young Scientists (B) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (K. M., grant number 15K16213). The Ministry of Education, Culture, Sports, Science and Technology of Japan had no role in the design, analysis or writing of this article.
K. M. contributed to the concept and design of the study, statistical analysis, data interpretation and manuscript writing. M. B. E. L. critically reviewed the manuscript. All the authors read and approved the final version of the manuscript.
There are no conflicts of interest.
Supplementary Material
For supplementary materials referred to in this article, please visit http://dx.doi.org/10.1017/S0007114515002706
