Hostname: page-component-7c8c6479df-ws8qp Total loading time: 0 Render date: 2024-03-18T21:57:57.074Z Has data issue: false hasContentIssue false

Dietary assessment methods for micronutrient intake in infants, children and adolescents: a systematic review

Published online by Cambridge University Press:  26 January 2010

Adriana Ortiz-Andrellucchi
Affiliation:
Nutrition Research Group, Department of Clinical Sciences, Centre for Health Sciences, University of Las Palmas de Gran Canaria, PO Box 550, 35080Las Palmas de Gran Canaria, Spain
Patricia Henríquez-Sánchez
Affiliation:
Nutrition Research Group, Department of Clinical Sciences, Centre for Health Sciences, University of Las Palmas de Gran Canaria, PO Box 550, 35080Las Palmas de Gran Canaria, Spain
Almudena Sánchez-Villegas
Affiliation:
Nutrition Research Group, Department of Clinical Sciences, Centre for Health Sciences, University of Las Palmas de Gran Canaria, PO Box 550, 35080Las Palmas de Gran Canaria, Spain
Luis Peña-Quintana
Affiliation:
Nutrition Research Group, Department of Clinical Sciences, Centre for Health Sciences, University of Las Palmas de Gran Canaria, PO Box 550, 35080Las Palmas de Gran Canaria, Spain
Michelle Mendez
Affiliation:
Center for Research in Environmental Epidemiology, Municipal Institute of Medical Research. Biomedical Research Park, Barcelona, Spain
Lluís Serra-Majem*
Affiliation:
Nutrition Research Group, Department of Clinical Sciences, Centre for Health Sciences, University of Las Palmas de Gran Canaria, PO Box 550, 35080Las Palmas de Gran Canaria, Spain
*
*Corresponding author: Lluis Serra-Majem, fax +34 928 453475, email lserra@dcc.ulpgc.es
Rights & Permissions [Opens in a new window]

Abstract

A systematic literature search identified studies validating the methodology used for measuring the usual dietary intake in infants, children and adolescents. The quality of each validation study selected was assessed using a European micronutrient Recommendations Aligned-developed scoring system. The validation studies were categorised according to whether the study used a reference method that reflected short-term intake ( < 7 d), long-term intake ( ≥ 7 d) or used biomarkers. A correlation coefficient for each nutrient was calculated from the mean of the correlation coefficients from each study weighted by the quality of the study. Thirty-two articles were included in the present review: validation studies from infants (1–23 months); child preschool (2–5 years); children (6–12 years); adolescents (13–18 years). Validation of FFQ studies in infants and preschool children using a reference method that reflected short-term intake showed good correlations for niacin, thiamin, vitamins B6, D, C, E, riboflavin, Ca, K, Mg, Fe and Zn (with correlations ranging from 0·55 for vitamin E to 0·69 for niacin).Regarding the reference method reflecting short-term intake in children and adolescents, good correlations were seen only for vitamin C (r 0·61) and Ca (r 0·51). Using serum levels of micronutrient demonstrated that the 3 d weighed dietary records was superior to the FFQ as a tool to validate micronutrient intakes. Including supplement users generally improved the correlations between micronutrient intakes estimated by any of the dietary intake methods and respective biochemical indices.

Type
Full Papers
Copyright
Copyright © The Authors 2010

Growth in children from birth through adolescence is an extremely complex process. It is influenced not only by the genetic make-up of the individual but also by environmental factors, medical illnesses and nutritional status(Reference Mascarenhas, Zemel and Stallings1). Dietary assessments among infants and preschool children are complicated by the facts that dietary habits change rapidly in infancy, parents may share the responsibility for the child with other adults, e.g. in day-care, and finally not all food served to the infants are consumed, resulting in their disposal(Reference Andersen, Lande and Arsky2).

In the school age years, children experience enormous cognitive, emotional and social growth and development. Children transition from consuming most food intake under adult control and supervision to taking increasing responsibility for their food choices. The cognitive abilities required to self-report food intake include an adequately developed concept of time, a good memory and attention span and knowledge of the names of food(Reference Baranowski and Domel3, Reference Livingstone and Robson4). The need for adult assistance in dietary reporting is also driven by the limited scope of the child's experience and knowledge of food preparation(Reference Sobo and Rock5).

During adolescence, children undergo profound biological, emotional, social and cognitive changes to reach adult maturity. Adolescents' need for energy and all nutrients significantly increases to support the rapid rate of growth and development. Moreover, although appetite and food intake tend to increase, psychosocial characteristics often lead to the development of high-risk nutritional behaviours such as excessive dieting, adoption of fad diets or excessive alcohol consumption. The high prevalence of overweight and obesity, eating disorders, adolescent pregnancy and the lack of consumption of five fruits and vegetables a day constitute some of the nutritional challenges facing adolescents(Reference Frank6).

Research conducted as part of the European Commission's European micronutrient Recommendations Aligned Network of Excellence(Reference Ashwell, Lambert and Alles7) has focused on extensive literature reviews addressing the validation of methods used to assess intake of micronutrients, n-3 fatty acids and of special population groups, including pregnant women, infants, children, adolescents and elderly people. In this review, studies validating dietary methods for assessing micronutrient intake in infants, children and adolescents are presented.

Material and methods

The research question applied to the systematic review was ‘which dietary methods are reliable for the assessment of micronutrient intake in infants, children and adolescents?’ The main stages of the review are illustrated in Fig. 1. The review included English, Spanish, French, Italian, Portuguese and German articles, without limits on time frame or country published before April 2008. Stage 1 of the review involved searching for publications using electronic databases (MEDLINE and EMBASE). The MeSH terms used in the general search were: nutritional assessment, diet, nutritional status, dietary intake, food intake, validity, validation study, reproducibility, replication study, correlation coefficient and correlation study in the title and abstract. As a second specific search, the following words were included: infants (1–23 months), preschool child (2–5 years), children (6–12 years), adolescents (13–18 years), ‘dietary assessment’, ‘dietary intake’, ‘nutrition assessment’, ‘diet quality’, reliability, reproducibility, validit* and correlate* as free text in the title and abstract. Additional publications were identified from references published in the original papers. At stage 2 of the review, the titles and abstracts were analysed by two independent reviewers and the exclusion criteria were applied (Table 1). At stage 3, studies that fulfilled the inclusion criteria were analysed for relevance to the research question.

Fig. 1 Main stages of the systematic review process.

Table 1 Inclusion and exclusion criteria

The selected studies were then classified into three different types according to the reference method applied in the validation studies: (1) reference method assessing intake of < 7 d (including 24 h dietary recall (24HR), estimated dietary records (EDR) and weighed dietary records (WDR)), classified as reflecting short-term intake; (2) reference method assessing intake of ≥ 7 d, reflecting more long-term intake; (3) reference method that employed the use of a biomarker (BM). Furthermore, the different studies included in this review were scored according to a quality score system developed by European micronutrient Recommendations Aligned. The studies were rated according to the sample size, the statistics used to validate the method, the procedure of data collection, the consideration or not of seasonality and the inclusion or not of vitamin supplement use. (For details see the article in this supplement ‘Evaluating the quality of dietary intake validation studies’). A total score was calculated according to the mean of the correlation coefficients weighted by the quality score of the validation study. It was considered a poor method for assessing specific nutrient intake when the correlation between methods was < 0·30. Methods whose correlations were between 0·30 and 0·50 were regarded as acceptable for assessing nutrient intake. Good methods were those whose correlations were between 0·51 and 0·70, and finally, when the correlation was >0·70 the method was considered very good.

Results

A total of thirty-two publications(Reference Andersen, Lande and Arsky2, Reference Marriott, Robinson and Poole8Reference Räsänen39) were selected for inclusion, with information on each validation study, ordered by publication year, summarised in Table 2. Fifteen of the publications showed results from European countries (Norway, Greece, Belgium, Italy, Denmark, United Kingdom and Finland), fifteen from American countries (United States of America, Brazil and Canada), one study from Australia and one study from New Zealand. The number of participants varied from 17 to 741 in the selected studies. In eight of the studies presented(Reference Moore, Braid and Falk11, Reference Magkos, Manios and Babaroutsi13Reference Bertoli, Petroni and Pagliato16, Reference Taylor and Goulding26, Reference Iannotti, Zuckerman and Blyer34, Reference Lytle, Nichaman and Obarzanek36), only one type of micronutrient was analysed, while in the rest of the publications included in this review, correlations for a wide variety of micronutrients were observed, and a total of twenty micronutrients were analysed. Tables 3 and 4 show information on the correlation between methods and other statistics in the validation studies in infants, children and adolescents for twelve vitamins and eight minerals, respectively.

Table 2 Characteristics of included studies

24 HR, 24 h diet recalls; EPIC, european prospective investigation of cancer; WDR, weighed dietary record; EDR, estimated dietary record.

Table 3 Validation studies in infants, children and adolescents: vitamins

WDR, weighed dietary record; CC, correlation coefficient; 24 HR, 24 h recall; RE, retinol equivalent; DH, diet history; BM, biomarker; EDR, estimated dietary record; YAQ, Youth/Adolescent Questionnaire; FC, food checklist; SW, semi-weighed method; SAW, Self-administered workbook; PFD, pre-coded food diary; HH measures, household measures.

Mean values were significantly different: *P < 0·05; **P < 0·01; *** P < 0·001.

Excluding vitamin supplementation.

Table 4 Validation studies in infants, children and adolescents: minerals

WDR, weighed dietary record; CC, correlation coefficient; 24 HR, 24 h recall; EDR, estimated dietary record; YAQ, Youth/Adolescent Questionnaire; BM, biomarker; HH measures, household measures; FC, food checklist; SW, semi-weighed method; DH, diet history; PFD, pre-coded food diary; IIC, intra-class correlation coefficients.

Mean values were significantly different: *P < 0·05; **P < 0·01; *** P < 0·001.

Infants

This group included infants aged 1–23 months. Of the thirty-two articles included in the present review, seven showed data on the validation of methods used to assess micronutrient intake in infants(Reference Andersen, Lande and Arsky2, Reference Marriott, Robinson and Poole8, Reference Marriott, Inskip and Borland9, Reference Williams and Innis17, Reference Marshall, Eichenberger Gilmore and Broffitt21, Reference Parrish, Marshall and Krebs22, Reference Blum, Wei and Rockett24). Evaluating the quality of these validation studies resulted in quality scores ranging from 2·5 to 5. All the studies evaluated micronutrient intake in infants using a FFQ, and only one article applied four 24HR as an additional dietary assessment method(Reference Parrish, Marshall and Krebs22). Different FFQ were validated for which wide variations in the number of food items were observed (7–191 items). In addition, five studies were classified into group 1 with a reference method that reflected short-term intake, in which one applied 24 h recalls(Reference Blum, Wei and Rockett24), three used WDR(Reference Marriott, Robinson and Poole8, Reference Marriott, Inskip and Borland9, Reference Marshall, Eichenberger Gilmore and Broffitt21) and one applied EDR(Reference Williams and Innis17). Likewise, another study was classified into group 2, where the reference method reflected long-term intake; this study used WDR as the reference method(Reference Andersen, Lande and Arsky2). Finally, two studies utilised BM as the reference method(Reference Williams and Innis17, Reference Parrish, Marshall and Krebs22), of which one(Reference Williams and Innis17) presented validation of more than one instrument. The number of repeated 24 h recalls ranged from 3 to 4 d of administration. Dietary records varying in the number of recording days (from 3 to 7 d) were used as the reference method in a total of five studies.

Comparison of different dietary assessment methods in infants by vitamins and minerals is presented in Fig. 2. This figure shows that WDR used as the reference method for evaluating FFQ presented better correlations for several micronutrients than the other methods in this population group. However, we must emphasise that this result is probably due to the fact that 57 % of the studies analysed in this group used WDR as the reference method.

Fig. 2 Comparison of different dietary assessment methods in infants (1–23 months) and preschool children (2–5 years) by vitamins and minerals (mean of quality weighted correlation coefficients) 24 HR, 24 h recall; EDR, estimated dietary record; WDR, weighed dietary record; BM, biomarker; DH, diet history. (a) Infants (three or more studies: vitamin D, four studies FFQ v. WDR; vitamin E, three studies FFQ v. WDR; vitamin C, three studies FFQ v. WDR; thiamin, three studies FFQ v. WDR; riboflavin, four studies FFQ v. WDR; calcium, four studies FFQ v. WDR; iron, four studies FFQ v. WDR). , FFQ v. 24 HR (one study(Reference Blum, Wei and Rockett24)); , FFQ v. WDR (four studies(Reference Andersen, Lande and Arsky2, Reference Marriott, Robinson and Poole8, Reference Marriott, Inskip and Borland9, Reference Marshall, Eichenberger Gilmore and Broffitt21)); , FFQ v. EDR (one study(Reference Williams and Innis17)); , FFQ v. BM (one study(Reference Parrish, Marshall and Krebs22)). (b) Preschool children (three or more studies: not for any micronutrient). , FFQ v. 24 HR (three studies(Reference Blum, Wei and Rockett24, Reference Iannotti, Zuckerman and Blyer34, Reference Stein, Shea and Basch37)); , FFQ v. WDR (two studies(Reference Andersen, Lande and Trygg18, Reference Andersen, Lande and Trygg19, Reference Marshall, Eichenberger Gilmore and Broffitt21)); , FFQ v. BM (one study(Reference Parrish, Marshall and Krebs22)); , 24 HR v. DH (one study(Reference Räsänen39)).

Table 5 presents the classification of the dietary methods utilised for studies in infants according to the mean of the correlation coefficients for each micronutrient weighted by the quality of different validation studies included in this review. Methods analysed met the criteria of having at least three studies, thus providing sufficient data to conduct quantitative estimates for each micronutrient(Reference Treadwell, Tregear and Reston40). Vitamin D and vitamin E intake analysed using FFQ v. WDR showed acceptable correlations. Comparing these methods, we observed that vitamin C, thiamin, riboflavin, Ca and Fe presented a good correlation. Additionally, when FFQ were validated considering WDR and EDR as the reference methods, only Ca's correlation increased slightly. The correlation for Fe was not modified, and for the rest of the micronutrients, there was insufficient data to conduct an analysis. (Table 5)

Table 5 Classification of the dietary assessment methods for infants, children and adolescents according to the weighted mean of the correlations of each micronutrient (including three or more studies)

WDR, weighed dietary record; EDR, estimated dietary record; 24 HR, 24 h recall; YAQ, Youth/Adolescent Questionnaire.

* Correlation V: very good (>0·7); G: good (0·51–0·70); A: acceptable (0·30–0·50); P: poor ( < 0·30)

Preschool children

For this review, the preschool children group included children aged 2–5 years. Of the thirty-two articles included in the present review, ten showed data on the validation of methods used to assess micronutrient intake in preschool children(Reference Holmes, Dick and Nelson10, Reference Huybrechts, De Bacquer and Matthys15, Reference Andersen, Lande and Trygg18, Reference Marshall, Eichenberger Gilmore and Broffitt21, Reference Parrish, Marshall and Krebs22, Reference Blum, Wei and Rockett24, Reference Taylor and Goulding26, Reference Iannotti, Zuckerman and Blyer34, Reference Stein, Shea and Basch37, Reference Räsänen39). Eight different FFQ had been validated(Reference Huybrechts, De Bacquer and Matthys15, Reference Andersen, Lande and Trygg18, Reference Marshall, Eichenberger Gilmore and Broffitt21, Reference Parrish, Marshall and Krebs22, Reference Blum, Wei and Rockett24, Reference Taylor and Goulding26, Reference Iannotti, Zuckerman and Blyer34, Reference Stein, Shea and Basch37), and a 24 h recall had been validated in two studies(Reference Holmes, Dick and Nelson10, Reference Räsänen39). Some articles presented validation of more than one instrument, of which one study also validated 24 h recalls(Reference Parrish, Marshall and Krebs22), and another study validated a food checklist and a semi-weighed method(Reference Holmes, Dick and Nelson10). After evaluating the quality of these validation studies, the quality scores obtained ranged from 2·5 to 5. Different FFQ were validated for which wide variations in the number of food items were observed (7–125 items). Eight studies were classified into group 1 with a reference method that reflected short-term intake, in which two used WDR(Reference Holmes, Dick and Nelson10, Reference Marshall, Eichenberger Gilmore and Broffitt21), another two applied EDR(Reference Huybrechts, De Bacquer and Matthys15, Reference Taylor and Goulding26), three used 24 h recalls(Reference Blum, Wei and Rockett24, Reference Iannotti, Zuckerman and Blyer34, Reference Jenner, Neylon and Croft38) and one applied a DH(Reference Räsänen39). Likewise, one study was classified into group 2, where the reference method reflected long-term intake, in which WDR were applied(Reference Andersen, Lande and Trygg18). Finally, one study utilised BM as the reference method, which presented validation of more than one instrument(Reference Parrish, Marshall and Krebs22). The number of repeated 24 h recalls ranged from 3 to 4 d of administration. Dietary records varying in the number of recording days (from 3 to 7 d) were used as the reference method in a total of four studies.

Comparison of different dietary assessment methods for vitamin and mineral intake in preschool children is presented in Fig. 2. This figure shows that there were not enough studies to conduct a comparison per micronutrient, as the minimum criterion of three studies per intake assessment method was not met. Moreover, when FFQ were validated applying WDR and EDR as the reference methods, only the correlation for Ca presented acceptable values (r 0·50; Table 5 and Fig. 3 ).

Fig. 3 Comparison of different dietary assessment methods in children (6–12 years) and adolescents (13–10 years) by vitamins and minerals (mean of quality weighted correlation coefficients) 24 HR, 24 h recall; EDR, estimated dietary record; WDR, weighed dietary record; DH, dietary history; YAQ, Youth/Adolescent Questionnaire. (a) Children (three or more studies: calcium, five studies FFQ v. 24 HR). , FFQ v. EDR (three studies(Reference Taylor and Goulding26, Reference Arnold, Rohan and Howe31, Reference Lytle, Nichaman and Obarzanek36)); , FFQ v. WDR (two studies(Reference Bertoli, Petroni and Pagliato16, Reference Lietz, Barton and Longbottom23)); , FFQ v.BM (one study(Reference Byers, Trieber and Gunter35)); , 24 HR v. DH (one study(Reference Räsänen39)); , YAQ v. 24 HR (one study(Reference Rockett, Breitenbach and Frazier29)); , FFQ v. 24 HR (five studies(Reference Moore, Braid and Falk11, Reference Magkos, Manios and Babaroutsi13, Reference Bertoli, Petroni and Pagliato16, Reference Field, Peterson and Gortmaker25, Reference Jenner, Neylon and Croft38)). (b) Adolescents (three or more studies: calcium: three studies FFQ v. 24 HR; three studies FFQ v. WDR). , FFQ v. 24 HR (three studies(Reference Moore, Braid and Falk11, Reference Harnack, Lytle and Story14, Reference Slater, Philippi and Fisberg20)); , FFQ v. WDR (three studies(Reference Bertoli, Petroni and Pagliato16, Reference Mølgaard, Sandström and Michaelsen27, Reference Andersen, Nes and Lillegaard33)); , 24 HR v. DH (one study(Reference Räsänen39)); , YAQ v. 24 HR (one study(29)).

Fig. 4 only shows FFQ validation studies that assessed micronutrient intake in infants and preschool children, using a short-term(Reference Marriott, Robinson and Poole8, Reference Marriott, Inskip and Borland9, Reference Huybrechts, De Bacquer and Matthys15, Reference Williams and Innis17, Reference Marshall, Eichenberger Gilmore and Broffitt21, Reference Blum, Wei and Rockett24, Reference Taylor and Goulding26, Reference Iannotti, Zuckerman and Blyer34, Reference Stein, Shea and Basch37) or a long-term(Reference Andersen, Lande and Arsky2, Reference Andersen, Lande and Trygg18) dietary assessment instrument or BM as a reference method(Reference Williams and Innis17, Reference Parrish, Marshall and Krebs22). In regards to the reference method that reflected short-term intake, good correlations were observed for niacin, thiamin, vitamins B6, D, C, E, riboflavin, Ca, K, Mg, Fe, and Zn. However, when the reference method used reflected long-term intake, good correlations were observed only for riboflavin and Fe. Additionally, when BM were used as the reference method, a good correlation was observed only for vitamin C. None of the micronutrients analysed showed correlations higher than 0·7 using a short-term or a long-term dietary assessment instrument or BM as the reference method. However, results presented in FFQ validation studies using short-term or long-term dietary instruments or BM as the reference methods based on correlations from only one or two studies should be viewed with caution (Fig. 4). To conduct micronutrient comparisons, there should be at least three or more studies to ensure the robustness of the results obtained.

Fig. 4 Validation of FFQ studies that assess micronutrient intake in infants (1–23 months) and preschool children (2–5 years) using as the reference method: short-term or long-term dietary instruments or biomarkers. Correlations: poor ( < 0·30), acceptable (0·30–0·50), good (0·51–0·70) and very good (>0·70). Three or more studies: sodium, vitamins B12, E, C, D, B6, zinc, iron, magnesium, potassium, calcium, riboflavin, thiamin, niacin. ■, Short-term intake ( < 7 d); , long-term intake ( ≥ 7 d); ▧, biomarkers.

Children

This group included children aged 6–12 years. Of the thirty-two articles included in the present review, seventeen showed data on validation of methods used to assess micronutrient intake in children(Reference Holmes, Dick and Nelson10Reference Harnack, Lytle and Story14, Reference Bertoli, Petroni and Pagliato16, Reference Lietz, Barton and Longbottom23, Reference Field, Peterson and Gortmaker25, Reference Taylor and Goulding26, Reference Rockett, Breitenbach and Frazier29Reference Bellù, Ortisi and Riva32, Reference Byers, Trieber and Gunter35, Reference Lytle, Nichaman and Obarzanek36, Reference Jenner, Neylon and Croft38, Reference Räsänen39). After evaluating the quality of these studies, the quality scores obtained ranged from 2·5 to 5. Eleven different FFQ(Reference Moore, Braid and Falk11, Reference Magkos, Manios and Babaroutsi13, Reference Harnack, Lytle and Story14, Reference Bertoli, Petroni and Pagliato16, Reference Lietz, Barton and Longbottom23, Reference Field, Peterson and Gortmaker25, Reference Taylor and Goulding26, Reference Bellù, Riva and Ortisi30Reference Bellù, Ortisi and Riva32, Reference Byers, Trieber and Gunter35, Reference Jenner, Neylon and Croft38) had been validated, and 24 h recalls had been validated in three studies(Reference Holmes, Dick and Nelson10, Reference Lytle, Nichaman and Obarzanek36, Reference Räsänen39). Different FFQ were validated for which wide variations in the number of food items were observed (10–175 items). Of these, it is worth pointing out the Youth/Adolescent Questionnaire (YAQ)(Reference Rockett, Breitenbach and Frazier29) that is a self-administered FFQ specifically designed for children aged 9–18 years. Another validated dietary method analysed included 4 d pre-coded food diaries(Reference Lillegaard, Løken and Andersen12). The pre-coded food diary is scanable and is developed to simplify the work of the respondents as well as of the researcher. Some articles presented validation of more than one instrument, of which one study also validated a food checklist and a semi-weighed method(Reference Holmes, Dick and Nelson10), and another study correlated data from WDR(Reference Lietz, Barton and Longbottom23). Eleven studies were classified into group 1 with a reference method that reflected short-term intake, in which six studies used 24HR(Reference Moore, Braid and Falk11, Reference Magkos, Manios and Babaroutsi13, Reference Harnack, Lytle and Story14, Reference Field, Peterson and Gortmaker25, Reference Rockett, Breitenbach and Frazier29, Reference Bellù, Riva and Ortisi30), two applied WDR(Reference Holmes, Dick and Nelson10, Reference Lillegaard, Løken and Andersen12), only one study used EDR(Reference Taylor and Goulding26), one applied a DH(Reference Räsänen39) and another study observed intakes(Reference Lytle, Nichaman and Obarzanek36). Likewise, five other studies were classified into group 3, where the reference method reflected long-term intake, two studies used WDR(Reference Bertoli, Petroni and Pagliato16, Reference Lietz, Barton and Longbottom23), two applied EDR(Reference Arnold, Rohan and Howe31, Reference Bellù, Ortisi and Riva32) and only one study utilised 24 h recalls(Reference Jenner, Neylon and Croft38) as the reference method. Finally, one study was validated using BM(Reference Byers, Trieber and Gunter35). Some articles applied more than one instrument as the reference methods, also employing the use of BM(Reference Lietz, Barton and Longbottom23). The number of repeated 24 h recalls ranged from 1 to 14 administration days. Dietary records varying in the number of recording days (from 1 to 14 d) were used as the reference method in a total of seven studies.

Fig. 3 presents the comparison of different dietary assessment methods in children by vitamins and minerals. In this figure, when the 24HR was used as the reference method, it seemed to obtain better correlations for several micronutrients. However, only five studies were included for measuring Ca intake, for which FFQ were validated against 24HR, presenting an acceptable correlation.

Additionally, when FFQ were validated utilising WDR and EDR as the reference methods, only the correlation for Ca presented an acceptable value (r 0·50). Similar correlations were observed when the validity of Ca intake methods using FFQ (FFQ and YAQ) was assessed by comparing them with Ca intake estimated by 24 HR (r 0·48 and r 0·49, respectively). The analysis for Ca was conducted as it met the minimum of having data from at least three studies. However, for the rest of the micronutrients, there were insufficient data to conduct a comparative analysis (Table 5).

Adolescents

For the present paper, the adolescent age group included children aged 13–18 years. Of the thirty-two articles included in the present review, ten showed data on validation of methods used to assess micronutrient intake in adolescents(Reference Holmes, Dick and Nelson10, Reference Moore, Braid and Falk11, Reference Harnack, Lytle and Story14, Reference Bertoli, Petroni and Pagliato16, Reference Slater, Philippi and Fisberg20, Reference Mølgaard, Sandström and Michaelsen27Reference Rockett, Breitenbach and Frazier29, Reference Andersen, Nes and Lillegaard33, Reference Räsänen39). Eight different FFQ(Reference Moore, Braid and Falk11, Reference Harnack, Lytle and Story14, Reference Bertoli, Petroni and Pagliato16, Reference Slater, Philippi and Fisberg20, Reference Mølgaard, Sandström and Michaelsen27Reference Rockett, Breitenbach and Frazier29, Reference Andersen, Nes and Lillegaard33, Reference Räsänen39) had been validated (of which one was the YAQ)(Reference Rockett, Breitenbach and Frazier29), and 24 h recalls had been validated in two studies(Reference Holmes, Dick and Nelson10, Reference Räsänen39). Some articles presented validation of more than one instrument, of which one study also validated a food checklist and a semi-weighed method(Reference Holmes, Dick and Nelson10). After evaluating the quality of these studies, the quality scores obtained ranged from 2·5 to 5. Seven studies were classified into group 1 with a reference method that reflected short-term intake, in which four studies used 24HR(Reference Moore, Braid and Falk11, Reference Harnack, Lytle and Story14, Reference Slater, Philippi and Fisberg20, Reference Rockett, Breitenbach and Frazier29), two studies applied WDR(Reference Holmes, Dick and Nelson10, Reference Mølgaard, Sandström and Michaelsen27) and another study administered a DH(Reference Räsänen39). In addition, two studies were classified into group 2 with a reference method that reflected long-term intake, in which WDR were used as the reference method(Reference Bertoli, Petroni and Pagliato16, Reference Andersen, Nes and Lillegaard33) and finally, one study used BM(Reference Green, Allen and O'Connor28). The number of repeated 24 h recalls ranged from 1 to 3. Different FFQ were validated for which wide variations in the number of food items were observed (10–190 items). Dietary records varying in the number of recording days (from 1 to 7 d) were used as the reference method in a total of four studies.

Comparison of different dietary assessment methods was limited in the adolescent group, due to the fact that there was an insufficient number of studies for each micronutrient that could provide conclusive results (Fig. 3). A good correlation (r 0·58) was observed, when FFQ were validated applying WDR as the reference method only for Ca. Additionally, when FFQ were validated considering 24HR as the reference method, the correlation for Ca presented an acceptable correlation (r 0·50). Similar correlations were observed when the validity of Ca intake using the FFQ (including the YAQ) was assessed by comparing them with Ca intake using the 24HR (r 0·52). There were not enough data to conduct an analysis for the rest of the micronutrients in this population group (Table 5).

Fig. 5 shows only FFQ validation studies that assessed micronutrient intake in children and adolescents, using a short-term(Reference Moore, Braid and Falk11, Reference Magkos, Manios and Babaroutsi13, Reference Harnack, Lytle and Story14, Reference Slater, Philippi and Fisberg20, Reference Field, Peterson and Gortmaker25Reference Mølgaard, Sandström and Michaelsen27, Reference Rockett, Breitenbach and Frazier29) or a long-term(Reference Bertoli, Petroni and Pagliato16, Reference Lietz, Barton and Longbottom23, Reference Arnold, Rohan and Howe31Reference Andersen, Nes and Lillegaard33, Reference Jenner, Neylon and Croft38) dietary assessment instrument or BM as the reference method(Reference Lietz, Barton and Longbottom23, Reference Green, Allen and O'Connor28, Reference Byers, Trieber and Gunter35). In regards to the reference method that reflected short-term intake, good correlations were observed for thiamin, vitamins B6, C, E, riboflavin, Ca, Mg, Cu and folate. However, when the reference method used reflected long-term intake, good correlations were observed only for Mg and β-carotene. Additionally, when BM were used as the reference method, correlations having a very good classification were not observed. None of the micronutrients analysed showed correlations higher than 0·7 using short or long-term dietary assessment instruments or BM as the reference method. Nevertheless, the results presented in FFQ validation studies using short-term or long-term dietary instruments or BM as the reference methods based on correlations from less than three studies should be viewed with caution (Fig. 5).

Fig. 5 Validation of FFQ studies that assess micronutrient intake in children (6–12 years) and adolescents (13–10 years) using short-term or long-term dietary instruments or biomarkers as the reference methods. Correlations: poor ( < 0·30), acceptable (0·30–0·50), good (0·51–0·70) and very good (>0·70). ■, Short-term intake ( < 7 d; three or more studies: vitamin C, iron, calcium, phosphorus); , long-term intake ( ≥ 7 d; three or more studies: vitamin C, calcium, riboflavin, thiamin); ▧, biomarkers.

Biomarkers

A total of five publications analysed BM(Reference Williams and Innis17, Reference Parrish, Marshall and Krebs22, Reference Lietz, Barton and Longbottom23, Reference Green, Allen and O'Connor28, Reference Byers, Trieber and Gunter35), which were used to validate five FFQ. Some articles presented validation of more than one instrument, of which one study also validated estimated dietary records(Reference Williams and Innis17) and two studies validated WDR using BM as the reference method(Reference Lietz, Barton and Longbottom23, Reference Green, Allen and O'Connor28). The BM analysed were: serum markers of Fe status (ferritin, Hb)(Reference Williams and Innis17), plasma levels of vitamins C, D, retinol, β-carotene and α-tocopherol (vitamin E)(Reference Parrish, Marshall and Krebs22), 24 h urine K(Reference Lietz, Barton and Longbottom23), serum folate and serum vitamin B12(Reference Green, Allen and O'Connor28) and serum levels of vitamins C, A and E(Reference Byers, Trieber and Gunter35).

Discussion

In the present review, thirty-two studies(Reference Andersen, Lande and Arsky2, Reference Marriott, Robinson and Poole8Reference Räsänen39) are described. The aim of this analysis was to determine the reliability of methods used to measure the usual intake of vitamins and minerals in infants, children and adolescents and how these were validated. The different studies included in this review were classified according to which reference method was used, those reflecting short-term intake, long-term intake or BM. To rate the different studies, a quality score system was developed by the European micronutrient Recommendations Aligned network. A total score was calculated according to the weighted mean of the correlations that had been adjusted by the quality of the different validation studies, and all the methods were scored into the categories: poor, acceptable, good or very good. Measuring dietary intake in very young children is difficult because of the rapid changes in the food habits of toddlers, the need to rely on parental reporting and the questionable ability of parents to accurately report their child's diet when other caregivers are also involved in feeding the child(Reference Parrish, Marshall and Krebs22). Adolescents' eating habits are highly influenced by family patterns and habits, their peer group, as well as by their increasing concern with body image. Habits such as meal skipping (particularly breakfast), consuming high-energy foods that are poor in nutrients and the tendency, particularly among girls, to restrict their diet and to go on diets are part of the repertoire of adolescent eating behaviour(Reference Slater, Philippi and Fisberg20).

Infants

This group included infants aged 1–23 months. Evidence for the long-term effects of infant nutrition on later health has given impetus to the need for developing methods to assess the diets of infant populations(Reference Owen, Martin and Whincup41). In general, for this group, the daily non-human milk intake in the previous day or week was estimated from the average total volume of bottle-feeds consumed per day; for breast-fed infants, the usual feeding length and number of feeds per day were recorded. In Andersen's study(Reference Andersen, Lande and Arsky2), the relative validity of a semi-quantitative FFQ (140-item semi-quantitative FFQ) used in 12-month-old infants was examined (with correlations ranging from 0·18 for vitamin D to 0·62 for Ca and Fe). This validation study indicated that the semi-quantitative FFQ used in infants overestimated the median intake of nutrients except for Ca intake. In a sample of fifty infants aged 6 months, Marriott et al. (Reference Marriott, Robinson and Poole8) compared the energy and nutrient intakes assessed by a newly developed, interviewer-administered, 34-item infant FFQ with intakes determined from 4 d WDR. Differences in intakes between methods were observed, and the agreement tended to be lower for breast-fed than for formula-fed infants. The correlation coefficients of this study compared favourably with other FFQ validation studies for young children. Reported correlation coefficients for energy and nutrients were 0·18–0·72 at 1 year in Andersen's study(Reference Andersen, Lande and Arsky2), and 0·26–0·63 for 1- to 5-year-olds in Blum's study(Reference Blum, Wei and Rockett24). In the latter study, an 84-item FFQ was used, with correlations ranging from 0·31 for Zn to 0·63 for Mg. Marriott et al. (Reference Marriott, Inskip and Borland9) also evaluated the relative validity of a 76-item FFQ for assessing micronutrient intakes in 12-month-old infants. They reported smaller correlations in infants aged 12 months (with correlations ranging from 0·24 for vitamin B12 to 0·75 for Na) than in the infants aged 6 months (with correlations ranging from 0·55 for Cu to 0·89 for thiamin). Marshall et al. (Reference Marshall, Eichenberger Gilmore and Broffitt21) compared nutrient intakes from beverages in a sample of 6-month-old infants using a 7-item FFQ and obtained coefficients for Ca and vitamin D intakes of 0·64 and 0·80, respectively. Similar results were shown for Marriott et al. (Reference Marriott, Robinson and Poole8) for these nutrients (Ca, r 0·78 and vitamin D, r = 0·83). The ability of the FFQ to accurately rank intakes of energy and all the nutrients in infants is enhanced by the quality and detail of the information collected, which includes information about brands and types of baby foods and milks used(Reference Marriott, Robinson and Poole8). In Williams & Innis(Reference Williams and Innis17), parents of 148 infants aged 8–26 months completed a 191-item FFQ and 3 d EDR for assessing Fe nutrition in infants (with correlations ranging from 0·64 for vitamin C and Fe to 0·75 for Ca). These results showed relative validity of a FFQ in comparison with 3 d EDR for estimating Fe intakes in a group of infants and suggested that while dietary assessments can serve as useful research tools to assess nutrient intakes in 8- to 26-month-old infants, they have limited value as tools to identify infants at risk of Fe deficiency.

Preschool children

In this review, the preschool children group included children aged 2–5 years. Many factors contribute to making intake assessment in this age group difficult: preschool children eat small amounts of food at frequent intervals, they are not able to complete questionnaires and their food habits and nutrient intakes may rapidly change(Reference Hertzler, Bowens and Hull42, Reference Stein, Shea and Basch43). In Huybrechts et al. (Reference Huybrechts, De Bacquer and Matthys15), the relative validity of a 47-item FFQ for measuring preschool children's usual Ca intake was assessed using parents or guardians as a proxy, and EDR were used as the reference method. Based on the comparison of means, the FFQ underestimated the mean Ca intake measured by the EDR. These findings are in contrast to the findings of another study(Reference Taylor and Goulding26), which reported overestimates of actual Ca intakes in young children (3–6 years) using a 35-item FFQ. Andersen et al. (Reference Andersen, Lande and Trygg18) presented the relative validity of food and nutrient intakes estimated by the 125-item FFQ against intakes from 7 d WDR applied in the nation-wide survey among 2-year-old children in Norway. The correlations in this study were also lower than those observed by Blum et al. (Reference Blum, Wei and Rockett24) However, the correlation coefficients for nutrient density values estimated from the two methods were much higher than those seen for absolute nutrient intakes, and for ten out of sixteen nutrients, the correlations were >0·50. The average correlation (median r 0·52) for nutrient density was similar to that observed among 12-month-old Norwegian infants (median r 0·50)(Reference Andersen, Lande and Arsky2). The agreement across quartiles between the two methods was on average (median) 36 %, which is similar to the results observed among 12-month-old infants (38 %)(Reference Andersen, Lande and Arsky2). In Andersen's study(Reference Andersen, Lande and Trygg18), agreement across quartiles increased when using nutrient density. Other authors studied 224 preschool children and obtained dietary data by interviewing the child's mother(Reference Stein, Shea and Basch37). The present study was performed to determine the utility of the Willett semi-quantitative FFQ for assessing habitual diets of preschool children. Results reported associations between nutrient intakes estimated by the FFQ and 24 h recalls, with correlations (adjusted for within-person variability and non-differential measurement error) in the range of 0·20–0·60(Reference Stein, Shea and Basch37). Blum et al. (Reference Blum, Wei and Rockett24) demonstrated that past dietary intake of children aged 1–5 years could be measured reasonably well with an 84-item FFQ completed by child's parent or guardian. Correlation coefficients between the dietary intakes assessed by the two methods ranged from 0·26 for fibre to 0·63 for Mg. All but three nutrients (protein, dietary fibre and Zn) had correlations of 0·47 or higher. After adjusting for energy intake and within-person variation, the average correlation was 0·52.

Children

This group included children aged 6–12 years. The food and nutrient intake of children is important not only for growth and development but also for present and future health, including the prevention of chronic diseases in adulthood(Reference Berenson44, Reference McGill, McMahan and Herderick45). To succeed in nutrition monitoring and epidemiologic research among large groups of children, it is necessary to have a dietary assessment method that is both valid and functional with young age groups(Reference Lillegaard, Løken and Andersen12). Almost all of the reviewed validity and reliability studies in children younger than age 9 included adult assistance in providing information on the child's intake. In general, the reviews concluded that children generally have difficulty in estimating portion sizes(Reference Cypel, Guenther and Petot46). Räsänen(Reference Räsänen39) compared the 24 h recall method with the diet history method as used in a food consumption survey for children. The correlation coefficients between the values obtained by the 24 h recalls and the diet history method varied from 0·20 (vitamin A) to 0·50 (energy). The history method used in this study gave consistently higher mean values than the 24 h recalls. Lytle et al. (Reference Lytle, Nichaman and Obarzanek36) found correlations for nutrient intake between 0·45 and 0·79 when comparing observed and recalled food intake in 8- to 10-year-old children. These are within the same range or somewhat higher than results obtained in Lillegaard's study(Reference Lillegaard, Løken and Andersen12), which evaluated intake of energy, macro- and micronutrients assessed from pre-coded food diaries and by using 4 d WDR as the reference method in 9-year-old Norwegians. The pre-coded food diary is scanable and is developed to simplify the work of the respondents as well as of the researcher. In Belgium, a self-administered computer dietary assessment program ‘Young Adolescents Nutrition Assessment Computer’ was developed, based on the concept of the 24HR. The Young Adolescents Nutrition Assessment Computer is a computer program designed for use with children and adolescents aged 11 years and over(Reference Vereecken, Covents and Matthys47). Vereecken et al. (Reference Vereecken, Covents and Matthys47) assessed the relative validity of the computerised 24 h recall Young Adolescents Nutrition Assessment Computer comparing results with a 1 d EDR and 24 h recall; the Spearman's correlations obtained for the 24 h recall (ranging from 0·44 to 0·86) were comparable to those observed by Lytle et al. (Reference Lytle, Nichaman and Obarzanek36) between 24 h recalled and observed nutrient intake of third-grade children (ranging between 0·45 and 0·79). In the study of Van Horn et al. (Reference Van Horn, Gernhofer and Moag-Stahlberg48), the correlations were slightly higher (0·64–0·96) using electronic methods such as telephone recalls and tape-recorded dietary records. Bellù et al. (Reference Bellù, Riva and Ortisi30) reported on the ability of a 116-item FFQ to estimate the mean nutrient intake at the population level for an Italian paediatric population (mean age = 9·3 years) comparing results with a 24 h recall. Moderate concordance between the two methods was found. Similar results were presented in by Stein et al. (Reference Stein, Shea and Basch37) In another study, Bellù et al. (Reference Bellù, Ortisi and Riva32) tested the validity for the assessment of individual nutrient intake by a 116-item FFQ for Northern Italian school children comparing results with 7 d EDR. In this study, the Pearson's correlation coefficients ranged between 0·30 and 0·58, and for some nutrients (vitamins A and B6) the correlation was found to be low.

Adolescents

For the present paper, the adolescent age group included youth aged 13–18 years. Adolescents are a group whose eating habits are characterised by factors such as irregular meals, snacking and meal-skipping. These eating habits are not easily conducive to diet assessment by a prospective methodology(Reference Lietz, Barton and Longbottom23). Slater et al. (Reference Slater, Philippi and Fisberg20) compared the 76-item FFQ developed for adolescents to data obtained from 24 h recalls, and similar values for energy, carbohydrates, total fat and Ca intake were observed, suggesting a high consistency in estimating these nutrients. However, there was a significant difference for seven remaining nutrients (protein, polyunsaturated fat, dietary fibre, cholesterol, retinol, vitamin C and Fe). Higher values (mean r 0·57) were observed with unadjusted correlations compared with results obtained by Rockett et al. (Reference Rockett, Breitenbach and Frazier29) (mean r 0·39) and Field et al. (Reference Field, Peterson and Gortmaker25) (0·27 for Ca, 0·25 for P, 0·20 for Fe, 0·19 for vitamin C). The YAQ is a youth-friendly FFQ that allows 9- to 18-year-old adolescents to report on their own diet(Reference Lamb, Ross and Brady49). Rockett et al. (Reference Rockett, Breitenbach and Frazier29) evaluated the questionnaire's validity by comparing the nutrient scores of the YAQ with those from the average of three 24 h recalls. Similar mean nutrients were found by both the methods. Correlation coefficients between the mean energy-adjusted nutrients computed by the two methods ranged from 0·21 for Na to 0·58 for folate. After correction for within-person error, the average correlation coefficient was 0·54, similar to that found for adults(Reference Rimm, Giovannucci and Stampfer50). In the validation study developed by Mølgaard et al. (Reference Mølgaard, Sandström and Michaelsen27), twenty-three children aged 13–14 years filled in an 88-item FFQ designed to assess Ca, protein and P intakes and kept 3 d WDR. Spearman rank correlations between nutrient intake values from the FFQ and WDR were 0·56–0·62 (mean 0·60). Another study showed similar findings for Ca intake (r 0·54)(Reference Andersen, Nes and Lillegaard33) using 7 d WDR as the reference method.

Biomarkers

Williams & Innis(Reference Williams and Innis17) showed the validity of nutrient intake estimates using a 191-item FFQ that was assessed by comparing data with biochemical measures of Fe status in infants. These results presented weak correlations for assessment methods (acceptable classification for FFQ v. BM r 0·33 and poor classification for a 3 d EDR v. BM r 0·19). This is the only study presented in this review that analyses biochemical measures of Fe status in infants and possibly, the lower correlation coefficients between Fe intake and serum ferritin observed in this study, as compared to previous studies with older subjects, involve factors inherent to studies in infants but not in older children and adults(Reference Fomon, Ziegler and Nelson51). Lietz et al. (Reference Lietz, Barton and Longbottom23) reported on the correlation coefficient between urinary K and dietary intake from a 130-item FFQ (r − 0·04) and from a 7 d WDR (r 0·78) in fifty children between 11 and 13 years. The latter correlation was higher than correlations reported in a previous study(Reference McKeown, Day and Welch52).

Green et al. (Reference Green, Allen and O'Connor28) validated a 116-item FFQ and a 3 d WDR by comparing nutrient intake estimates using these methods with serum folate and serum vitamin B12 concentrations in 105 females aged 16–19 years. Using serum folate concentration as the sole biochemical criterion, it appeared that the 3 d WDR was superior to the FFQ as a tool to predict folate intakes. The correlations between folate intakes and serum folate as determined by 3 d WDR (adjusted r 0·65) showed higher correlations than those determined by FFQ (r 0·48; supplement users were included in these analyses). Vitamin B12 intake as determined by either of the dietary methods showed only a modest association with serum vitamin B12 concentrations, when supplement users were included in the analyses (3 d WDR r 0·32; FFQ r 0·25). Another author(Reference Jacques, Sulsky and Sadowski53) using a FFQ reported a correlation of r 0·60 between folate intakes and plasma folate concentrations and a correlation of r 0·35 between vitamin B12 intakes and serum vitamin B12 concentrations in a group of 139 adults aged 40–83 years.

The correlations observed by Parrish et al. (Reference Parrish, Marshall and Krebs22) compared the reported dietary intake in infants and preschool children using a 111-item FFQ with plasma concentrations of the micronutrient. The highest correlation between plasma concentrations and dietary intake as measured by the FFQ was for vitamin C (r 0·51), and the correlations were weak or absent for vitamin D, retinol and β-carotene. Byers et al. (Reference Byers, Trieber and Gunter35) administered a 111-item FFQ to ninety-seven parents of children aged 6–10 years to evaluate their children's usual dietary intake and compared results with children's serum levels of vitamins C, A and E. The correlations observed in this study between circulating levels of vitamin C (r 0·37) and vitamin E (r 0·32), and the parent's reports of their children's diets were similar to those observed in dietary validation studies conducted in adults(Reference Coates, Eley and Block54). However, it is important to keep in mind that the results presented in FFQ validation studies using BM as the reference method based on correlations from only one or two studies should be interpreted with caution (Figs. 4 and 5).

Conclusion

When comparing different validation methods in infants, the reviewed studies showed that the FFQ was the method used in all the studies to assess micronutrient intake in infants. Different FFQ were validated for which wide variations in the number of food items were observed (7–191 items). The ability of the FFQ to accurately rank intakes of energy and all the nutrients in infants is enhanced by the quality and detail of the information collected, which includes information about brands and types of baby foods and milks used(Reference Marriott, Robinson and Poole8). The WDR used as the reference method for evaluating FFQ present better correlations for several micronutrients than other methods in this population group. However, we must emphasise that this result is probably due to the fact that 57 % of the studies analysed in infants used WDR as the reference method. Vitamin D and vitamin E intake analysed using FFQ v. WDR showed acceptable correlations, and good correlations were observed for vitamin C, thiamin, riboflavin, Ca and Fe. For the rest of the micronutrients, there were insufficient data to conduct an analysis (three or more studies were needed for each micronutrient).

In the preschool children cluster, different FFQ were validated for which wide variations in the number of food items were observed (7–125 items).The FFQ was used as the method to assess micronutrient intake in 80 % of the studies in preschool children. Comparison of different dietary assessment methods was difficult in this group, as there were not enough studies to analyse for each micronutrient. Except for the measurement of Ca intake, acceptable correlations (r 0·50) were observed when FFQ were validated applying WDR and EDR as the reference methods. This age group typically eat more frequently, and since 50 % of the preschool children consume two meals outside of their homes, parents may not be able to accurately describe what was actually consumed(Reference Iannotti, Zuckerman and Blyer34).

The results presented in the FFQ validation studies for infants and preschool children, using long-term dietary instruments or BM as the reference methods were based on only one or two correlation studies and as such, there were insufficient data to reach robust conclusions. In regards to the reference method that reflected short-term intake in the infants and preschool children group, good correlations were observed for niacin, thiamin, vitamins B6, D, C, E, riboflavin, Ca, K, Mg, Fe, and Zn (with mean weighted correlations ranging from 0·55 for vitamin E to 0·69 for niacin). For the analysis of these micronutrients, data from three or more studies were included (the study quality scores ranging from 2·5 to 5·5).

For the children group, 65 % of the reviewed studies used FFQ to assess micronutrient intake for which wide variations in the number of food items were observed (10–175 items). When comparing different validation methods in this group, only an acceptable correlation (r 0·50) was observed for measuring of Ca intake, when FFQ were validated applying WDR and EDR as the reference methods. Similar correlations were observed when the validity of Ca intake estimates using the FFQ (including YAQ) was assessed by comparing them with the Ca intake estimates using a 24HR (r 0·49). For the other micronutrients, there were insufficient data to compare.

Assessing the validation of different dietary assessment methods in adolescents was difficult as there were not enough studies to compare for each micronutrient. A good correlation (r 0·58) was observed when FFQ were validated applying WDR as the reference method only for Ca intake. Additionally, when FFQ were validated using a 24 HR as the reference method, correlations for Ca presented acceptable values (r 0·50). In the adolescent group, 80 % of the reviewed studies used FFQ to assess micronutrient intake for which wide variations in the number of food items were observed (10–190 items).

There were not enough data to contrast FFQ studies that assessed micronutrient intake in children and adolescents using long-term dietary instruments (except for thiamin, riboflavin, Ca and vitamin C intake that showed acceptable correlations) or BM as the reference methods. (Only one or two studies, thus being insufficient to reach strong conclusions). Regarding the reference method that reflected short-term intake in children and adolescents, good correlations were observed only for vitamin C (r 0·61) and Ca (r 0·51). In the analysis of these micronutrients, data from three or more studies were included (the study quality scores ranging from 2 to 4·5).

In the studies reviewed, FFQ comprised the dietary method that was most utilised to assess the micronutrient intake in these groups, in which it is of utmost importance to recognise methodological aspects such as food composition databases used for analysis, portion size assessment and the time periods between the two dietary assessment methods(Reference Lietz, Barton and Longbottom23). On the other hand, dietary assessment in children and adolescents using electronic methods for deriving self-reports, such as the tape-recorded food record or the computer dietary assessment program, may help improve compliance and frequency of record keeping in these age groups(Reference Van Horn, Gernhofer and Moag-Stahlberg48).

The micronutrients analysed in this review using BM as the reference method were: vitamins A, C, D, E, and B12, retinol, β-carotene, folate and K. In general, the best correlations were observed when the validity of nutrient intake estimates using WDR was assessed by comparing them with micronutrient serum levels. Validation of FFQ studies that assess micronutrient intake in the infant and preschool children group using BM as the reference method presented good correlations only for vitamin C. Poor rankings (r < 0·3) were observed for retinol, β-carotene and K, and acceptable rankings were observed only for vitamin E and Fe. In the children and adolescent group, the validation of FFQ studies that assess micronutrient intake comparing them with biochemical measures of micronutrient status presented poor correlations for vitamins A and B12 and acceptable correlations for folate, vitamins E and C. However, we must emphasise once more that the data presented in FFQ validation studies using BM as the reference method showed correlations from only one or two studies; as such, this information should be viewed with caution. Including supplement users generally improved the correlations between micronutrients intakes estimated by either of the dietary intake methods and their respective biochemical indices. Nutrient BM are appealing as a comparison method because measurement errors are uncorrelated with reporting errors(Reference Willett, Sampson and Stampfer55). However, given that blood concentrations may be a result of absorption and metabolism and not only reflecting intake, nutrient BM may not always be an appropriate method of comparison(Reference Parrish, Marshall and Krebs22).

Acknowledgements

The studies reported herein have been carried out within the European micronutrient Recommendations Aligned Network of Excellence (www.eurreca.org), financially supported by the Commission of the European Communities, specific Research, Technology and Development (RTD) Programme Quality of Life and Management of Living Resources, within the Sixth Framework Programme, contract no 036196. This report does not necessarily reflect the Commission's views or its future policy in this area. A. O.-A. wrote the first manuscript; A. O.-A., P. H.-S. and A. S.-V. contributed to the planning of the search and analysed the articles included in this review; M. M. analysed the articles included in this review; L. P.-Q. and M. M. revised and discussed previous drafts; L. S.-M. contributed to the planning of the search, decided the analysis and presentation of the results and created the quality assessment tool of the articles, and revised and discussed previous drafts. There are no conflicts of interest to report. The authors of the present paper would like to thank Dr Margaret Ashwell, Dr Janet Lambert, Dr Adriënne Cavelaars, Dr Olga Souverein and Mrs Sandra Crispim for their technical contribution and advice to this publication. The authors also thank the Health Sciences Library of the University of Las Palmas de Gran Canaria, Lourdes Ribas, and Nuria Melián for their contribution in the collection of articles analysed in this review. Special thanks to Joy Ngo, RD from the Nutrition Research Foundation (FIN) for her help in editing the English version of the manuscript.

References

1Mascarenhas, MR, Zemel, B & Stallings, VA (1998) Nutritional assessment in paediatrics. Nutrition 14, 105115.Google Scholar
2Andersen, LF, Lande, B, Arsky, GH, et al. (2003) Validation of a semi-quantitative food-frequency questionnaire used among 12-month-old Norwegian infants. Eur J Clin Nutr 57, 881888.Google Scholar
3Baranowski, T & Domel, SB (1994) A cognitive model of children's reporting of food intake. Am J Clin Nutr 59, S212S217.Google Scholar
4Livingstone, MB & Robson, PJ (2000) Measurement of dietary intake in children. Proc Nutr Soc 59, 279293.Google Scholar
5Sobo, EJ & Rock, CL (2001) ‘You ate all that!?’: caretaker-child interaction during children's assisted dietary recall interviews. Med Anthropol Q 15, 222244.Google Scholar
6Frank, GC (1997) Methodological issues regarding eating behavior of high-risk adolescents. Ann N Y Acad Sci 817, 6682.Google Scholar
7Ashwell, M, Lambert, JP, Alles, MS, et al. (2008) How we will produce the evidence-based EURRECA toolkit to support nutrition and food policy. Eur J Nutr 47, Suppl. 1, 216.Google Scholar
8Marriott, LD, Robinson, SM, Poole, J, et al. (2008) What do babies eat? Evaluation of a food frequency questionnaire to assess the diets of infants aged 6 months. Public Health Nutr 11, 751756.Google Scholar
9Marriott, LD, Inskip, HM, Borland, SE, et al. (2008) What do babies eat? Evaluation of a food frequency questionnaire to assess the diets of infants aged 12 months. Public Health Nutr 12, 967972.Google Scholar
10Holmes, B, Dick, K & Nelson, M (2008) A comparison of four dietary assessment methods in materially deprived households in England. Public Health Nutr 11, 444456.Google Scholar
11Moore, M, Braid, S, Falk, B, et al. (2007) Daily calcium intake in male children and adolescents obtained from the rapid assessment method and the 24-hour recall method. Nutr J 6, 2428.Google Scholar
12Lillegaard, IT, Løken, EB & Andersen, LF (2007) Relative validation of a pre-coded food diary among children, under-reporting varies with reporting day and time of the day. Eur J Clin Nutr 61, 6168.Google Scholar
13Magkos, F, Manios, Y, Babaroutsi, E, et al. (2006) Development and validation of a food frequency questionnaire for assessing dietary calcium intake in the general population. Osteoporos Int 17, 304312.Google Scholar
14Harnack, LJ, Lytle, LA, Story, M, et al. (2006) Reliability and validity of a brief questionnaire to assess calcium intake of middle-school-aged children. J Am Diet Assoc 106, 17901795.Google Scholar
15Huybrechts, I, De Bacquer, D, Matthys, C, et al. (2006) Validity and reproducibility of a semi-quantitative food-frequency questionnaire for estimating calcium intake in Belgian preschool children. Br J Nutr 95, 802816.Google Scholar
16Bertoli, S, Petroni, ML, Pagliato, E, et al. (2005) Validation of food frequency questionnaire for assessing dietary macronutrients and calcium intake in Italian children and adolescents. J Pediatr Gastroenterol Nutr 40, 555560.Google Scholar
17Williams, PL & Innis, SM (2005) Food frequency questionnaire for assessing infant iron nutrition. Can J Diet Pract Res 66, 176182.Google Scholar
18Andersen, LF, Lande, B, Trygg, K, et al. (2004) Validation of a semi-quantitative food-frequency questionnaire used among 2-year-old Norwegian children. Public Health Nutr 7, 757764.Google Scholar
19Andersen, LF, Lande, B, Trygg, K, et al. (2004) Validation of a semi-quantitative food-frequency questionnaire used among 2-year-old Norwegian children – Corrigendum. Public Health Nutr 12, 10261027.Google Scholar
20Slater, B, Philippi, ST, Fisberg, RM, et al. (2003) Validation of a semi-quantitative adolescent food frequency questionnaire applied at a public school in São Paulo, Brazil. Eur J Clin Nutr 57, 629635.Google Scholar
21Marshall, TA, Eichenberger Gilmore, JM, Broffitt, B, et al. (2003) Relative validation of a beverage frequency questionnaire in children ages 6 months through 5 years using 3-day food and beverage diaries. Am Diet Assoc 103, 714720.Google Scholar
22Parrish, LA, Marshall, JA, Krebs, NF, et al. (2003) Validation of a food frequency questionnaire in preschool children. Epidemiology 14, 213217.Google Scholar
23Lietz, G, Barton, KL, Longbottom, PJ, et al. (2002) Can the EPIC food-frequency questionnaire be used in adolescent populations? Public Health Nutr 5, 783789.Google Scholar
24Blum, RE, Wei, EK, Rockett, HR, et al. (1999) Validation of a food frequency questionnaire in Native American and Caucasian children 1 to 5 years of age. Matern Child Health J 3, 167172.Google Scholar
25Field, AE, Peterson, KE, Gortmaker, SL, et al. (1999) Reproducibility and validity of a food frequency questionnaire among fourth to seventh grade inner-city school children: implications of age and day-to-day variation in dietary intake. Public Health Nutr 2, 293300.Google Scholar
26Taylor, RW & Goulding, A (1998) Validation of a short food frequency questionnaire to assess calcium intake in children aged 3 to 6 years. Eur J Clin Nutr 52, 464465.Google Scholar
27Mølgaard, C, Sandström, B & Michaelsen, KF (1998) Evaluation of a food frequency questionnaire for assessing of calcium, protein and phosphorus intakes in children and adolescents. Scand J Nutr 42, 25.Google Scholar
28Green, TJ, Allen, OB & O'Connor, DL (1998) A three-day weighed food record and a semiquantitative food-frequency questionnaire are valid measures for assessing the folate and vitamin B-12 intakes of women aged 16 to 19 years. J Nutr 128, 16651671.Google Scholar
29Rockett, HR, Breitenbach, M, Frazier, AL, et al. (1997) Validation of a youth/adolescent food frequency questionnaire. Prev Med 26, 808816.Google Scholar
30Bellù, R, Riva, E, Ortisi, MT, et al. (1996) Validity of a food frequency questionnaire to estimate mean nutrient intake of Italian school children. Nutr Res 16, 197200.Google Scholar
31Arnold, JE, Rohan, T, Howe, G, et al. (1995) Reproducibility and validity of a food-frequency questionnaire designed for use in girls age 7 to 12 years. Ann Epidemiol 5, 369377.Google Scholar
32Bellù, R, Ortisi, MT, Riva, E, et al. (1995) Validity assessment of a food frequency questionnaire for school-age children in Northern Italy. Nutr Res 15, 11211128.Google Scholar
33Andersen, LF, Nes, M, Lillegaard, IT, et al. (1995) Evaluation of a quantitative food frequency questionnaire used in a group of Norwegian adolescents. Eur J Clin Nutr 49, 543554.Google Scholar
34Iannotti, RJ, Zuckerman, AE, Blyer, EM, et al. (1994) Comparison of dietary intake methods with young children. Psychol Rep 74, 883889.Google Scholar
35Byers, T, Trieber, F, Gunter, E, et al. (1993) The accuracy of parental reports of their children's intake of fruits and vegetables: validation of a food frequency questionnaire with serum levels of carotenoids and vitamins C, A, and E. Epidemiology 4, 350355.Google Scholar
36Lytle, LA, Nichaman, MZ, Obarzanek, E, et al. (1993) Validation of 24-hour recalls assisted by food records in third-grade children. The CATCH Collaborative Group. J Am Diet Assoc 93, 14311436.Google Scholar
37Stein, AD, Shea, S, Basch, CE, et al. (1992) Consistency of the Willett semiquantitative food frequency questionnaire and 24-hour dietary recalls in estimating nutrient intakes of preschool children. Am J Epidemiol 15, 667677.Google Scholar
38Jenner, DA, Neylon, K, Croft, S, et al. (1989) A comparison of methods of dietary assessment in Australian children aged 11–12 years. Eur J Clin Nutr 43, 663673.Google Scholar
39Räsänen, L (1979) Nutrition survey of Finnish rural children. VI. Methodological study comparing the 24-hour recall and the dietary history interview. Am J Clin Nutr 32, 25602567.Google Scholar
40Treadwell, JR, Tregear, SJ, Reston, JT, et al. (2006) A system for rating the stability and strength of medical evidence. BMC Med Res Methodol 6, 5270.Google Scholar
41Owen, CG, Martin, RM, Whincup, PH, et al. (2005) Effect of infant feeding on the risk of obesity across the life course: a quantitative review of published evidence. Pediatrics 115, 13671377.Google Scholar
42Hertzler, AA, Bowens, J & Hull, S (1993) Preschoolers' reporting of food habits. J Am Diet Assoc 93, 11591161.Google Scholar
43Stein, AD, Shea, S, Basch, CE, et al. (1991) Variability and tracking of nutrient intakes of preschool children based on multiple administrations of the 24-hour dietary recall. Am J Epidemiol 134, 14271437.Google Scholar
44Berenson, GS (2002) Childhood risk factors predict adult risk associated with subclinical cardiovascular disease. The Bogalusa Heart Study. Am J Cardiol 90, 3L7L.Google Scholar
45McGill, HC Jr, McMahan, CA, Herderick, EE, et al. (2000) Origin of atherosclerosis in childhood and adolescence. Am J Clin Nutr 72, Suppl. 5, 13071315.Google Scholar
46Cypel, YS, Guenther, PM & Petot, GJ (1997) Validity of portion-size measurement aids: a review. J Am Diet Assoc 97, 289292.Google Scholar
47Vereecken, CA, Covents, M, Matthys, C, et al. (2005) Young adolescents' nutrition assessment on computer (YANA-C). Eur J Clin Nutr 59, 658667.Google Scholar
48Van Horn, LV, Gernhofer, N, Moag-Stahlberg, A, et al. (1990) Dietary assessment in children using electronic methods: telephones and tape recorders. J Am Diet Assoc 90, 412416.Google Scholar
49Lamb, MM, Ross, CA, Brady, HL, et al. (2007) Comparison of children's diets as reported by the child via the Youth/Adolescent Questionnaire and the parent via the Willett food-frequency questionnaire. Public Health Nutr 10, 663670.Google Scholar
50Rimm, EB, Giovannucci, EL, Stampfer, MJ, et al. (1992) Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am J Epidemiol 135, 11141126.Google Scholar
51Fomon, SJ, Ziegler, EE, Nelson, SE, et al. (1981) Cow milk feeding in infancy: gastrointestinal blood loss and iron nutritional status. J Pediatr 98, 540545.Google Scholar
52McKeown, NM, Day, NE, Welch, AA, et al. (2001) Use of biological markers to validate self-reported dietary intake in a random sample of the European Prospective Investigation into Cancer United Kingdom Norfolk cohort. Am J Clin Nutr 74, 188196.Google Scholar
53Jacques, PF, Sulsky, SI, Sadowski, JA, et al. (1993) Comparison of micronutrient intake measured by a dietary questionnaire and biochemical indicators of micronutrient status. Am J Clin Nutr 57, 182189.Google Scholar
54Coates, RJ, Eley, JW, Block, G, et al. (1991) An evaluation of a food frequency questionnaire for assessing dietary intake of specific carotenoids and vitamin E among low-income black women. Am J Epidemiol 134, 658671.Google Scholar
55Willett, WC, Sampson, L, Stampfer, MJ, et al. (1985) Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 122, 5165.Google Scholar
Figure 0

Fig. 1 Main stages of the systematic review process.

Figure 1

Table 1 Inclusion and exclusion criteria

Figure 2

Table 2 Characteristics of included studies

Figure 3

Table 3 Validation studies in infants, children and adolescents: vitamins

Figure 4

Table 4 Validation studies in infants, children and adolescents: minerals

Figure 5

Fig. 2 Comparison of different dietary assessment methods in infants (1–23 months) and preschool children (2–5 years) by vitamins and minerals (mean of quality weighted correlation coefficients) 24 HR, 24 h recall; EDR, estimated dietary record; WDR, weighed dietary record; BM, biomarker; DH, diet history. (a) Infants (three or more studies: vitamin D, four studies FFQ v. WDR; vitamin E, three studies FFQ v. WDR; vitamin C, three studies FFQ v. WDR; thiamin, three studies FFQ v. WDR; riboflavin, four studies FFQ v. WDR; calcium, four studies FFQ v. WDR; iron, four studies FFQ v. WDR). , FFQ v. 24 HR (one study(24)); , FFQ v. WDR (four studies(2,8,9,21)); , FFQ v. EDR (one study(17)); , FFQ v. BM (one study(22)). (b) Preschool children (three or more studies: not for any micronutrient). , FFQ v. 24 HR (three studies(24,34,37)); , FFQ v. WDR (two studies(18,19,21)); , FFQ v. BM (one study(22)); , 24 HR v. DH (one study(39)).

Figure 6

Table 5 Classification of the dietary assessment methods for infants, children and adolescents according to the weighted mean of the correlations of each micronutrient (including three or more studies)

Figure 7

Fig. 3 Comparison of different dietary assessment methods in children (6–12 years) and adolescents (13–10 years) by vitamins and minerals (mean of quality weighted correlation coefficients) 24 HR, 24 h recall; EDR, estimated dietary record; WDR, weighed dietary record; DH, dietary history; YAQ, Youth/Adolescent Questionnaire. (a) Children (three or more studies: calcium, five studies FFQ v. 24 HR). , FFQ v. EDR (three studies(26,31,36)); , FFQ v. WDR (two studies(16,23)); , FFQ v.BM (one study(35)); , 24 HR v. DH (one study(39)); , YAQ v. 24 HR (one study(29)); , FFQ v. 24 HR (five studies(11,13,16,25,38)). (b) Adolescents (three or more studies: calcium: three studies FFQ v. 24 HR; three studies FFQ v. WDR). , FFQ v. 24 HR (three studies(11,14,20)); , FFQ v. WDR (three studies(16,27,33)); , 24 HR v. DH (one study(39)); , YAQ v. 24 HR (one study(29)).

Figure 8

Fig. 4 Validation of FFQ studies that assess micronutrient intake in infants (1–23 months) and preschool children (2–5 years) using as the reference method: short-term or long-term dietary instruments or biomarkers. Correlations: poor ( < 0·30), acceptable (0·30–0·50), good (0·51–0·70) and very good (>0·70). Three or more studies: sodium, vitamins B12, E, C, D, B6, zinc, iron, magnesium, potassium, calcium, riboflavin, thiamin, niacin. ■, Short-term intake ( < 7 d); , long-term intake ( ≥ 7 d); ▧, biomarkers.

Figure 9

Fig. 5 Validation of FFQ studies that assess micronutrient intake in children (6–12 years) and adolescents (13–10 years) using short-term or long-term dietary instruments or biomarkers as the reference methods. Correlations: poor ( < 0·30), acceptable (0·30–0·50), good (0·51–0·70) and very good (>0·70). ■, Short-term intake ( < 7 d; three or more studies: vitamin C, iron, calcium, phosphorus); , long-term intake ( ≥ 7 d; three or more studies: vitamin C, calcium, riboflavin, thiamin); ▧, biomarkers.