The inclusion of whole-grain-rich foods in the daily diet is widely promoted, with epidemiological studies showing regular consumption of whole grain foods (three servings or 48 g/d) to be associated with a reduced risk of CVD, diabetes(Reference McKeown, Meigs and Liu1–Reference Jensen, Koh-Banerjee and Franz3), certain cancers(Reference Slavin4) and obesity(Reference Harland and Garton5). A number of randomised controled trials have also produced encouraging results, with increasing intakes favourably influencing LDL-cholesterol, systolic blood pressure, insulin and C-reactive protein(Reference Pereira, Jacobs and Pins6–Reference Giacco, Clemente and Cipriano8). However, much of the current research has concentrated on the adult population, with few studies considering the effects of whole grain consumption in younger age groups(Reference Gold, Wong and Tong9, Reference Steffen, Jacobs and Murtaugh10).
Whole grains have been defined by the American Association of Cereal Chemists(11) as consisting of ‘the intact, ground, cracked or flaked caryopsis whose principal anatomical components (the endosperm, germ and bran) are present in the same relative proportions as they exist in the intact caryopsis’. However, the definition of what qualifies as a whole grain food is widely debated, and has resulted in a variety of health and content claims. One such health claim, in relation to heart disease and certain cancers, defines whole grain foods as those containing 51 % or more whole grain ingredients(Reference Richardson12). In comparison, to use a whole grain stamp on food packaging in Canada and the USA, the food must contain at least 8 g of whole grain ingredients per serving(13). Few official government regulations define how much whole grain must be in a product to be deemed a whole grain food. As a result, studies assessing population whole grain intakes have differed in their criterion as to what qualifies as a whole grain food, with some studies including all foods containing a source of whole grain(Reference Alexy, Zorn and Kersting14), while others only include foods made with a certain amount of whole grain(Reference Thane, Jones and Stephen15).
Dietary recommendations for daily whole grain intakes have been proposed, with suggested intakes ranging from 48 g in the USA and Australia(16, Reference Griffiths17), to 75 g in Sweden and Denmark(Reference Mejborn, Biltoft-Jensen and Trolle18), to 115 g in the Netherlands(19); however, the UK(20) and Germany(21) have a more vague approach, and suggest that ‘cereal products such as bread, pasta, rice and grain should preferably be whole grain’. Despite these recommendations, present intakes of whole grain in adults remain low, with research showing daily intakes in the USA(Reference O'Neil, Zanovec and Cho22) and UK(Reference Thane, Jones and Stephen23) to be 11 and 23 g, respectively. Intakes of whole grain in children and teenagers are also reported to be low, with daily intakes of 9·8, 13 and 26·4 g in the USA(Reference Zanovec, O'Neil and Cho24), UK(Reference Thane, Jones and Stephen15) and Germany(Reference Alexy, Zorn and Kersting14), respectively, and just 5 and 8 % of children and teenagers are meeting recommendations in Denmark(Reference Mejborn, Biltoft-Jensen and Trolle18) and France(25), respectively.
To date, there is no information on patterns of whole grain consumption in the Irish population, and so the aim of the present study was to quantify the intake of whole grains in the diets of Irish children and teenagers, and to assess the major sources of intake across age groups and mealtimes using data from the National Children's Food Survey (NCFS) and the National Teens' Food Survey (NTFS).
Methodology
National Children's Food Survey and National Teens' Food Survey
The NCFS and the NTFS were cross-sectional surveys conducted in representative samples of Irish children and teenagers between 2003–4 and 2005–6, respectively (Irish Universities Nutrition Alliance; IUNA, http://www.iuna.net). These studies were conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving human subjects/patients were approved by the St James's Hospital and Federated Dublin Voluntary Hospitals Joint Research Ethics Committee (NCFS) and the University College Cork Clinical Research Ethics Committee of the Cork Teaching Hospitals (NTFS). Written informed consent was obtained from all participants and their parents as appropriate. In total, 1035 participants took part (594 children: 293 boys, 301 girls and 441 teenagers: 224 males, 217 females). The age of the children was in the range from 5 to 12 years and that of the teenagers from 13 to 17 years. The overall response rate was 66 % for the NCFS and 63 % for the NTFS. The samples were representative of both population groups with respect to sex, age, location and social class, as per the most recent Irish census(Reference Central26). More detailed information about the population groups can be found at http://www.iuna.net
Data collection
Food and beverage intake was determined using a 7 d weighed food diary for the NCFS and a 7 d semi-weighed food diary for the NTFS. Detailed information was recorded regarding the amount and types of all foods and beverages consumed over the seven consecutive survey days (at brand level where possible). Food and beverage intakes were assessed using WISP version 3.0 (Tinuviel Software), which uses data from the McCance and Widdowson's ‘The Composition of Foods’ fifth and sixth editions(Reference Holland, Welch and Unwin27, Reference Agency28), plus supplemental volumes(Reference Holland, Unwin and Buss29–Reference Holland, Welch and Buss37). Food package labelling was also used to assess food and beverage intakes where necessary. Each food and beverage was allocated an individual food code and subsequently assigned to one of sixty-eight food groups for ease of analysis, e.g. savoury snacks and ready-to-eat breakfast cereals (RTEBC).
In addition to comprehensive dietary information, detailed data on socio-economic level and lifestyle were also collected. Full study details, including quantification methods used, are described by Joyce et al. (Reference Joyce, McCarthy and Gibney38).
The National Food and Ingredient database
During the NCFS and the NTFS, food packaging of the consumed foods and beverages were collected and used to record information about the products. The information gathered created the Irish National Food and Ingredient Database(Reference Gilsenan, Lambe and Gibney39). General information recorded included brand name, product description, portion weight, country of origin and manufacturers' details. In addition, ingredient listings and nutritional information per 100 g of product were recorded. Where possible, foods were recorded at brand level and allocated an individual brand ID that could be linked back to actual consumption patterns.
Calculation of whole grain intakes
The NCFS and the NTFS databases were initially searched to identify potential foods made with whole grain. From the sixty-eight original food groups, fourteen were identified as containing whole grain.
All foods within each of these fourteen food groups were then checked for the presence of whole grain and a new whole grain database was created. All foods made with whole grain ingredients were included in the present study. The grains included in the present study were grouped into one of six groups: wheat, oats, rice, maize, barley and rye (Table 1).
The amount of whole grain in each product per 100 g was recorded in the database. As the NCFS and the NTFS collected information at brand level where possible, values for whole grain could be ascribed at a brand-specific level. These values were obtained either from the food package labelling (Irish National Food and Ingredient Database) using quantitative ingredient declaration percentages or directly from the manufacturer. Where brand-specific data were not available, the whole grain content of a comparable product was used (if the ingredient list of a known whole grain product was identical to that of an unknown product); an average of similar products was used where such detail was lacking. A recipe database previously created from the NCFS and the NTFS was also used where appropriate, taking water losses from cooking and processing into account. The whole grain values collected per 100 g of product were then used to calculate the actual amounts of whole grain consumed.
For ease of analysis, the original sixty-eight food groups were reduced to ten food groups, one of which was a whole grain food group. This whole grain food group was subsequently broken down into the following seven subgroups: (1) rice, pasta and their dishes; (2) breads and rolls; (3) RTEBC; (4) other breakfast cereals (including porridge and cooked breakfast cereals); (5) sweet biscuits, cereal bars and desserts; (6) savoury biscuits, crackers and popcorn; and (7) meat products, tofu and yoghurts.
Statistical analysis
Statistical analyses were carried out using PASW Statistics version 18.0 (SPSS, Inc.). The mean and standard deviations for whole grain consumed according to sex, age group, social class, eating location and day of the week were calculated for both the total population and consumers only. Differences within the groups were detected using independent t tests or by one-way ANOVA as appropriate. Independent samples t tests were also used to test for significant differences in mean daily intakes of macro- and micronutrients for consumers and non-consumers of whole grain. In addition, the percentage contribution of each whole grain food group was analysed and the type of grains consumed were identified. Tertile analysis was also carried out in relation to whole grain intakes and values were adjusted for age and sex using ANCOVA. The Bonferroni post hoc test was used to identify the significant differences between each of the groups.
Further analysis was carried out to determine the proportion of Irish children and teenagers meeting whole grain recommendations set out by the United States Department of Agriculture. The data were assessed using both the recommendation of three servings (or 48 g)/d(16) and using recommendations accounting for differences in energy intake(16, Reference Gidding, Dennison and Birch40). For both methods, two approaches were used to assess adherence. First, the percentage of individuals in a population who met the dietary target for whole grain for the total population was calculated and referred to as ‘approach 1’. However, population goals relate to the recommended mean intake of a population and it is not necessary for all individuals to achieve this intake. Compliance was also, therefore, estimated by the method of Wearne & Day(Reference Wearne and Day41) and referred to as ‘approach 2’. This approach calculated the maximum size of a subgroup of the population, known as compliers, whose mean intake equals the population dietary recommendation. To calculate the percentage of compliers meeting the recommendation, mean daily whole grain intakes of each individual were ranked in descending order from the highest to the lowest. The mean intake of the group was calculated. The individual with the lowest mean intake of whole grain was then removed and the calculation repeated. This procedure continued until the mean result of the group was ≥ 48 g. The percentage of children and teenagers achieving the whole grain recommendation was then calculated.
Results
Whole grain intake
Tables 2 and 3 present the mean daily intakes of whole grain (g/d) for all children and teenagers. The data are described for sex, age group, social class, eating location and day of the week, for both the total population and whole grain consumers only. Overall, 90·5 % of children and 86·6 % of teenagers were consumers of whole grain. Mean daily intakes were 18·5 and 23·2 g/d for children and teenagers, respectively, in the total population, rising to 20·5 g/d (children) and 26·9 g/d (teenagers) in consumers only. In both age groups, males consumed significantly higher amounts of whole grain than females (22·0 g/d in children and 28·6 g/d in teenagers v. 15·1 g/d in children and 17·7 g/d in teenagers). Intakes of whole grain also tended to increase with age, although this was only significant for teenagers, with values in ‘consumers only’ increasing from 23·1 g/d (13–14 years) to 29·9 g/d (15–17 years). The home was the primary location for whole grain consumption (P< 0·001), with minimal amounts eaten at other homes (e.g. home of friends, relatives and child minders) and outside of the home (e.g. restaurants). There was no significant difference between whole grain consumption at weekends in comparison to weekdays.
a,b Mean values with unlike superscript letters were significantly different between mean daily intakes of whole grain (ANOVA with Bonferroni correction).
Mean values were significantly different: * P< 0·05, ** P< 0·01, *** P< 0·001, NS (P≥ 0·05).
† Age group 1: 5–8 years; age group 2: 9–12 years.
a,b Mean values with unlike superscript letters denote significant differences between mean daily intakes of whole grain (ANOVA with Bonferroni correction).
Mean values were significantly different: * P< 0·05, ** P< 0·01, *** P< 0·001, NS (P≥ 0·05).
† Age group 1: 13–14 years; age group 2: 15–17 years.
Sources of whole grain intake
For both children and teenagers, wheat was the major grain contributing to whole grain intake, providing 65 and 67 % of whole grain intakes, respectively. Oats were the second highest contributor (16·2 and 18·4 %, respectively), with maize, rice, barley and rye providing lesser amounts of whole grain to the diet (Fig. 1).
Mean daily whole grain intakes (g/d) and the percentage contribution (%) of the seven whole grain food groups to total whole grain intake are presented for whole grain consumers and across tertiles of intake in Table 4. For both children and teenagers, the main food group contributor to whole grain intakes was the RTEBC group (59·3 and 44·3 %, respectively), followed by bread and rolls (14·4 and 26·5 %, respectively). Across the tertiles of intake, those in the highest tertile had significantly greater intakes of the food groups: ‘RTEBC’; ‘bread and rolls’; ‘savoury biscuits, crackers and popcorn’; and ‘other breakfast cereals’, in comparison to those children and teenagers in the lowest tertile.
% Cons, % consumers of each whole grain food group; % Ctr, % contribution of food groups to whole grain intakes; NA, not available.
a,b,c Mean values with unlike superscript letters were significantly different between the mean values of whole grain intake across the tertiles of whole grain intake (adjusted for sex) (ANCOVA, Bonferroni post hoc test).
Mean values were significantly different: * P< 0·05; *** P< 0·001; NS (P≥ 0·05).
As expected, for both children and teenagers, the majority of whole grain was consumed at breakfast (62·4 and 54·0 %), with more than 75 % of whole grain eaten at this meal coming from RTEBC. The next most popular time to consume whole grain foods was as a snack, followed by lunch and dinner. This was evident for both children and teenagers. For children, the most popular whole grain food consumed as a snack was ‘savoury biscuits, crackers and popcorn’ (48 %), while ‘bread and rolls’ (33 %) was the most popular choice for teenagers. ‘Bread and rolls’ was also the main contributor to whole grain intakes at lunch for both children (54 %) and teenagers (79 %), while the ‘rice, pasta and their dishes’ food group was the main contributor at dinner (37 and 74·3 % for children and teenagers, respectively) (data not shown).
Whole grain intakes and dietary recommendations
The percentage of Irish children and teenagers meeting whole grain recommendations(16) using both ‘approach 1’ and ‘approach 2’, are presented in Table 5. Using approach 1, which assesses individual compliance with the recommendations, over 5 % of children and almost 15 % of teenagers were meeting the whole grain recommendation of three servings (48 g)/d, while those meeting recommendations based on energy intake were 7·7 % for children and 12 % for teenagers. Using approach 2, which focuses on the proportion of the population that achieved the recommendations, 17 % of the children and 39 % of teenagers satisfied the 48 g/d recommendation, while 20 % of children and 32 % of teenagers satisfied the recommendations based on energy intakes.
* The recommendation suggests average daily intake amounts for whole grain at energy intakes of 4184 kJ (1000 kcal) to 13 389 kJ (3200 kcal)(Reference Griffiths17), and for those consuming < 4184 kJ ( < 1000 kcal)(Reference Wearne and Day41).
Consumers v. non-consumers of whole grain
A comparison of nutrient intakes between non-consumers and consumers of whole grain across the tertiles of intake for children and teenagers are presented in Table 6. For both children and teenagers, significant differences were seen for mean daily intakes of energy (MJ) and for a range of nutrients (fibre, P, K, Mg, riboflavin and niacin) across the tertiles of intake or when comparing consumers of whole grain with non-consumers of whole grain. The percentage energy contribution derived from added sugars for both children and teenagers was significantly lower in the highest tertile of whole grain intake in comparison to non-consumers of whole grain (P< 0·05).
a,b,c Mean values with unlike superscript letters were significantly different between non-consumers and tertiles of whole grain intakes for mean intakes of whole grain food groups (adjusted for sex) (ANCOVA, Bonferroni post hoc test).
Mean values were significantly different: * P< 0·05; ** P< 0·01; *** P< 0·001; NS (P≥ 0·05).
Discussion
The present paper analysed whole grain intakes in the diets of Irish children and teenagers. Mean daily intakes were 18·5 g/d for children and 23·2 g/d for teenagers, with 90 % of children and 86 % of teenagers being consumers of whole grain. Comparable studies around the world have found similarly low results, with mean daily intakes of 9·8, 13·0 and 26·4 g/d in the USA(Reference Zanovec, O'Neil and Cho24),UK(Reference Thane, Jones and Stephen15) and Germany(Reference Alexy, Zorn and Kersting14), respectively. Consumer rates of children and teenagers in these aforementioned studies are reported to have been 55, 73 and 80 %, respectively, all lower than that observed in this Irish population. However, there is currently no universally agreed list of whole grain foods and scientific studies can apply differing criteria when identifying a whole-grain-containing food, making direct comparisons between studies difficult. Moreover, a UK study only included foods made from a minimum whole grain content of 10 %(Reference Thane, Jones and Stephen15), in comparison to the present study and that of Alexy et al. (Reference Alexy, Zorn and Kersting14) (a German study), which included all sources of whole grain. The aim of the present study was to quantify whole grain intakes, and therefore it was deemed appropriate to include all sources of whole grain. Nevertheless, intakes were well below the dietary recommendation of 48 g/d set out by the United States Department of Agriculture(16), with just a maximum of 17 % of children and 39 % of teenagers observed to satisfy this recommendation. The suitability of applying this recommendation to younger children, where portion sizes are typically lower, is questionable; therefore, intakes were also assessed using recommendations that accounted for energy intake(16, Reference Gidding, Dennison and Birch40). Using this second approach, similar numbers of children (20 %) and teenagers (32 %) satisfied the recommendations. Dietary recommendations for whole grain intake have been developed alongside growing evidence that increasing intakes of whole grain can protect against disease, specifically CVD(Reference Mellen, Walsh and Herrington42), type 2 diabetes(Reference Sun, Spiegelman and van Dam43) and certain cancers, e.g. colorectal cancer(Reference Aune, Chan and Lau44). It is likely that the mechanisms by which these benefits are likely to occur (improved inflammatory status(Reference Masters, Liese and Haffner45), insulin response(Reference Pereira, Jacobs and Pins6), vascular function(Reference Flint, Hu and Glynn46) and lipid profile(Reference Newby, Maras and Bakun2)) may also play a role in younger cohorts, or at least pave the way for a healthier adulthood. Furthermore, a recent review investigating the link between whole grain consumption and colorectal cancer concluded that a diet rich in whole grains (90 g/d) could bring about a 20 % reduced risk of developing colorectal cancer. The review also highlighted the protective effects of increasing fibre intakes, particularly cereal fibre (10 g/d), to reduce the risk by a further 10 %(Reference Aune, Chan and Lau44).
The composition of whole grains (the germ, the endosperm and the outer bran layers) makes them a rich source of vitamins, minerals, fibre and antioxidants. Even within the boundaries of the present study, we found that children and teenagers who consumed whole grain had more favourable intakes of fibre and a number of micronutrients in comparison to non-consumers of whole grain. The strong association between whole grain and fibre was clearly evident in the present study, with a stepwise increase observed across the tertiles of whole grain intake. These results are comparable to an American study, which also showed a positive relationship between whole grain and fibre consumption(Reference Zanovec, O'Neil and Cho24). Although previous research in the present study population has showed fibre intakes to be poor(Reference Deasy, Walton and Hannon47, Reference Bannon, Walton and Hannon48) and below dietary recommendations(Reference Williams, Bollella and Wynder49), the present study suggests that inclusion of whole-grain-rich foods in the daily diet presents a viable way of boosting fibre intakes as well as other micronutrients.
The major food group contributing to whole grain intake was RTEBC (52 %). Other countries have also presented similarly high results, with RTEBC contributing 56 % to whole grain intakes in British children and teenagers(Reference Thane, Jones and Stephen15), while in America, this figure is 31 %(Reference Harnack, Walters and Jacobs50). The high consumption rates of whole grain RTEBC observed in this Irish population was no surprise (72 % of children and 59 % of teenagers), as earlier research has shown over 90 % of Irish children and 81 % of teenagers to be consumers of RTEBC(Reference Hannon and Flynn51, 52). European studies examining the change in food choices over time have also reported intakes of RTEBC to be rising(Reference Alexy, Zorn and Kersting14, Reference Lang, Thane and Caroline53).
There are a number of strategies that could be followed to increase whole grain intakes in this population. For whole grain breakfast cereals, bread and rolls, the potential remains to increase the range of whole grain varieties available and/or to encourage an increased frequency of consumption of existing foods. Another option involves the food groups ‘savoury biscuits, crackers and popcorn’ and ‘sweet biscuits, cereal bars and desserts’. Typically, these foods are deemed as ‘treat foods’ due to their fat, salt and/or sugar content, yet both groups made small contributions to whole grain intake (12 % on average), and over 34 and 49 % of the present study sample were consumers of these food groups. A practical way to increase whole grain intakes in this population is to act on the finding that whole grain foods are poorly consumed outside of the home, and to work with various stakeholders to increase the availability of whole-grain-containing foods in restaurants and catering establishments. On a positive note, the overall high rate of whole grain consumers, as seen in the present study, contests previous views that inferior taste and texture may be the reason for such poor compliance(Reference Cleveland, Moshfegh and Albertson54). It may also reflect the 20-fold increase in availability of products made with whole grain in the last decade(55). Such increasing availability, coupled with growing consumer awareness, suggests a potential to further improve compliance to recommendations in future years.
There are a number of strengths and limitations associated with the present study that deserve mentioning. First, a definite strength of this analysis was the thorough data collection methodology that was followed, providing very detailed information on whole grain intakes. Previous similar studies have used 24 h recall(Reference Zanovec, O'Neil and Cho24), FFQ(Reference McKeown, Meigs and Liu1, Reference Jacobs, Meyer and Kushi56) and 3 d weighed food diaries(Reference Alexy, Zorn and Kersting14). The present study, however, used 7 d weighed (children) and semi-weighed (teenagers) food diaries, with continuous involvement of the research team (four visits per subject). In addition, over 46 % of whole grain data were collected at the brand level.
In terms of limitations, data received from manufacturers and food labels must be accepted with caution. Further investigation of whole grain pasta consumed in the present study labelled as 100 % durum whole wheat semolina was found to contain just 88 % whole grain content by alkylresorcinol analysis. A second limitation of the present study is the possibility of under-reporting, which would result in the underestimation of whole grain intakes. In addition, despite best efforts to encourage respondents to maintain usual eating practices, the burden of completing 7 d dietary records is acknowledged. Furthermore, the lack of involvement from very low socio-economic groups and those with poor or self-conscious health lifestyles must also be taken into account; these problems being not uncommon in large food-consumption surveys.
In conclusion, despite high consumer rates of whole grain, the percentage of Irish children and teenagers meeting whole grain recommendations advised by the United States Department of Agriculture(16) is poor. RTEBC were the greatest provider of whole grain, with wheat being the primary source. In the present study, whole grain intake was associated with a more nutrient-dense diet. To further improve whole grain intakes in Irish children and teenagers, a combination of effective nutritional education programmes and increased product availability and variety are required.
Acknowledgements
The present analysis was supported by Cereal Partners Worldwide. A. P. N., B. A. Mc. N. and M. J. G. received the research funding; N. F. C. D. conducted the analysis for the paper and has no conflict of interest; the analysis was supervised by B. A. Mc. N. and A. P. N.; F. T. and H. S. are employed by Cereal Partners Worldwide and helped in designing the study and observed the analysis closely over the course of the study; M. J. G. was the principal investigator of the NCFS and the NTFS. All authors reviewed the manuscript. We acknowledge the assistance of Tim E. Grant (statistician at CStar, http://www.CStar.ie) for his advice on statistical components of the study.