Eating frequency, especially breakfast skipping and snack consumption, has been reported to be associated with nutrient intake and diet quality in epidemiological and experimental studies; however, the findings were not consistent(Reference Kerver, Yang and Obayashi1–Reference Whybrow and Kirk10). For example, eating frequency is related to higher intake of carbohydrate, dietary fibre and some vitamins and minerals among adults in USA and Australia(Reference Kerver, Yang and Obayashi1,Reference Leech, Livingstone and Worsley3) . In USA adults, higher eating frequency has also been associated with better dietary quality, as indicated by the HEI-2010 score(Reference Murakami and Livingstone2). On the other hand, breakfast skipping is associated with a decrease in the daily intake of energy, dietary fibre, vitamin B1, Ca and potassium among Japanese and Koreans adults under 50 years old both in epidemiological and interventional studies(Reference Murakami, Livingstone and Fujiwara4–Reference Yoshimura, Hatamoto and Yonekura6). However, a study in the UK reported that daily energy intake of university students is not affected with or without breakfast consumption(Reference Halsey, Huber and Low7). Furthermore, studies reported inconsistent results on the association between snack consumption and intake of nutrients and food groups. For example, snack consumption was associated with a higher consumption of nutrient-rich foods, such as fruit and milk, and a lower intake of meat and added sugars in USA adults(Reference Zizza and Xu9). Snack consumption was not associated with macro- and micronutrient intake and by itself does not lead to poor diet quality among female students aged 17–26 years in UK(Reference Whybrow and Kirk10). However, snack consumption was associated with a higher intake of daily energy and confectionaries and a lower micronutrient intake among British and Finland adults(Reference Ovaskainen, Reinivuo and Tapanainen8,Reference Murakami and Livingstone11) . Inconsistent results may be that studies have focused only on the total number of eating frequency or on a specific meal (i.e. breakfast skipping or snack consumption).
The frequency of meal or snack consumption differs according to age. Among the elderly in USA, the percentage of those who skip breakfast or consume snacks is low(Reference Krok-Schoen, Jonnalagadda and Luo12). In contrast, among young adults, the breakfast skipping rate was higher than other age groups(Reference Zizza, Siega-Riz and Popkin13), and the proportion of snack consumers was also high(Reference Fayet-Moore, McConnell and Cassettari14). Energy intake from meals or snacks also differs by age. Even among young adults who consume breakfast, the energy intake from breakfast (approximately 15 % of the daily energy intake) was less than that of the elderly (> 20 %)(Reference Howarth, Huang and Roberts15). In addition, the energy intake from snacks (20 % of the daily energy intake) among young adults aged 19–29 years(Reference Zizza, Siega-Riz and Popkin13) was approximately the same as that from breakfast among the elderly(Reference Howarth, Huang and Roberts15). Breakfast skippers were also likely to consume snacks daily among Japanese adults(Reference Yokoyama, Onishi and Hosoda16). These results suggest that assessments of energy, nutrient and food group intake, especially in young adults, should consider their situation regarding breakfast and snack consumption, to identify targets for nutritional intervention. Although one study has investigated the association between specific meals, snack patterns (i.e. breakfast, dinner and twice snack consumption; breakfast, lunch, dinner and snack consumption and lunch, dinner and twice snack consumption) and overall dietary intake among American adults(Reference Kerver, Yang and Obayashi1), no study has examined the association of breakfast and snack consumption with daily nutrient or food group intake, with a particular focus on young adults, who are more likely to breakfast skippers and snack consumers(Reference Zizza, Siega-Riz and Popkin13,Reference Fayet-Moore, McConnell and Cassettari14) .
Therefore, the current study aimed to evaluate the association of breakfast and snack consumption with daily intake of nutrients and food groups, applying breakfast consumers without snack consumption as the reference group, among Japanese young adults aged 20–39 years who participated in the 2012 National Health and Nutrition Survey (NHNS).
Methods
Data source and study population
The NHNS is a nationally representative cross-sectional annual survey conducted by local public health centres under the supervision of the Ministry of Health, Labour, and Welfare. The present study was based on data from the 2012 NHNS conducted between 25 October and 7 December 2012. The reason for using the 2012 data in the present study is the larger sample size of 2012 and 2016 survey, but the survey was not conducted in several regions in 2016 due to the earthquake and Typhoon(17,18) . The details of the 2012 NHNS have been described elsewhere(17,Reference Saito, Imai and Htun19) . Briefly, the participants, who included households and their family members (aged > 1 year as of 1 November 2012) in 475 areas, were randomly selected by single-stage clusters from the general census areas in the 2010 National Census(20). The 2012 NHNS consisted of a physical examination, dietary survey and lifestyle questionnaire. A total of 12 750 out of 24 555 eligible households (51·9 %) and 32 228 people participated in the dietary survey(17). The current study included 5905 adults aged 20–39 years. We excluded lactating or pregnant women who may have changed their usual dietary habits (n 344)(Reference Takimoto, Yoshiike and Katagiri21). We also excluded participants with missing data required for analysis in the present study, such as body height and/or body weight, waist circumference, smoking status, habitual alcohol consumption and daily step counts (n 2707). Additionally, we excluded those who skipped lunch and/or dinner (n 119). Thus, the final sample consisted of 3079 Japanese adults aged 20–39 years (1420 men and 1659 women) (Fig. 1).
Flow chart of participant inclusion.
This survey was conducted according to the guidelines laid down in the Declaration of Helsinki, all participants gave informed consent to the local government based on the Health Promotion Act(22). Permission to use the 2012 NHNS data was obtained from the Ministry of Health, Labour, and Welfare, under Article 33 of the Statistics Act, and only anonymised information was available for this study. In accordance with the Ethical Guidelines of Epidemiological Research, our study was exempted from the application of these guidelines as only anonymised data were used.
Dietary assessment
Dietary intake data were collected using a one-day semi-weighed household dietary record, excluding Sundays and public holidays. Prior to completing the survey, trained fieldworkers (mainly registered dieticians) provided an outline of the survey and explained to the participants, especially the main record-keepers (members who are usually responsible for preparing meals), how to complete the dietary record. The main record-keepers in the household were instructed to weigh all foods and beverages consumed by the household members and the amount of food waste and leftovers. When the main record-keepers ate with the household members, then they recorded their names and weights on recording forms. Additionally, the main record-keepers recorded the approximate proportions of food consumed by each household member when members shared foods from the same dish to enable estimation of individual intake. If weighing was not possible because the meal was consumed away from home, the portion size consumed, or quantity of food, and details of any leftovers were estimated by the main record-keepers. When individual household members were not with the main record-keeper at the meal, the individual reported the amount of food or beverage to the main record-keeper. Participants, who were unable to report food or beverage intake on the survey day, were regarded as non-respondents.
Trained fieldworkers visited each household and checked for missing information and errors. In accordance with the survey manual of the NHNS(17), for the weights of foods and beverages that were not measured, the trained fieldworkers converted these estimates of portion sizes or quantity of foods into weights of foods. Then, the trained fieldworkers coded each food item according to the NHNS food number lists based on the 2010 Standard Tables of Food Composition in Japan(23) to calculate the intake of energy and nutrients. The trained fieldworkers inputted collected dietary intake data using software specifically developed for the NHNS; then, the data were compiled by trained investigators at the central office to create an overall dietary data set(17).
Energy and nutrients were calculated based on the 2010 standard tables of food composition in Japan. Food items were classified into nineteen groups based on standard tables(23). The percentage of daily energy intake using reported values for each macronutrient was also calculated for protein, total fat (fat), SFA and carbohydrates.
Additionally, inadequate intake of each nutrient was determined by comparing nutrient levels with the relevant dietary reference value according to the Japanese dietary reference intakes, 2020(24), using a previously reported method(Reference Okubo, Sasaki and Murakami25–Reference Kohri, Kaba and Itoh27). In the Japanese dietary reference intakes, different types of dietary reference values were established according to their purpose. The estimated average requirement is set to prevent insufficient intake of nutrients, whereas the tentative dietary goal to prevent lifestyle-related diseases is set to prevent non-communicable diseases. Nutrient intake inadequacy was defined as follows: intake level below estimated average requirement was considered inadequate using the cut-point method for the following thirteen nutrients: vitamin A (as retinol activity equivalents), vitamin B1, vitamin B2, niacin (as niacin equivalent), vitamin B6, vitamin B12, folate, vitamin C, Ca, Mg, Fe, Zn and Cu. Regarding Fe intake, we applied the value < 9·3 mg/d as recommended by the WHO (probability of inadequacy as 50 % by assuming bioavailability of Fe as 15 %)(28) for women aged 20–49 years according to the previous studies(Reference Kobayashi, Asakura and Suga26,Reference Saito, Okada and Tarui29) . For the following seven nutrients outside the range of dietary goal values was considered inadequate; the dietary goal values used were the percentage of protein energy intake in (13–20 % energy), fat (20–30 % energy), SFA (≤ 7 % energy) and carbohydrate (50–65 % energy) and the intake level for total dietary fibre (men: ≥ 21 g/d, women: ≥ 18 g/d), Na (as salt-equivalent, men: < 7·5 g/d; women: < 6·5 g/d) and potassium (men: ≥ 3000 mg/d; women: ≥ 2600 mg/d). In the present study, crude values were used instead of the density values for all nutrients, except for macro nutrients, because it is not appropriate to underestimate the influence of breakfast and snack skipping although reporting errors could not be taken into account.
The definition of meal and snack consumption was based on the participant’s identification; if any foods or beverages that contain energy and were listed in the meal (i.e., breakfast, lunch and dinner) or snack section of the self-reported dietary record, they were categorised as meals or snacks(Reference Murakami, Livingstone and Fujiwara4). Participants were classified into four groups according to with or without breakfast and snack consumption (both breakfast and snack consumption (B+S+), breakfast consumption without snacking (B+S–), breakfast skipping and snack consumption (B–S+) and breakfast skipping without snacking (B–S–)).
Other variables
Body height (to the nearest 0·1 cm) and weight (to the nearest 0·1 kg) were measured for approximately 90 % of the participants by trained field workers according to standardised procedures. For the remaining participants, height and weight were measured either by other household members at home or self-reported. BMI was calculated as weight (kg) divided by height (m) squared. Additionally, participants wore a pedometer around their waist and measured the number of steps on a single day and then recorded the number in the dietary intake questionnaire. Information on smoking status (current smoker or not (those who quit smoking or never smoked)) and alcohol drinking habits during the preceding month (yes (drinking ≥ 20 g of pure alcohol at least three times a week) or not), occupation (professional/manager, sales/service/clerical, security/transportation/labour and others including student, housekeeper and not in paid employment) and living alone (yes or no) were collected using a self-administered questionnaire.
Statistical analysis
All statistical analyses were stratified according to sex. The differences in characteristics among the four groups based on with or without breakfast and snack consumption were compared using the χ 2 test for categorical variables and ANOVA for continuous variables. Differences in daily intake of energy, nutrients and food groups based on with or without breakfast and snack consumption were assessed using the Dunnett test referenced to the B+S– group. The nutritional inadequacy of each nutrient intake was represented as the proportion of participants whose intake was below the estimated average requirement or outside the range of the dietary goal. Logistic regression analysis was used to examine the difference in the prevalence of meeting dietary reference intakes in the other groups compared to the B+S– group. The confounding factors considered were age group (20–29 and 30–39 years), occupation (professional/manager, sales/service/clerical, security/transportation/labour and others), living alone (yes or no), current smoker (yes or no), habitual alcohol drinking (yes or no) and energy intake (except for the analysis of energy itself and nutrients provided as % energy (protein, fat, saturated fatty acids and carbohydrate)). These variables have been reported as factors that affect breakfast skipping or snack consumption(Reference Murakami, Livingstone and Fujiwara4,Reference Mishra, Lee and Hayashi30,Reference Tanaka, Tsuji and Kusuhara31) . All statistical analyses were performed using SAS statistical software (version 9.4; SAS Institute Inc.). All reported P values were two-tailed. P < 0·05 was considered statistically significant in the χ 2 and ANOVA, and P < 0·001 was considered statistically significant in the Dunnett test to minimise of Type 1 error with reference to a previous study(Reference Fujiwara, Okada and Okada32,Reference Fujiwara, Okada and Okada33) . This study was limited to participants aged 20–39 years, and analyses were performed according to sex. The proportion of participants who were in the same group from the same household was approximately 2·5 % in the present study. Therefore, households were not included as the clusters, although the NHNS is a household survey.
Results
The basic characteristics of the participants stratified by breakfast and snack consumption are shown in Table 1. The proportions of breakfast skippers in men and women were 11·8 % and 6·6 %, respectively, whereas those for snack consumers were 55·3 % and 68·2 %, respectively. Men in the B–S– group (mean age: 29·1 years, sd: 6·2) and women in the B–S+ group (mean age: 27·9 years, sd: 5·7) were the youngest. The proportion of participants within the normal BMI range was approximately 70 % in both men and women. The proportion of underweight in men was the highest in the B–S– group, whereas no significant difference in terms of BMI was observed in women. There was a significant difference in terms of occupation among the four groups, in both men and women. The proportion of men working in sales, services or clerical was the highest in the B–S– group, whereas that in security, transportation or manual labour was the highest in the other three groups. In all four groups, women mostly worked in sales, services or clerical, especially those in the B–S– group. Both men and women participants living alone were more likely to be breakfast skippers, irrespective of snack consumption. The number of daily steps in the B–S– group was significantly the lowest among men but not among women. The highest proportion of current smokers and habitual alcohol drinkers among women were in the B–S+ and B–S– groups, respectively. However, waist circumference did not significantly differ based on breakfast and snack consumption in both sexes.
Characteristics of study participants stratified by breakfast and snack consumption among 1420 Japanese men and 1659 Japanese women aged 20–39 years from the NHNS 2012 data
* Means for continuous values were compared using the analysis of variance, and proportions for categorical values were compared using the χ 2 test among the four groups.
Table 2 shows the daily energy and nutrient intake based on the dietary record stratified by breakfast and snack consumption. Men who consumed both breakfast and snacks (B+S+) had higher intake of energy, vitamin B2, Ca, Mg and potassium but lower intakes of protein (as % energy) and Zn than men who consumed breakfast but not snacks (B+S–). On the other hand, for only energy intake, the breakfast and snack skippers group (B–S–) had lower energy intake than the B+S– group. Women who consumed breakfast and snacks (B+S+) had higher intakes of energy, folate, Mg, dietary fibre and potassium but lower intake of protein (as % energy) than women who are non-snack consumers (B+S–). While women in the B–S– group had higher saturated fatty acid intake (as % energy) and lower energy, Cu, dietary fibre and potassium intake than those in B+S- group. No difference was observed in the intake between breakfast skippers with snack consumption (B–S+) and those in the B+S– group among both men and women.
Daily energy and nutrient intake of study participants stratified by breakfast and snack consumption among 1420 Japanese men and 1659 Japanese women aged 20–39 years from the NHNS 2012 data
(mean values and standard eror)
NE, niacin equivalent; RAE, retinol activity equivalent; SFA, saturated fatty acid.
* Indicates significant difference referenced to the B+S– groups by Dunnett test in the adjusted model with confounding variables of age, occupation (professional/manager, sales/service/clerical, security/transportation/manual labour and others), living alone (yes or no), current smoker (yes or no), habitual alcohol drinker (yes or no) and energy intake (except for the analysis on energy itself and nutrients provided as % energy (protein, fat, saturated fatty acid and carbohydrate)).
† Sum of retinol, β-carotene/12, α-carotene/24 and cryptoxanthin/24.
‡ Sum of niacin and protein/6000.
The multivariate-adjusted OR for nutritional inadequacy based on with or without breakfast and snack consumption are shown in Table 3. Among men who consumed breakfast, those who also consumed snacks (B+S+) were less likely to have inadequate intake for vitamin B2, vitamin C, Ca, Mg and potassium but a higher prevalence of inadequate intake of Zn and saturated fatty acid than non-snack consumers (B+S–). On the other hand, only the prevalence of inadequate folate intake was higher in the B–S– group than in the B+S– group. Among women, breakfast and snack consumers (B+S+) were more likely to have a lower prevalence of inadequate intake of vitamins A, B2, B6 and C, folate, Mg, dietary fibre and potassium than the B+S– group. A lower prevalence of inadequate protein intake and a higher prevalence of inadequate niacin intake were observed in the B–S+ group than in the B+S– group. The B–S– group was more likely to have a higher prevalence of inadequate folate and Cu intake than the B+S– group.
Multivariate-adjusted odds ratios for the presence of not meeting estimated average requirement (EAR) and tentative dietary goal to prevent lifestyle-related diseases (DG) among 1420 Japanese men and 1659 Japanese women aged 20–39 years from the NHNS 2012 data
(Odd ratio (OR) and 95 % confidence intervals (CI))
DRI, dietary reference intakes; SFA, saturated fatty acid.
Percentage of subjects whose intake was in the range of DG or above the EAR.
Each energy-adjusted nutrient intake (unit/d) was compared with each DRI value (unit/d)(22), using the cut-point method.
* Adjusted for confounding variables of age category (20–29 and 30–39 years), occupation (professional/manager, sales/service/clerical, security/transportation/manual labour and others), living alone (yes or no), current smoker (yes or no), habitual alcohol drinker (yes or no) and energy intake (except for the analysis on energy itself and nutrients provided as %energy (protein, fat, saturated fat and carbohydrate)).
† The probability of inadequacy > 50 % for menstruating women whose bioavailability of iron is 15 % (< 9·3 mg/d) was considered inadequate for women aged 20–49 years.
Food group intake based on breakfast and snack consumption is shown in Table 4. Men in the B+S+ group had higher intakes of grain, fruits, dairy products, confectionaries, tea, coffee, cocoa and soft drinks than men in the B+S− group. While women who consumed breakfast and snacks (B+S+) had higher intakes of fruits, confectionaries, tea, coffee, cocoa and soft drinks than those consume breakfast but not snacks (B+S–). Participants in the B–S+ group had higher intakes of confectionaries than those in the B+S– group among both men and women. Additionally, food group intake did not significantly differ between the B+S– and B–S– groups in both men and women.
Daily food group intake of study participants stratified by breakfast and snack consumption among 1420 Japanese men and 1659 Japanese women aged 20–39 years from the NHNS 2012 data
(Numbers; mean values and standard error)
* Indicates significant difference referenced to the B+S− groups by Dunnett test in the adjusted model with confounding variables of age, occupation (professional/manager, sales/service/clerical, security/transportation/manual labour and others), living alone (yes or no), current smoker (yes or no), habitual alcohol drinker (yes or no) and energy intake (except for the analysis on energy itself and nutrients provided as % energy (protein, fat, saturated fatty acid and carbohydrate)).
Discussion
To our knowledge, this is the first study to investigate daily nutrient intake by focusing on breakfast and snack consumption among young Japanese adults. We found that breakfast and snack consumers (B+S+) had higher levels of several vitamins and minerals than those who consumed only three meals a day (B+S–). Furthermore, the nutrient intake level of breakfast consumers without snack consumption (B+S–) was similar to that of breakfast skippers who consumed snacks (B–S+).
The proportion of breakfast skippers in the present study (men; 11·8 % and women; 6·6 %) was similar to that of a previous Japanese study (men; 10·6 % and women; 6·4 %)(Reference Sakurai, Yoshita and Nakamura34). Additionally, the proportions of snack consumers in this study (men; 55·3 % and women; 68·2 %) were lower than that of previous studies for USA adults (approximately 80–90 %)(Reference Kerver, Yang and Obayashi1,Reference Zizza and Xu9) . This difference may be attributed to the higher proportion of lunch (approximately 25 %) and dinner (10 %) skippers in the USA(Reference Kerver, Yang and Obayashi1) than Japanese in this study (5 %). The characteristics of breakfast skippers in this study were consistent with previous studies which reported that breakfast skippers were younger, living alone and smokers(Reference Mishra, Lee and Hayashi30,Reference Sakurai, Yoshita and Nakamura34–Reference Nakamoto, Tanaka and Ono37) .
In the current study, breakfast and snack consumption was positively associated with energy intake similar to a previous study(Reference Leech, Livingstone and Worsley3). The higher energy intake in B+S+ groups could be explained by energy intake from confectionaries and soft drinks because snack consumers had more intakes of these food groups than non-snack consumers as previous studies, irrespective of dietary survey methods such as 24-h dietary recall and dietary record(Reference Murakami and Livingstone11,Reference Zizza, Siega-Riz and Popkin13,Reference Hartmann, Siegrist and van der Horst38,Reference Murakami, Shinozaki and Livingstone39) .
In contrast, the highest prevalence of underweight was observed in men but not in women in the B–S– group as well as a previous study which reported that eating frequency was positively associated with BMI only in USA men(Reference Murakami and Livingstone2). Underreporting is more common among young Japanese women(Reference Murakami, Livingstone and Okubo40), especially those who are overweight(Reference Murakami, Sasaki and Okubo41). In fact, 20 % of the women in the B–S– group were overweight, although their energy intake was the lowest on the dietary record day. In addition, differences in social desirability by gender, such as underreporting of energy intake among women, have been reported(Reference Hebert, Ma and Clemow42). Thus, caution should be observed in interpreting our results on women. The lower energy intake in the B–S– group suggests that the lower energy intake due to missing breakfast and snacks was not compensated for by lunch and dinner. However, because the results of this study were obtained from a one-day dietary record, similar trends observed in studies of habitual food intake need to be assessed.
The intake of folate, vitamin C, Ca, Fe, potassium and dietary fibre among breakfast and snack consumers were higher than those consuming three meals a day in USA adults(Reference Kerver, Yang and Obayashi1). Additionally, for young USA adults, snacks have been reported to be contributed to the intake of vitamin B2, vitamin C, niacin, Ca and Fe(Reference Khan and Lipke43), whereas breakfast skipping has been reported to be associated with inadequate intake of vitamin A, vitamin C and Mg(Reference Deshmukh-Taskar, Radcliffe and Liu44). These results from 24-h recall data are consistent with our results from dietary record data. Moreover, regarding the intake of food groups, the B+S+ group had higher intake of fruits than the B+S- group. Fruits contain a large amount of vitamins and minerals(45), and their consumption was higher among snack consumers(Reference Zizza and Xu9). In addition, women with a higher frequency of snack consumption were more likely to be health conscious(Reference Hartmann, Siegrist and van der Horst38), and the frequency of snack consumption in women was higher than that in men(Reference Murakami and Livingstone46,Reference Murakami, Livingstone and Masayasu47) . These reports may partly explain why the number of nutrients meeting adequate intake was higher among women than among men in this study. Moreover, among only men, Zn intake was lower and inadequate in the B+S+ group than the B+S– group. Zn content varies from food to food, irrespective of food group(23). The secondary data for the NHNS did not have access to detailed food records. Detailed food sources of Zn in breakfast and snacks may need to be explored. Overall, snack consumption may lead to higher nutritional intake.
On the other hand, the lower intake of Cu, dietary fibre and potassium was observed among women in the B–S– group than those in the B+S– group. Among USA and Korean adults, breakfast skippers had lower intake of vitamin B1, niacin, folate, vitamin C and Ca than breakfast consumers, irrespective of dietary survey methods such as 24-h dietary recall and dietary record(Reference Min, Noh and Kang5,Reference Zhang, Cordeiro and Liu48) . In addition, among Japanese adults, breakfast skippers had lower intake of folate, vitamin C, Ca, Mg, potassium and Fe than breakfast consumers(Reference Murakami, Livingstone and Fujiwara4). However, breakfast skippers in these previous studies included both snack and non-snack consumers, unlike the classification in this study, which could explain the discrepancy in results. In young adults who skip breakfast and snacks, especially women, need to be considered in the future.
No difference was found in nutrient intake between the B+S– and B–S+ groups. In contrast to the results of this study, American adults aged ≥ 20 years who skipped breakfast and consumed at least two snacks per day had lower intake of nutrients except for salt than those who consumed only three meals per day(Reference Kerver, Yang and Obayashi1). Although the frequency of snack consumption per day was not examined in this study, the B–S+ group included those who snack at least once a day, whereas the previous study did not include those who consumed snacks only once(Reference Kerver, Yang and Obayashi1). This may partly explain the difference in the results. Thus, whether snack content changes according to the timing and frequency of snacks should be considered in future studies.
This study had some limitations. First, the participants were randomly selected from nationally representative households in Japan; however, only 51·9 % of the sampled households participated, and the individual-level response rate was unknown. This might have introduced some bias in the estimation of average intake in Japanese adults. Second, a dietary intake derived from a one-day weighed dietary record is unlikely to represent the usual intake. Therefore, caution is needed for interpreting our results, especially those regarding nutrient inadequacy. Incidentally, it is noted that the one-day household-based dietary record method used in NHNS has been compared with individual dietary records of Japanese participants, and the correlation coefficients of the intake of total energy and macronutrients were high (0·89–0·91)(Reference Iwaoka, Yoshiike and Date49). Thus, this method may be valid for estimating individual intake. Third, we only analysed those who consumed three meals a day because this study aimed to assess nutrient intake and nutrition adequacy based on the difference in breakfast and snack consumption. This might have introduced some bias regarding nutrient intake. Fourth, meal frequency and snack frequency have been linked to education and income(Reference Murakami and Livingstone2). However, we did not investigate the education and income levels of the participants; thus, we could not consider these relationships. Therefore, their influence on the results could not be determined. Fifth, we did not examine the timing and frequency of snack consumption. Snack timing and frequency have been positively associated with the intake of fruit and dairy products and negatively associated with the fish intake(Reference Murakami and Livingstone2,Reference Murakami, Shinozaki and Livingstone50) . Sixth, a large number of participants (n 1373) were excluded from the present analysis because of missing information on their height and weight. The proportion of individuals with missing value was higher among breakfast skippers; however, there were no differences in macronutrient intakes between excluded persons and participants in this study (data not shown). Therefore, the exclusion of people with missing values from the analysis would not have a significant impact on the results of this study. Lastly, the number of people varied in each group, and the number of participants was small, especially the women in the B–S– groups, although the required sample size was B–S– (n 34) and B+S– (n 458) that was estimated using G × Power 3 with effect size 0·5, power 0·8 and allocation ratio n2/n1 12·5(Reference Faul, Erdfelder and Lang51). Further study may need to be conducted in target population with a large sample size where the statistical power of the B–S–group is sufficient.
This cross-sectional study showed that breakfast and snack consumption is related to nutrient intake among Japanese adults. In young Japanese adults whose energy intake is within an appropriate range, snack consumption could supplement nutrients that may not be adequately consumed by three meals. Overall, Japanese adults are likely to consume nutrient-rich snacks, and further investigation on the nutrients and consumption of snacks is warranted.
Acknowledgements
Financial support: This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors. Authorship: All the authors designed the research. M. M. analysed the data and wrote the first draft of the manuscript. E. O., R. T. and A. F. took part in the interpretation of the data and provided critical revisions of the manuscript for important intellectual content. H. T. had the primary responsibility for the final content. All authors read and approved the final manuscript.
Conflict of interest:
There are no conflicts of interest.
