Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

From milk to solids: a reference standard for the transitional eating process in infants and preschool children in Japan


Objective: This paper aims to establish a potential reference standard for the process of transition from milk to solid food in infants and preschool children in Japan, using the transitional food process (TFP) scale described by Sakashita et al. The background for variation and delay in the process are also discussed.

Design: A randomized sample survey covering entire Japan.

Setting: Mailing self-completion of questionnaires.

Subjects: Randomized sample of 14 000 children aged 0–6 y and their family from 13 prefectures in Japan, namely Hokkaido, Aomori, Iwate, Niigata, Tokyo, Saitama, Fukui, Nagano, Nagoya, Hyogo, Yamaguchi, Kagoshima, and Okinawa.

Methodology: Questionnaires requesting the TFP scale and background factors were sent to 14 000 children and families. The percentile ages were calculated. An eating ability index (EAI: number of accepted foods/total number of foods) × 100) was calculated. Regression analysis by analysis of covariance (ANCOVA; SPSS, 1997) was used to determine the influence of background factors on EAI.

Results: From the 6747 (48.2%) effective answers received, percentile curves of the acceptability of each food on the scale were drawn, and used as initial reference standards. The 50 percentile age range of these 20 standard foods covers from 5 to 42 months after birth. The sensitive period for increasing the acceptance of foods in children was between 6 months and 2½y of age. ANCOVA regression model (R2=0.605) showed that EAI was mostly influenced by age (P=0.000), followed by feeding style (P=0.000), infant food preparation (P=0.000), information source (P=0.000), and birth order (P=0.003). The dominant cause of digestive system problems was shown to be infection, not too-hard food. It seems that breast feeding, bottle feeding with chewing-type nipples, and the manner of preparing infant foods from the family table promote the progress in acceptance. Children whose mothers followed the information given in books or magazines showed a slower progress.

Conclusions: It seems appropriate to use this reference standard in the study of the transitional process from milk to solid food in infants and preschool children in Japan.

Sponsorship: Grant-in-Aid for Scientific Research, provided by the Japanese Ministry of Education, Culture, Sports, Science and Technology Project No. 07838030.


Many children are known to have difficulties in eating in Japan (Maternal and Child Health Division, 1986; Yokomizo, 1992). Yokomizo (1992) found that at 2 y of age, 37.5% of preschool children swallowed without chewing, or were unable to chew and swallow some foods. Kindergarten nurses reported that 34.6% of children aged 3–6 y did not chew properly, or ate slowly, retaining food in the side oral cavity, or frequently spitting out food. It has been suggested that such problems are due to underdevelopment of the masticatory system and immaturity of the chewing and swallowing ability caused by soft diet (Inoue, 1993; Sakashita et al, 1998). The underdevelopment of the masticatory system seems to be on the increase among young Japanese (Inoue, 1993), for example, 63.1% (N=495) of young Japanese people have crowding of the teeth (Sakashita et al, 1997), while more than 33.5% (N=391) of high school students have a juvenile temporomandibular joint problem (Inoue, 1993). With respect to the causes of these problems, the present diet, with its too-soft and too-nutritious food, bottle feeding and long-lasting pureed baby food have all been suspected (Inoue et al, 1995). Experiments have shown teat-fed mice to have a better developed masticatory system than bottle-fed ones (Ito et al, 1982). A diet of liquidized or pureed food reduced the height of the mandible, the length of the body of the mandible and the condylar width, and widened the gonial angle (Ito et al, 1988). It also reduced the size of the masseter and temporal muscles as well as that of the salivary glands (Ikeda, 1998). Experiments with rats have also shown that a liquid diet after weaning affects the critical-period imprinting of both mastication and the motor performances of the jaw and tongue muscle (Liu et al, 1998).

Before the causes of eating problems can be discussed, it is necessary to clarify the actual condition and variance of the progress in food types among children. For this reason, a scale for use in assessing the transitional process from milk to solid food in infants and preschool children, the transitional food process (TFP) scale, was developed (Sakashita et al, 2001,2003).

To introduce the proper food at the proper time is surely important not only for the masticatory system but also for the lifelong health of the whole body. Experience in the early stages of life is likely to influence diet in later life (Westenhoefer, 2001). Our previous studies suggest that intake of fiber-rich food, such as various kinds of vegetables, would promote masticatory development and reduce dental disease (Sakashita et al, 1996,1997). Eating more fiber-containing foods like vegetables, fruit, and legumes and less animal products could help decrease metabolic risk factors involved in obesity and obesity-related disease (Rolland-Cachera et al, 1997; James et al, 1997; Pereira & Ludwig, 2001). Figures provided by the Maternal and Child Health Division, the Ministry of Health and Welfare Japan (1995) show that 81.4% of mothers with children under 6 y of age were reported to worry about children's diet, such as unbalanced intake, playing while eating, slow eating, and so on. This reference standard can help mothers by suggesting what and when food should be given to children. The eating ability reference would also help child-care providers or pediatricians to identify individual children who fall outside the normal distribution of eating ability, and then offer advice on which foods to introduce and when. Furthermore, it can be used in research or in public health surveillance to identify populations of children whose distribution of food acceptability differs substantially from that of the reference standard.

In the present study, we attempted to establish a potential reference standard by first documenting variation in progress in the transition to solid food. We also checked on problems with the digestive system during the transitional process, and analyzed background history as a variance factor. The latter analysis was conducted by considering the inter-relationship between food progress and background, including feeding style history, manner of preparing the infant food, and information source for the infant food used.

Subjects and methods


To avoid any regional bias, the subjects for the present study were drawn from different regions across Japan, namely Hokkaido, Aomori, Iwate, Niigata, Tokyo, Saitama, Fukui, Nagano, Nagoya, Hyogo, Yamaguchi, Kagoshima, and Okinawa (Figure 1). A total of 14 000 children aged between 2 and 71 months and their families were randomly selected from health center lists according to the population size in each district. The results of a previous study, in which children aged from 2 to 46 months were studied (Sakashita et al, 2001), had suggested the need for a wider age range, since some children even older than 36 months could not eat several foods. Therefore, in this study, the upper subject age level was set at 6 y.

Figure 1

Location of samples prefectures.

The questionnaire

In the previous study (Sakashita et al, 2001), 159 items from among more than 600 kinds of daily food were selected for further evaluation as possible candidates for standard food. They were selected using the following criteria: daily use, geographical commonality, ready availability, and mild taste. In the second study (Sakashita et al, 2003), 20 reference foods were chosen from the 159, based on the period when 50% of the subjects could eat them, and on the food groups classified by cluster analyses using the Varclus procedure of SAS (Hartman & Sarle, 1988). The validity of the scale was tested by Pearson's correlation coefficient between the scale score of selected food items and the total score of all 159 food items. The total score for the 159 items was estimated using selected model food items in the general linear models (GLM) procedure of SAS statical package. Reliability was tested using Cronbach's coefficient.

Questionnaires were sent to the families, who were asked to return them by mail. They were asked to record their child's present ability to eat each of 20 food items listed in the questionnaire. The child's developmental stage for eating each food was categorized according to the method used in our previous findings (Sakashita et al, 2001), but adding the category ‘often spat out’, following advice received from several mothers and public health nurses during the previous investigation. The categories used were thus:

  1. 1

    Not yet experienced

  2. 2

    Offered but unable to eat

  3. 3

    Often spat out

  4. 4

    Chewed but not swallowed

  5. 5

    Able to eat

Food that was decided not to be given to children because of allergies or certain diseases, or which the family as a whole did not eat, were also recorded to allow them to be excluded so as not to create a bias in the standard process of food intake.

Any history of problems with the digestive system was noted, together with suspected causes.

As a background to the child's progress, we asked for birth order (ie first-born, second-born, third-born, etc), a history of the feeding style, how the infant's food was prepared, and which information sources on infant food, if any, had been consulted.

The history of breast- or bottle-feeding included information about the main types of artificial nipples used. Breast fed here refers to breast feeding only, and mixed feed to a mixture of breast fed and bottle fed, whether complementary foods were introduced or not. Exclusively bottle-fed subjects, and those given mixed feed, but where breast feeding stopped within 3 months after birth, were categorized as ‘mainly bottle fed.’ Four main types of artificial nipple were found to have been used, namely, nipple with a round hole at the top (Pigeon Co., Japan), nipple with a crosscut at the top (Jex Co., Japan), NUK nipple (originally from Germany, Mituwa Co., Japan), and the chewing-type nipple with a pair of semilunar valves inside the nipple (Beanstalk Nipple, Otsuka Pharmaceutical Co., Japan). The nipple used for the longest period was considered the main nipple.

The preparation of infant food was categorized as follows:

  1. 1

    Specially cooked for the baby

  2. 2

    Canned baby food bought commercially

  3. 3

    Food from the family table

The main sources of information on infant food were categorized as follows:

  1. 1

    Book or magazine

  2. 2

    Medical doctor or public health nurse

  3. 3

    Mother, sisters, or friends

  4. 4

    Own experience only

The questions on food preparation and information sources used were added after the active investigation had begun. These questions were therefore not asked in 31.7% (4440/14000) of the total sample.


The percentile ages when 25, 50, and 75% of the children could eat a given food were calculated from the percentage curves of those who could eat a given food (category 5 in the questionnaire).

An eating ability index (EAI) was defined as a means of summarizing the total eating ability and calculated as (Number of accepted foods/Total number of foods provided daily at the family table) × 100. Foods not given to the children because of allergies or certain diseases, or which the family would not normally eat, were excluded. Regression analysis and analysis of variance by analysis of covariance (ANCOVA), with estimated marginal means and comparing main effects options (using SPSS v. 7.5, 1997), were employed to determine the influence of background factors on EAI. ANCOVA provides regression analysis and analysis of variance for one dependent variable through combinations of categorical and continuous predictor variables. EAI was predicted using categorical predictors (birth order, digestive problems, feeding style, feeding style at 1 week of age, infant food preparation, and information source), and continuous predictors (month of age). The effect of each categorical level with a significant impact on EAI was analyzed by ANCOVA comparisons of estimated marginal means for each level, adjusted for age.


Subjects (Table 1)

Table 1 Population make-up

The questionnaires were returned by 6840 mothers (48. 9%) out of 14 000. Some data were considered inappropriate due to low birth weight (under 1500 g), or a history of serious disease, and this was removed from the data set so as not to create a bias in the standard process of food intake. The number of effective returns was 6747 (48.2%), of which 4512 (66.9 % of returns) contained replies to additional questions about food preparation methods and the information source consulted. For the purposes of the ANCOVA analysis, responses listing unspecified nipples (28 cases) and/or imprecise information sources (e.g. ‘TV’, 14 cases) were excluded and a total of 4470 responses were used.

Table 1 gives data on subjects by age group and sex. As there was no significant sex difference, the information was pooled. A total of 3374 (50.0%) were first-born children and the rest second or later children.

The type of household and parents’ working condition of subjects’ parents are shown in Table 2, together with data from a national survey.

Table 2 Type of household and parental employment

Age percentiles

The distribution of eating conditions for 20 reference food, given according to age, is shown in Figure 2. The 25, 50, and 75 age percentiles are shown in Figures 2 and 3. More than half of the infants had been introduced to rice gruel before 6 months of age. At 6 months, most of all had been offered this food, with 83.3% (204/245) being able to eat it. Rice gruel, noodles, and boiled rice were readily accepted by the children, as indicated by the relatively small range between 25 and 75 percentiles and the low incidence of children who were offered, but did not accept these foods. Boiled minced meat, sliced apple, sliced bread, and sliced ham were also consumed with a minimum of trouble. It appeared that some children could not easily accept cucumber sticks, leaf vegetables such as 3–4 cm spinach or boiled burdock, boiled Japanese leek, 3–4 cm fried cabbage, shredded raw cabbage, sliced meat, or beef steak. There is a wide age range between the 25, 50, and 75 age percentiles for these food items.

Figure 2

Distribution of eating condition according to age.

Figure 3

A potential reference standard with 25, 50, and 75 age percentiles.


EAI percentiles of 25, 50, and 75 are shown in Figure 4. From 6 to 12 months after birth, the number of accepted foods rapidly increased; after 3 y, the progress slowed down.

Figure 4

The 10, 25, 50, 75, and 90 percentiles of EAI.

Problems with the digestive system

The number of subjects who experienced trouble with the digestive system (eg vomiting, diarrhea, or anorexia) was 18.3% (1233/6747) of the total number of subjects and 25.5% (1074/4212) of children aged 1–3 y. The troubles increased gradually from 6 months; the average age for such episodes was 8 months and the problems decreased from the age of 1 y. In all, 66.2% of the respondents (816/1233) assumed the difficulties to be due to infections, for example, diarrheal disease, the common cold, or influenza. In 15.4% of the cases (190/1233), food was indicated as the cause. The most frequently suspected foods were milk and other dairy products with 18.4% of the cases (35/190), followed by fruit with 11.0% (21/190).

Background to the variance

History of feeding style

The data for age 4 months and above showed that, at 4 months of age, 34.0% (2269/6674) were breast fed, 37.6% (2509/6674) were bottle fed, and the rest were mixed. Figures for age 14 months and above showed that, after 13 months, 64.6% (602/932) of children who had been breast fed were no longer fed in this manner, with 72.1% (338/469) being the figure for those who had been bottle fed. Most of the mothers stopped breast- or bottle-feeding before their babies were 18 months old. At 6 months, 87.9% (58/66) of breast-fed infants were introduced to rice gruel, while the figure for bottle-fed ones was 89.5% (94/105). Bottle-fed children tended to be introduced to Japanese noodles earlier than the breast-fed ones, whereas in the cases of spinach and minced meat, no significant difference was found.

ANCOVA results

As shown in Table 3, the adjusted model gave a good prediction of EAI (adjusted R2=0.604, P=0.000). Age, feeding style, infant food preparation, information source, and birth order had a significant effect on EAI, while digestive problems and feeding style for 1 week after birth were not significantly effective. Type III sum of squares shows the contribution for prediction, with age being the most effective variable, followed by feeding style, infant food preparation, etc.

Table 3 ANCOVA results for predicting EAI

The effects of categorical levels were compared by ANCOVA, using EAI estimated marginal means, adjusted for age (Table 4). Breast fed and mainly bottle fed with the chewing-type nipple marked a higher EAI, while mainly bottle fed with circular hole nipple and crosscut ones are lower (F=36.959, P=0.000). Children who were offered food prepared from the family table marked the highest EAI, while those offered canned baby food marked the lowest (F=17.627, P=0.000). Information was mostly obtained from books or magazines, and seldom from doctors or public health nurses. Information found in books or magazines was a negative factor on EAI, while advice from relatives or friends, or own experience was a positive factor (F=11.980, P=0.000). First-born children tended to have a lower EAI than second or later children (F=19.336, P=0.000).

Table 4 Effects of categorical level on EAI using estimated marginal meansa in ANCOVA


Evaluation of subjects and reference foods

In this study, socioeconomic class was not specifically addressed. Compared to the national survey (Statistics and Information Department, 2001) on types of household and parental employment, subjects were thought to be representative of Japanese households with child(ren) under 6 y of age (Statistics and Information Department, 2001). In the present study, a postal questionnaire was used to collect a large quantity of data, covering a wide age range in children in an economical manner. The response rate in Japan to postal questionnaires rarely exceeds 50% (Marui, 1982), and this study reflected this trend, with a return of less than 50%. Although there was no clearly observable bias in the present sample, it would be desirable to control the results through further study.

The selected 20 food items are very popular foods in Japan; they do not have a particularly strong taste, nor are they seasonal, and they are easily obtained. This is evident from the fact that all of the food items, except whole apple, were introduced to most of the children by the age of 3 y. The 50 percentile ages in relation to the 20 food items cover a wide spectrum, ranging from 5 months (rice gruel) to 42 months (boiled Japanese leek). Whole apple most frequently produced a negative response. Apple is a very popular fruit in Japan and in our previous study a negative response was rare (Sakashita et al, 2001). This difference may be due to the market tendency to concentrate on relatively flawless, large (about 10 cm diameter), and expensive apples, which tends to drive smaller apples off the market. It may also be linked to mothers being especially careful to cut the apples into easy, bite-sized pieces. Thus, although we could not pinpoint the reason for the observed difference, a possible solution would be to remove whole apple as a selected food.

As there is a wide intercultural variation in the transition from milk to solid food, this reference standard is culturally specific to Japan, not only in the types of foods used in the index but also in what is considered culturally acceptable timing. This particular standard should not be used in a study in a different country. However, it would be possible to create the same kind of scale using the same method.

A further point of concern is the reliability of the method. Since the present study was carried out using a questionnaire, the reliability of the information may have been limited and may not always have reflected the true dietary condition of children. However, there seems little scope for significant error when reporting the present condition of an infant's ability, or nonability, to eat a particular popular food. For recall data, several authors believe maternal recall of infant feeding events to be accurate (Persson & Carlgren, 1984; Launer et al, 1992), while Quandt (1987) found it to be inaccurate. The amount of recall data relied on in the survey would therefore have to be limited.

Overview of transitional process

The results obtained in the present study showed that the number of accepted foods increased most rapidly between 6 months and 1 y of age, and continued to increase throughout the first 2 y.

One of the first solid foods introduced to an infant is rice gruel (Imamura, 1987). More than half of the infants in this study had this food introduced below 6 months of age, which is the same figure as reported in a national survey (Maternal and Child Health Division, 2001). In that survey, it was reported that infants started to accept pulp-like gruel at average 5.1±1.5 months and stopped depending on milk for their main calorie intake at average 11.9±1.5 months. In the present study, most of the food items were accepted by around 2.5 y. After that, the rate of progress slowed down, which may indicate this period to be particularly sensitive in terms of food acceptance.

Skinner et al (1997) reported that dislike of vegetables in the second year of life was an area of concern. Food such as leafy vegetables and sliced meat were not accepted easily in this study. Such foods share a richness in fiber, and must be sufficiently chewed before they can be swallowed. although vegetable foods are not particularly hard in texture (Brandt et al, 1963; Sherman, 1969), their rough natural fiber requires thorough mastication before swallowing. This explanation seems to be supported by the finding, in this study, that sliced processed meat (ham) and minced meat were readily accepted when introduced, but ordinary sliced meat was not.

Problems with the digestive system

The transitional process is dependent not only on the development of the masticatory system but also on the infant's ability to digest and absorb the food (Lebenthal, 1985; Milla, 1986). A WHO publication (2001) recommends breast feeding alone for 6 months, with the introduction of complementary food and continued breast feeding thereafter. In developing countries, the most important potential advantage of exclusive breast feeding for 6 months, as opposed to exclusive breast feeding for 4 months followed by partial breast feeding till the age of 6 months, lies in combating infectious disease morbidity and mortality (WHO, 2001). In Japan, however, where the risk of diarrheal disease from complementary food is thought to be very low, the introduction of pureed food at 5 months has been recommended (Maternal and Child Health Division, 1995), and widely accepted (Maternal and Child Health Division, 2001). It has also been argued that the texture of hard food would cause problems with the digestive system (Imamura, 1987). However, the digestive ability increases after 6 months (Grand et al, 1976), and by the eighth or ninth month after birth, the maturing gastroenteric and renal systems enable the infant to digest a variety of foods (Bronner & Paige, 1992).

In the present study, milk and dairy products were suspected of causing gastric trouble. We found little evidence that hard food texture itself led to any problems with the digestive system. Common sense would dictate that an infant should not be forced to eat food before 6 months, but thereafter there is no need to hesitate in introducing food that requires chewing as and when the baby wants, as long as the child is watched for any urgent case of choking and their health condition.

Background to the various progress trends

If one of the main reasons why a child cannot accept a certain food is insufficient development of masticatory ability, then the amount of exercise given to the masticatory system should be discussed. In the present study, breast-fed children marked a higher EAI than bottle-fed ones, except for those using chewing-type nipples. Experiments have shown breast-fed mice to have a better developed masticatory system than bottle-fed mice (Ito et al, 1982), although in the case of humans, both confirmatory (Barrett & Hanson, 1978; Stanley & Lundeen, 1980; Randolph & Dennison, 1981) and negative reports (Simpson & Cheung, 1976) exist. The masticatory system of breast-fed babies has plenty of opportunity to develop well, because the masseter muscle works more actively at breast feeding than at bottle feeding (Inoue et al, 1995).

Acceptance of food is also governed by the early flavor experiences provided through mother’s milk. Sullivan and Birch (1994) reported that infants fed breast milk showed greater increases in acceptance of vegetables than formula-fed infants, and suggested the reason to be the variety of flavor in mother’s milk. According to Mennella (1995), breast feeding generally seems to facilitate acceptance of different foods, with the greater variation in breast milk flavor compared with formula deemed to be the reason.

Among the artificial nipples, only the chewing-type nipple seemed to promote feeding progress, while those with a round hole seemed to hinder it. The chewing-type nipple is equipped with two semilunar valves inside the teat. These valves are designed so as to be closed when the baby takes the nipple between its jaws, causing the milk to be ejected by jaw movements similar to those used when feeding from the mother's breast. The jaw movements in a baby feeding from an ordinary, round-hole nipple have been found to be quite different. The masticatory movement is very weak, and negative pressure would play a greater role in sucking milk (Sakashita et al, 1996).

The manner of preparation of the infant food appears to influence the progress to solid food. Preparations from the family table food seem to promote eating progress. Such preparations would give the infant plenty of opportunity to experience a variety of food textures at an early stage. The timing of food introduction also seems important (Northstone et al, 2001). Infants who were introduced to lumpy, solid foods before 6 months of age consumed a greater variety of family foods. Infants who had such foods introduced at 10 months or later were significantly less likely to accept family foods and more likely to have feeding difficulties, when compared to those who were first given them between 6 and 9 months (Northstone et al, 2001).

With regard to the information sources used, it was ironical that children whose mothers followed the information in books or magazines showed slower eating progress, while children whose mothers followed advice from relatives or friends, or purely their own experience, progressed faster. The latter will have had access to a large amount of personal experience in the preparation of weaning food suitable for their own children's development, which likely enhanced that development. Most magazines and books follow the guidelines set by the Maternal and Child Health Division, Ministry of Health and Welfare, Japan (1995). This guideline recommends the introduction of pureed, gruel-type food once a day at around 5 months, the purpose being that of making the infant accustomed to food taste rather than supplying nutrition. At about 6 months, giving pureed food twice a day is recommended. At about 7 months, soft food that can be pressed down by the infant's tongue should be introduced, and at 9 months, food that can be compressed by the gums should be given three times a day. The aim is to arrive, at about 9 months, at the ability to consume food that can be crushed by the gums three times a day. When such food becomes children's main source of nutrition, the end of the weaning process would reached between the ages of 12 and 15 months (average age: 13 months).

The Japanese publication (Mutou, 2003) recommends introducing solids earlier than the WHO recommendation of 2001, but recommended keeping the food texture the same for a long time, and there is scant information on when to introduce adult-type food. As seen in the present study, many children do not conform to the guideline. For instance, 25% of children were able to eat boiled potato cubes at 8 months, while many remained unable to eat this until after 13 months and could not accept 3 cm pieces of leaf vegetable until 2 y of age. A child develops its eating behavior dynamically, day by day, based on trial and error. In traditional Japan, the family has a meal together, when the infant can watch others’ eating behavior and has the opportunity to try out table food. People would allow an infant to eat rubbery or hard foods such as dried squid, dried sweet potato, or salt-reduced Takuan (dried radish pickle), and let them enjoy chewing them (Mizuno et al, 2000). These types of food have now mostly been replaced by plastic teething rings or hard rubber toys. These may bring some relief from the discomfort of teething (Betz et al, 1994), at the same time promoting the development of the masticatory system and helping the infant to become ready for chewing and swallowing.

The development of the masticatory system is promoted by the stimulation of mechanical function (Watt & Williams, 1951; Beecher & Corruccini, 1981). This may suggest that the insufficient use of the masticatory system in the early stages of life leads to unrecoverable loss of development.

Of course, there may be other reasons why foods are refused, for example, a dislike of the taste, an unfavorable smell, and unfamiliar appearance. Food refusal could also be influenced by social and emotional development. Children's food preferences are reported to resemble those of their family members (Pliner & Pelchar, 1986; Fisher et al, 2002). Imposition of stringent parental controls can potentiate food preference, limit children's acceptance of a variety of foods, and disrupt individual regulation of energy intake (Birch & Fisher, 1998). The TFP scale may also be used in further studies.

Practical use of the reference

At about 2.5 y of age, the rate of progress slowed down, which may indicate the period before 2.5 y to be sensitive in terms of food acceptance. In other words, if the timing is missed, the food that was not taken by this time might remain untaken throughout later life.

It was pointed out that child-feeding practices play a causal role in the development of food intake control, and perhaps in the etiology of problems of energy balance, especially childhood obesity (Birch, 1998). Fiber-containing foods like vegetables, fruit, and legume could help decrease metabolic risk factors involved in obesity and obesity-related disease (Rolland-Cachera et al, 1997; James et al, 1997; Pereira & Ludwig, 2001). However, a recent taste test showed that children from generally obese families had a higher preference for fatty foods, with less liking for vegetables (Wardle et al, 2001). Also, in a different research project, clinically obese individuals found fruit and vegetables to be less pleasant, and they tended to dislike tough and fibrous texture (Cox et al, 1998). Early experience of any particular food is considered to be imprinted on the memory, and may remain as a preference for a long time (Haller et al, 1999; Owada et al, 2000). For better health, vegetable and fruit should be well introduced in children's diet.

Using this reference enables the child-care provider or pediatrician to identify, at an early stage, individual children who fall outside the normal distribution of food acceptance and provide advice on which foods to introduce when. This reference standard also can be used in research or in public health surveillance to identify populations of children whose distribution of food acceptance differs substantially from the reference standard distribution. Current data, although limited, suggest that this reference can be useful in analyzing the relationship between the progress of the transitional process and its background, by using quantitative data embodied in the EAI. With further study, the background factor in eating ability could become clear by using this scale. It therefore seems that this reference standard would be a useful tool in a number of different types of enquiries, for example, the study of metabolic risk factors in early life.


Using 20 selected items of food as standards, the transitional process from liquid to solid food in infants and preschool children in Japan was studied, with the following conclusions:

  1. 1

    In child care, mass screening and daily practice, it is possible to identify children who fall outside the reference distribution of food acceptance and to help assess the delay of eating ability, using a scale based on 20 standard foods.

  2. 2

    The 50 percentile age range of these 20 standard foods covers from 5 to 42 months after birth.

  3. 3

    Most of the food items were accepted by around 2½y. After that, the rate of progress slowed down, which may indicate this period to be particularly sensitive in terms of food acceptance.

  4. 4

    The predominant cause of digestive problems, as reported by care givers, was infection, not too hard food.

  5. 5

    The transitional process is fastest in breast-fed babies and in babies bottle fed with chewing-type nipples and slowest in babies bottle fed with regular, round-hole nipples.

  6. 6

    Preparation of the infant food from the family table seems to promote progress. Children whose care giver (mainly mothers) followed the information given in books or magazines showed a slower progress.


  1. Barrett RH & Hanson ML (1978): Oral Myofunctional Disorders. St Louis, MO: The CV Mosby Co.

    Google Scholar 

  2. Beecher RM & Corruccini RS (1981): Effects of dietary consistency on craniofacial and occlusal development in the rat. Angle Orthod. 51, 61–65.

    CAS  PubMed  Google Scholar 

  3. Betz CL, Hunsberger MM & Wright S (1994): Anticipatory guidance: teething. In Nursing Care of Children, p 170. Philadelphia: WB Saunders Company.

    Google Scholar 

  4. Birch LL & Fisher JO (1998): Development of eating behaviors among children and adolescents. Pediatrics 101, 539–549.

    CAS  Google Scholar 

  5. Birch LL (1998): Development of food acceptance patterns in the first years of life. Proc. Nutr. Soc. 57, 617–624.

    CAS  Article  Google Scholar 

  6. Brandt MA, Skinner EZ & Coleman JA (1963): Texture profile method. J. Food Sci. 28, 404–409.

    Article  Google Scholar 

  7. Bronner YL & Paige DM (1992): Current concepts in infant nutrition. J. Midwifery 37, 43–58.

    Google Scholar 

  8. Cox DN, van Galen M, Hedderley D, Perry L, Moore PB & Mela DJ (1998): Sensory and hedonic judgments of common foods by lean consumers and consumers with obesity. Obes. Res. 6, 438–447.

    CAS  Article  Google Scholar 

  9. Fisher JO, Mitchell DC, Smiciklas-Wright H & Birch LL (2002): Parental influences on young girls’ fruit and vegetable, micronutrient, and fat intakes. J. Am. Diet. Assoc. 102, 58–64.

    Article  Google Scholar 

  10. Grand RJ, Watkins JB & Torti FM (1976): Development of the human gastrointestinal tract. A review. Gastroenterology 70, 790–810.

    CAS  PubMed  Google Scholar 

  11. Haller R, Rummel C, Henneberg S, Pollmer U & Koster EP (1999): The influence of early experience with vanillin on food preference later in life. Chem. Senses Aug. 24, 465–467.

    CAS  Article  Google Scholar 

  12. Hartman WM & Sarle WS (1988): Varcluster. In SAS Statistical Text, eds. KP Ingraham, RC Luginbuhl, SD Schlotzhauer & H Watts, pp 995–1014. North Carolina: SAS Institution.

    Google Scholar 

  13. Ikeda K (1998): Development of jaw muscle's function in rats fed a kneaded diet. Orthod. Waves 57, 163–172.

    Google Scholar 

  14. Imamura E (1987): Weaning instruction (in Japanese). Crit. Pediatr. 50, 68–71.

    Google Scholar 

  15. Inoue N (1993): Collapse of dentition in Japan. In Culture of Food and Oral Health in Maori, ed. N Inoue, pp 67–78. Tokyo: Therapeia Publishing Co.

    Google Scholar 

  16. Inoue N, Sakashita R & Kamegai T (1995): Reduction of masseter muscle activity in bottle fed babies. Early Hum. Dev. 42, 185–193.

    CAS  Article  Google Scholar 

  17. Ito G, Kuroe K, Inoue N & Kamegai T (1982): Experimental approach to jaw reduction. J. Dent. Res. 61, 596 (abstract).

    Google Scholar 

  18. Ito G, Mitani S & Kim JH (1988): Effects of soft diet on craniofacial growth in mice. Ant. Anz. 165, 151–166.

    CAS  Google Scholar 

  19. James WP, Nelson M, Ralph A & Leather S (1997): Socioeconomic determinants of health. The contribution of nutrition to inequalities in health. Br. J. Med. 314, 1545–1549.

    CAS  Article  Google Scholar 

  20. Launer LJ, Forman MR, Hundt GL, Sarov B, Berendes HW & Naggan L (1992): Maternal recall of infant feeding events is accurate. J. Epidemiol. Community Health 46, 203–206.

    CAS  Article  Google Scholar 

  21. Lebenthal E (1985): Impact of digestion and absorption in the weaning period on infant feeding practice. Pediatrics 75, 201–213.

    Google Scholar 

  22. Liu ZJ, Ikeda K, Harada S, Kasahara Y & Ito G (1998): Functional properties of jaw and tongue muscles in rats fed a liquid diet after being weaned. J. Dent. Res. 77, 366–376.

    CAS  Article  Google Scholar 

  23. Marui E (1982): Statistical data and research. In Statistics, eds Y Toyokawa & H Yanai, pp 6–24. Tokyo: Gendai-Suugaku.

    Google Scholar 

  24. Maternal and Child Health Division, The Ministry of Health and Welfare Japan (1986): Report on the Survey of Infant Nutrition, pp 1–102. Tokyo, Japan: The Ministry of Health and Welfare.

  25. Maternal and Child Health Division, The Ministry of Health, Labour and Welfare Japan (1995): Basic of Weaning Food. Tokyo, Japan: The Ministry of Health, Labour and Welfare.

  26. Maternal and Child Health Division, The Ministry of Health, Labour and Welfare Japan (2001): Surveys on the Growth of Infants and Preschool Children. Tokyo, Japan: The Ministry of Health, Labour and Welfare.

  27. Mennella JA (1995): Mother’s milk: a medium for early flavor experiences. J. Hum. Lact. 11, 39–45.

    CAS  Article  Google Scholar 

  28. Milla PJ (1986): The weanling's gut. Acta Pediatr. Scand. Suppl. 323, 5–13.

    CAS  Article  Google Scholar 

  29. Mizuno T, Watanabe K & Sakashita R (2000): Finger food. In Eat, Play and Development, pp 34–36. Saitama: Kaiensha.

    Google Scholar 

  30. Mutou S (2003): Child Nutrition, pp 1–81. Tokyo: Asakura-shoten.

    Google Scholar 

  31. Northstone K, Emmett P & Nethersole F (2001): The effect of age of introduction to lumpy solids on foods eaten and reported feeding difficulties at 6 and 15 months. J. Hum. Nutr. Diet. 14, 43–54.

    CAS  Article  Google Scholar 

  32. Owada M, Aoki K & Kitagawa T (2000): Taste preferences and feeding behaviour in children with phenylketonuria on a semisynthetic diet. Eur. J. Pediatr. 159, 846–850.

    CAS  Article  Google Scholar 

  33. Pereira MA & Ludwig DS (2001): Dietary fiber and body-weight regulation. Observations and mechanisms. Pediatr. Clin. N. Am. 48, 969–980.

    CAS  Article  Google Scholar 

  34. Persson LA & Carlgren G (1984): Measuring children's diets: evaluation of dietary assessment techniques and childhood. Int. J. Epidemiol. 13, 506–517.

    CAS  Article  Google Scholar 

  35. Pliner P & Pelchar ML (1986): Similarities in food preferences between children and their siblings and parents. Appetite 7, 333–342.

    CAS  Article  Google Scholar 

  36. Quandt SA (1987): Material recall accuracy for dates of infant feeding transitions. Hum. Organ. 46, 152–160.

    Article  Google Scholar 

  37. Randolph PM & Dennison CI (1981): Diet, Nutrition and Dentistry. St Louis, MO: The CV Mosby Co.

    Google Scholar 

  38. Rolland-Cachera MF, Deheeger M & Bellisle F (1997): Nutrient balance and body composition. Reprod. Nutr. Dev. 37, 727–734.

    CAS  Article  Google Scholar 

  39. Sakashita R, Kamegai T & Inoue N (1996): Masseter muscle activity in bottle feeding with the chewing type bottle teat: evidence from electromyographs. Early Hum. Dev. 45, 83–92.

    CAS  Article  Google Scholar 

  40. Sakashita R, Inoue N, Pan Q & Zhu H (1997): Diet and discrepancy between tooth and jaw size in the Yin-Shang period of China. Am. J. Phys. Anthrop. 103, 497–505.

    CAS  Article  Google Scholar 

  41. Sakashita R, Inoue N, Watanabe K, Watanabe M & Kamegai T (1998): Electromyographic evidences for sufficient and insufficient developmental space in juvenile jaws. Orthod. Waves 57, 409–416.

    Google Scholar 

  42. Sakashita R, Inoue N & Tastuki T (2001): Studies on transitional process from milk to solid food in infants. Kumamoto J. Maternal Health 5, 24–32.

    Google Scholar 

  43. Sakashita R, Inoue N & Tastuki T (2003): Selection of reference foods for a scale of standards for use in assessing the transitional process form milk to solid food in infants and pre-school children. Eur. Clin. Nutr. 57, 803–809.

    CAS  Article  Google Scholar 

  44. Sherman P (1969): A texture profile of foodstuffs based upon well-defined rheoloigical properties. J. Food Sci. 34, 458–462.

    Article  Google Scholar 

  45. Simpson WJ & Cheung DK (1976): Developing infant occlusion related feeding methods and oral habit. J. Can. Dent. Assoc. 3, 124.

    Google Scholar 

  46. Skinner JD, Carruth BR, Houck KS, Coletta F, Cotter R, Ott D & McLeod (1997): Longitudinal study of nutrient and food intakes of infants aged 2 to 24 months. J. Am. Diet Assoc. 97, 496–504.

    CAS  Article  Google Scholar 

  47. Stanley EO & Lundeen DJ (1980): Tongue thrust in breast-fed and bottle-fed school children: a cross-cultural investigation. Int. J. Oral Myol. 6, 6.

    CAS  PubMed  Google Scholar 

  48. Statistics and Information Department, Minister's Secretariat, Ministry of Health, Labour and Welfare, Japan (2001): The Basic Survey on National Life, pp 134–147. Tokyo: Health & Welfare Statistics Association.

  49. Sullivan SA & Birch LL (1994): Infant dietary experience and acceptance of solid foods. Pediatrics 93, 271–277.

    CAS  Google Scholar 

  50. Watt DG & Williams HM (1951): The effects of the physical consistency of food on the growth and development of the mandible and maxilla of the rat. Am. J. Orthod. 73, 895–928.

    Article  Google Scholar 

  51. Westenhoefer J (2001): Establishing good dietary habits—capturing the minds of children. Public Health Nutr. 4, 125–129.

    CAS  Article  Google Scholar 

  52. Wardle J, Sanderson S, Gibson EL & Rapoport L (2001): Factor-analystic structure of food preferences: I. Four-year-old children in the UK. Appetite 37, 217–223.

    CAS  Article  Google Scholar 

  53. World Health Organization (2001): The optimal duration of exclusive breastfeeding. Results of a WHO systematic review. World Health Organization

  54. Yokomizo M (1992): Research on development of masticatory behavior in kindergartners (in Japanese). J. Dent. Health 42, 227–306.

    Article  Google Scholar 

Download references


This research was supported by a Grant-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology in Japan, No. 07838030: a study on the developmental assessment of masticatory system in children from the aspect of morphological, functional, and behavioral development.

We express our sincere appreciation for the cooperation of the mothers of children studied. We are also indebted to the directors, public health nurses, and staff of public health centers in each district for their kind cooperation, support, and understanding of the authors’ aims. Finally, we thank Ms Veronica Hunt, Natural History Museum, London, for linguistic advice and comments on this paper.

Author information




Guarantor: Reiko Sakashita.

Contributors: RS participated in the study design, collected the data, carried out the statistical analysis, participated in the discussion of results, and wrote the paper. NI participated in the study design, coordinated the study, participated in the discussion of results, and reviewed the paper. TK coordinated the study and participated in the discussion of results.

Corresponding author

Correspondence to R Sakashita.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Sakashita, R., Inoue, N. & Kamegai, T. From milk to solids: a reference standard for the transitional eating process in infants and preschool children in Japan. Eur J Clin Nutr 58, 643–653 (2004).

Download citation


  • infant
  • children
  • development
  • food
  • mastication

Further reading


Quick links