Usage of commercial infant foods internationally

The usage of commercial infant foods has been reported by national feeding and cohort studies in several developed countries, although dietary collection methods, sampling and timeframes differ between studies, so direct comparison is not always possible. Additionally, as the focus of these studies is often breast- and formula-feeding practices and the age of introduction of solid food, not all published research specifically differentiates between commercially prepared and homemade infant food, so precise data are not always available.

National feeding data from the UK indicate that when questioned about the previous day’s dietary intake, a greater proportion of infants aged 4–6 months had been fed commercially prepared baby food than homemade baby food (38 % compared with 28 %)⁽ Reference McAndrew, Thompson and Fellows ^{39 )}. In addition, almost half (45 %) of mothers of 8- to 10-month-old infants use commercially prepared baby foods at least once per day⁽ Reference McAndrew, Thompson and Fellows ^{39 )}. Differences were observed according to maternal occupational status and ethnicity⁽ Reference McAndrew, Thompson and Fellows ^{39 )}, with those in the ‘managerial/professional’ job categories and the ‘Chinese and other’ ethnic groups less likely to use commercially prepared infant foods. However, in contrast, research from a large infant cohort study in the South of England that used principal component analyses to analyse food frequency data found clear differences in preference for wet and dry commercial infant food at age 6 months, but the pattern was not associated with many of the maternal and family characteristics considered⁽ Reference Robinson, Marriott and Poole ^{40 )}. The heterogeneous results reported are probably due to the differences in population sampling, dietary collection and statistical analysis methods used.

Similar trends are evident in other European countries. In Ireland, a birth cohort study found that 63·2 % of 6-month-old infants consumed a sweetened commercially produced cereal for breakfast, with 30–31 % consuming commercially prepared products at lunch and evening times⁽ Reference Tarrant, Younger and Sheridan-Pereira ^{41 )}. Another Irish birth cohort study reported that 49 % of foods eaten by infants in the first 6 weeks of weaning were homemade⁽ Reference O’Donovan, Murray and Hourihane ^{42 )}. In France commercial infant foods are estimated to account for 27–28 % of energy intake at ages 6–11 months, the majority of parents (63 %) offer their child commercially prepared baby foods 4–7 d/week, with only 24 % never using them⁽ Reference Bresson and Le Bris ^{43 )}. In a German birth cohort study, analysis of all food diaries completed indicated that 94·4 % of infants consumed at least one commercially prepared baby food product within a 3 d period, whereas homemade complementary food was eaten exclusively by only 5·6 % of participants⁽ Reference Foterek, Hilbig and Alexy ^{44 ,} Reference Foterek, Hilbig and Alexy ^{45 )}. Participants with higher consumption of commercial infant food were significantly older, breastfed for a shorter duration and were more likely to have mothers with a lower educational status. In Italy, a birth cohort study of 400 infant and mother pairs reported that commercially prepared infant foods were consumed in significantly higher quantities by infants who were breastfed than non-breastfed⁽ Reference Pani, Carletti and Knowles ^{46 )}. By using an estimated food diary approach, this study was able to quantify that commercial baby foods contributed a higher energy content than that of homemade foods. However, it must be stated that of these three studies, only one recruited⁽ Reference Bresson and Le Bris ^{43 )} a cross-sectional nationally representative sample.

From an economic and availability perspective, sales of commercial infant food in the twenty-seven countries of the European Union were reported to be 1271 million Euro in 2011⁽ Reference Ghisolfi, Bocquet and Bresson ^{47 )}. In Germany, the number of commercial infant products on sale increased between 2010 and 2012 from 276 to 309 jarred vegetable–potato–meat meals, demonstrating increased availability of products.

In developed countries outside of Europe, there is a similar pattern of consumption. Although the national infant feeding survey in Australia focused on breast and formula feeding and did not specifically collect data on commercial infant food^{( 48 )}, the baby and toddler food market is reported to have grown by a rate of 4·8 % per year in the last 5 years⁽ Reference Dunford, Louie and Byrne ^{36 )}. In the USA, 73–95 % of infants between the ages of 4 and 12 months consumed commercially produced baby foods when a national cross-sectional feeding survey was conducted in 2002⁽ Reference Briefel, Reidy and Karwe ^{49 )}. When repeated in 2008, the five most frequently consumed vegetables by infants aged 4–9 months were commercially prepared, rather than fresh⁽ Reference Siega-Riz, Deming and Reidy ^{50 )}. Although the proportion of infants aged 6–9 months consuming commercially prepared fruit products decreased from 66·4 to 50·2 %, between 2004 and 2008, respectively, four of the top five most frequently consumed fruits were commercially prepared rather than fresh in 2008⁽ Reference Siega-Riz, Deming and Reidy ^{50 )}. A limitation of this study, however, is that the data were collected using a 24 h recall, which may not be reflective of food group intake over a longer time period. Looking at sales figures in the USA, they have risen from 36·7 to 55 billion US dollars per annum from 2010 to 2015^{( 51 )}.

Perceptions of commercial infant foods

Studies assessing maternal perceptions of commercial infant foods have taken place in developed countries including Scotland⁽ Reference Hoddinott, Craig and Britten ^{52 ,} Reference Synnott, Bogue and Edwards ^{53 )}, England⁽ Reference Maslin, Galvin and Shepherd ^{54 ,} Reference Caton, Ahern and Hetherington ^{55 )}, France⁽ Reference Betoko, Charles and Hankard ^{56 –} Reference Maier, Chabanet and Schaal ^{58 )}, USA⁽ Reference Kim, Whaley and Gradziel ^{59 )}, Australia⁽ Reference Boak, Virgo-Milton and Hoare ^{60 )} Germany, Italy, Spain and Sweden⁽ Reference Synnott, Bogue and Edwards ^{53 )}. All of these studies, with the exception of Kim et al. ⁽ Reference Kim, Whaley and Gradziel ^{59 )}, used a focus group or structured interview approach, which enables in-depth analysis of opinions and attitudes.

Perceptions of infant feeding and commercial baby food are influenced by educational level, parity, previous experience of weaning and cultural factors⁽ Reference Maslin, Galvin and Shepherd ^{54 ,} Reference Betoko, Charles and Hankard ^{56 ,} Reference Maier, Chabanet and Schaal ^{58 )}, with second-time mothers and those from lower socio-economic groups more likely to perceive commercially prepared infant foods positively than first-time mothers or those from higher socio-economic groups⁽ Reference Maslin, Galvin and Shepherd ^{54 )}. Homemade foods are generally viewed as the ideal food by most mothers, due to the freshness of ingredients, taste⁽ Reference Synnott, Bogue and Edwards ^{53 )}, ‘avoidance of chemical in jars’⁽ Reference Hoddinott, Craig and Britten ^{52 )} and low cost; however, there is disagreement whether fresh or commercially prepared products are cheaper⁽ Reference Boak, Virgo-Milton and Hoare ^{60 )}. In most studies commercial baby foods are perceived negatively; as ‘bland’ and ‘unauthentic’⁽ Reference Schwartz, Madrelle and Vereijken ^{57 )} or only used in ‘an emergency’⁽ Reference Caton, Ahern and Hetherington ^{55 )}, with some participants saying they felt ‘a bit guilty’ using prepared foods to feed their baby⁽ Reference Hoddinott, Craig and Britten ^{52 )}. Similarly a questionnaire-based study in the USA indicated that many mothers have a preference for fresh fruits and vegetables over jarred baby foods, with mothers of older infants (9–11 months) reporting a significantly higher preference than mothers of younger infants⁽ Reference Kim, Whaley and Gradziel ^{59 )}. However, the study by Kim et al. ⁽ Reference Kim, Whaley and Gradziel ^{59 )} recruited a sample of low-income mothers who received supplemental food package and overall 83·7 % of respondents were ‘very satisfied’ with the jarred fruit and vegetables received.

Several of the same studies have also noted positive perceptions of commercial infant food by some participants. Perceived advantages of commercially produced infant foods are portability and convenience, with preparation of homemade food viewed as laborious and wasteful by some⁽ Reference Hoddinott, Craig and Britten ^{52 ,} Reference Synnott, Bogue and Edwards ^{53 ,} Reference Schwartz, Madrelle and Vereijken ^{57 )}. This is in agreement with the overall trend towards increased reliance on readymade foods across all ages generally⁽ Reference Jabs and Devine ^{61 )}. Access to fresh fruit and vegetables, leading to availability and perishability concerns, is also noted to differ depending on urban or rural location⁽ Reference Boak, Virgo-Milton and Hoare ^{60 )}. Betoko et al. ⁽ Reference Betoko, Charles and Hankard ^{56 )} reported that increased use of commercially prepared vegetables and fruit purées was explained by an awareness of nutritional advice about infant feeding, coupled with a lack of time and culinary skills to implement the advice. Indeed, some research has reported that commercial infant foods are perceived as superior to homemade foods by some mothers, describing them as ‘safer’ and possibly composed of better ingredients⁽ Reference Maslin, Galvin and Shepherd ^{54 )}. This is especially applicable to organic foods, which are viewed as ‘natural’⁽ Reference Hoddinott, Craig and Britten ^{52 )}. Commercially prepared infant foods are also seen as providing an opportunity to try out new foods that the family would not normally consume (for example, pumpkin)⁽ Reference Hoddinott, Craig and Britten ^{52 )}.

Two European studies⁽ Reference Synnott, Bogue and Edwards ^{53 ,} Reference Maier, Chabanet and Schaal ^{58 )} that recruited mothers and infants from different countries enabled exploration of cultural influences on commercial infant foods. Maier et al. ⁽ Reference Maier, Chabanet and Schaal ^{58 )} conducted structured interviews with two groups of mothers of infants aged 4–9 months: one group in Dijon, France, the other group in Aalen, Germany. Clear between- and within-group differences in weaning practices were found, with 68 % of Aalen mothers reported to prepare baby food at home more than once per week compared with 46 % of Dijon mothers. Distinct cultural differences were also reported by Synnott et al. ⁽ Reference Synnott, Bogue and Edwards ^{53 )}, who compared mothers from five different European countries. For example, all of the Italian participants chose to prepare homemade food for their infants compared with Swedish parents who were more likely to supplement home-prepared foods with commercially prepared foods. The study also highlighted that different factors were influential for each country when purchasing food for the infant. In Germany, Scotland and Sweden, health was considered the most important issue, followed by taste and organic ingredients. In Italy, the priorities were health, followed by method of production and brand compared with Spain, where the three most important factors were health, taste and brand.

Limitations of these studies are that they generally have used a small sample size, due to the qualitative approach. With the exception of Hoddinott et al. ⁽ Reference Hoddinott, Craig and Britten ^{52 )}, the cited studies have only included mothers as participants, rather than fathers or other caregivers and as with all health-related studies, selection bias and/or social desirability bias cannot be ruled out, with only those interested in infant feeding and nutrition likely to have taken part. However, overall these studies do provide rich insight into reasons why commercially prepared infant food is used so broadly, as well as the perceived disadvantages, such as taste and nutritional content.

Taste and variety of ingredients used in commercial infant foods

By incorporating a wide variety of fresh foods, ideally complementary feeding should provide a platform for the establishment of balanced taste preferences. It is known that newborn infants have an innate preference for sweet tastes and innate rejection of bitter tastes, which has developed from an evolutionary perspective to seek out energy and reject toxins⁽ Reference Beauchamp and Mennella ^{62 )}. However, these innate preferences can be manipulated with exposure to different tastes in the early stages of weaning; hence it is recommended that bitter-tasting vegetables may need to be offered several times before acceptance is achieved⁽ Reference Birch and Doub ^{63 ,} Reference Nicklaus ^{64 )}. The exposure effect has been described as consistent, powerful and universal⁽ Reference Schwartz, Scholtens and Lalanne ^{65 )}. A landmark study demonstrated that repeated exposure to similar foods in the early stages of complementary feeding can increase preference within a period of 10 d⁽ Reference Birch, Gunder and Grimm-Thomas ^{66 )}, although this study used banana and peas, rather than commercially prepared foods.

Research in the UK by Garcia et al. ⁽ Reference García, Raza and Parrett ^{22 ,} Reference Garcia, McLean and Wright ^{25 )} reported that nearly two-thirds of the 329 commercial baby foods studied were sweet, with a distinct lack of bitter vegetables. The six most common fruit and vegetables used were sweet (apple, banana, tomato, mango, carrot and sweet potato), with green vegetables such as broccoli or spinach rarely incorporated into products. In total, fruit juice was added to 18 % of products and 8·5 % of savory products had added fruit, giving them a sweet taste. This ubiquitious use of sweet flavours to mask the taste of bitter vegetables may be due to commercial pressure to manufacture instantly palatable foods⁽ Reference Garcia, McLean and Wright ^{25 )}.

It is unclear whether increased reliance on commercial infant foods reduces or increases the diversity of foods introduced during the weaning period. In theory, consumption of a varied diet should reduce the risk of developing a deficiency or excess of any particular nutrient⁽ Reference Ruel ^{67 )}, with dietary variety shown to correlate strongly with dietary adequacy in toddlers⁽ Reference Cox, Skinner and Carruth ^{68 )}. Less food diversity, defined broadly as the consumption of a narrow range of foods⁽ Reference Ruel ^{67 )}, in early life has also been associated with increased risk of any asthma, atopic asthma, wheeze and allergic rhinitis in a large birth cohort study⁽ Reference Nwaru, Takkinen and Kaila ^{69 )}.

A German study investigating food diversity in commercial infant foods reported that homemade infant meals used twenty-six different vegetables, compared with seventeen different vegetables used in commercially prepared food, with the majority of meals based on carrot⁽ Reference Mesch, Stimming and Foterek ^{24 )}, a finding also reported in a UK study⁽ Reference Carstairs, Craig and Marais ^{38 )}. Despite this, there was no difference in the variety of vegetables consumed at 6 or 9 months of age. Indeed, by 12 months of age, those fed commercial meals consumed a greater variety of vegetables. This was attributed to maternal confusion around infant feeding guidelines and that for practical food preparation reasons, infants who are fed homemade food may be fed the same homemade meal on three consecutive days, due to food being prepared in bulk.

Similarly, in a low-income sample of mothers and infants from the USA, infants aged 6–12 months who received commercial baby foods consumed a greater variety of fruits and vegetables, than those who did not⁽ Reference Hurley and Black ^{70 )}, even when adjusting for infant age, maternal education and ethnicity. Looking at longer-term outcomes, a longitudinal UK study reported that feeding home-cooked fruit or vegetables during infancy was associated with increased uptake and variety of fruit and vegetables eaten at the age of 7 years, whereas feeding commercially prepared fruit and vegetables during infancy was not⁽ Reference Coulthard, Harris and Emmett ^{71 )}. A proposed explanation for this was that commercially prepared fruit and vegetables are likely to have a uniform taste and texture, whereas those cooked at home or eaten raw will vary according to whether they are in season and the cooking method. It is also possible that the specific combination of ingredients in commercially prepared baby food may mask or interfere with learning about the particular flavour of single vegetables⁽ Reference Caton, Ahern and Hetherington ^{55 )}. A German study found no association between commercially prepared food intake in infancy and fruit and vegetable variety intake at preschool age in girls; however, in boys there was an association with reduced vegetable variety score⁽ Reference Foterek, Hilbig and Alexy ^{45 )}. The reason for this was unclear.

In summary, the existing research base underlines the fact that the development of dietary variety and taste preference is complex and multifactorial and it is not yet clear what role commercial infant food plays in either the short or long term.

Meat and fish content of commercial infant foods

Looking more broadly at food groups other than fruit and vegetables, concerns have also been raised about the limited inclusion of fish in commercially prepared infant foods⁽ Reference Mesch, Stimming and Foterek ^{24 )}. A study in Scotland highlighted the lack of infant seafood-based foods in the UK, finding that of 341 main meals available, only 3·8 % were seafood-based, compared with 30·2 % poultry-based, 35·5 % meat-based and 30·5 % vegetable-based⁽ Reference Carstairs, Marais and Craig ^{72 )}. This is seen as an important issue, as underexposure to the distinctive taste of fish may lead to reduced preference in child- and adulthood. When the study was updated in 2015, the proportion of meals containing seafood had increased to 6·3 %. However, it must be noted that this study only focused on one country, and availability of seafood-based meals in different parts of the world may be different and influenced by cultural preferences.

Fish, specifically oily fish, is of particular nutritional relevance in infancy due to the iodine⁽ Reference Zimmerman ^{73 )} and long-chain PUFA content and associated health outcomes⁽ Reference Koletzko, Lien and Agostoni ^{74 ,} Reference Riediger, Othman and Suh ^{75 )}. Although the iodine content of infant formula milk is regulated, there is no recommendation regarding minimum iodine fortification of commercial infant food in the European Union^{( 6 , 76 )}. In terms of essential fatty acids, the concern regarding a lack of sufficient long-chain PUFA in commercial infant food was noted more recently by Loughrill & Zand⁽ Reference Loughrill and Zand ^{77 )}. The contribution of fish-based meals to essential fatty acid intake was found to be low, providing only 19·9 and 3·41 % of requirements for EPA and DHA, respectively, which may be because the meals analysed were only composed of approximately 10 % of fish by weight⁽ Reference Loughrill and Zand ^{77 )}.

On the contrary, it could also be argued that non-consumption of fish is common in all infants and young children, regardless of whether they are fed predominantly homemade or prepared baby food. National UK dietary data reported that after disaggregation of composite dishes, mean consumption of fish from all sources ranged from 1 g/d for children aged 4–6 months to 6 g/d for those aged 12–18 months. The proportion of infants and young children consuming fish and fish products increased with age from 13 % at 4–6 months to 53 % at 12–18 months. This delayed introduction may be due to confusion and change in infant feeding guidelines for allergy prevention⁽ Reference Mesch, Stimming and Foterek ^{24 )}.

Red meat is also a source of long-chain PUFA, with lamb often recommended as a first meat for infants in some countries, including Italy⁽ Reference Nudda, McGuire and Battacone ^{78 )}. A study comparing the n-3 PUFA content of fresh lamb with a lamb-containing commercially prepared infant meal found a threefold higher content in fresh lamb. This may be due to the common use of vegetable oil as an ingredient in homogenised infant meat products which modifies the fatty acid composition, or due to the origin of lamb meat used in commercially prepared products⁽ Reference Nudda, McGuire and Battacone ^{78 )}. Of note, the quantity of long-chain PUFA in lamb-based commercially prepared foods was higher than that previously identified in beef-based products.

In terms of quantifying the amount of meat in an infant food product, a European Union directive^{( 6 )} states that if meat, poultry or fish are mentioned first in the name of the product, whether or not the product is presented as a meal, then the named meat, poultry or fish shall constitute not less than 10 % by weight of the total product. If meat, poultry or fish are mentioned, but not first in the name of the product, then it shall constitute not less than 8 % by weight of the total product. Following on from this stipulation, a German study concluded that the low meat composition of many commercial infant meals may increase the risk of marginal Fe status in older infants who were breastfed for 4–6 months⁽ Reference Dube, Schwartz and Mueller ^{79 )}. In Australia, Mauch et al. ⁽ Reference Mauch, Perry and Magarey ^{80 )} reported that commercial infant foods were the most common source of meat/meat alternatives consumed at age 5·5 months, but by 14 months mixed meals such as Bolognese were more common. The study concluded that parents should encourage meat ‘in a recognisable form’ and as one of the first complementary foods.

Nutritional content of commercial infant foods

Tables 1 and 2 provide a summary of studies that have investigated the nutritional content of commercial infant foods from 1997 to 2016. Studies were heterogeneous in design, assessing different types and numbers of food (main meals, desserts, snacks), obtained from several different countries. The studies have been divided into two tables, broadly dependent on the objectives of the study.

Table 1 Summary of studies on macronutrient composition with emphasis on sugar, fat and salt content

RNI, reference nutrient intake.

Table 2 Summary of studies on micronutrient and trace element composition

RNI, reference nutrient intake.

Contrasting methods have been used dependent on the objective of the study, which limits the ability to directly compare results and generalise findings. The majority of the studies investigating energy, sugar and salt content relied on nutritional content information provided by food labels⁽ Reference García, Raza and Parrett ^{22 ,} Reference Ruel ^{67 )}, which could be subject to substantial error, depending on the accuracy of the labelling information provided. Research investigating micronutrient, trace element or toxicity levels which undertook independent laboratory analysis of samples is arguably more objective⁽ Reference Mir-Marqués, González-Masó and Cervera ^{27 –} Reference Bosscher, Van Cauwenbergh and Van der Auwera ^{29 ,} Reference Melø, Gellein and Evje ^{81 –} Reference Loughrill, Govinden and Zand ^{84 )}. However, the methods, criteria and analysis standards may differ between studies. Overall there is noted to be a paucity of studies directly comparing the nutritional content of commercial and homemade infant foods, with only two studies identified that directly compared equivalent products using laboratory analysis⁽ Reference van den Boom, Kimber and Morgan ^{85 ,} Reference Randhawa, Kakuda and Wong ^{86 )}. The same conclusion was reached by a recent report, which reported that the overall evidence on nutritional composition was of low quality and direct comparison of commercially prepared infant foods with homemade foods was often lacking⁽ Reference Tzioumis, Kay and Wright ^{35 )}. In recognising the paucity of studies that directly compare commercially prepared infant with homemade infant foods, it must be highlighted that infant eating patterns are not necessarily dichotomous, i.e. that infants may be fed a combination of commercially prepared and homemade products and the proportion of each may vary at different developmental stages. Additionally, there are ethical concerns regarding infant feeding studies and therefore randomised controlled trials are probably not a suitable or practical study design to implement.

Limitations of the review

Although the present article has aimed to appraise the current literature, there are limitations and restrictions to this review. As set out in the introduction, the review has focused on complementary infant feeding in children under the age of 1 year old in developed countries. We have not included studies related to infant beverages, including infant formula or juice, as the remit would have been too broad. There are also varying degrees of legislation governing the composition, fortification and marketing of commercial infant foods in different countries; therefore findings and summaries reported may not be relevant to other countries. As previously mentioned, a wide range of categories of foods is available, which may have changed over time. Some of the studies cited have calculated the nutritional content of commercially prepared foods based on theoretical daily intakes using estimated portion sizes; however, the use and consumption of commercial and homemade foods are not dichotomous. There is also a paucity of data regarding the contribution of energy, macro- and micronutrients from commercial and homemade foods. Fundamentally, infant feeding practices are complicated by confounding variables including socio-economic and cultural factors, beliefs, attitudes and maternal diet; therefore it is not always possible to explain dietary patterns. As this is a narrative review, rather than a systematic review, it cannot be guaranteed that all the existing literature has been explored; however, an extensive literature search was undertaken.

Conclusion and future research needs

It is clear that usage of commercially prepared infant foods is very pervasive in many developed countries. Research has highlighted concerns about the altered nutritional intake, sweet taste, food diversity and toxicity of commercial infant food and the effect this could have on long-term dietary intake and health. However, the evidence base is unequivocal, complicated by different regulations between countries and a lack of randomised controlled trials. Commercially prepared baby foods have practical advantages and may improve nutritional intake and dietary variety in some situations and population groups. Overall, there are very few studies directly comparing homemade and commercial infant foods and a lack of longitudinal studies to draw firm conclusions on whether commercial infant foods are predominantly beneficial or unfavourable to infant health. It is therefore important for further high-quality research to be conducted.