Introduction

Long-chain polyunsaturated fatty acids (LC-PUFA), and docosahexaenoic acid (DHA) in particular, are considered to have a beneficial impact on brain and visual development in infancy [1–3]. During the first months after delivery, breast milk or formula milk (in the case of fortification) are important sources of DHA [4]. However, DHA requirements are still high during the second 6 months after delivery when it is recommended that breast milk and formula be successively replaced by complementary food. For infants >6 months, the European Food Safety Authority (EFSA) considers a daily intake of 100 mg DHA as adequate [5]. This raises the question as to whether DHA contents of complementary foods, on which we focused in the present survey, are sufficient to achieve the requirements [6].

A significant source of preformed DHA in solid form in infant nutrition is fatty fish, like salmon [7]. However, it was formerly recommended in Germany to avoid fish in complementary feeding for allergy prevention reasons. In contrast, in a new evidence-based guideline on allergy prevention from 2009, the Society of Paediatric Allergy and Environmental Medicine concluded early fish consumption could actually have beneficial effects on atopic diseases. Accordingly, fish should be part of complementary feeding [8]. The Research Institute of Child Nutrition (FKE) recommends in its Food Based Dietary Guidelines that infants be given fish (ca. 25 g) once per week [9]. As a recent survey on commercial and homemade baby meals in Germany only found a very limited selection of fish-containing meals, it is supposed that fish in infant nutrition is still uncommon in Germany [10].

Alternative to the intake of preformed DHA, n-3 fatty acid precursors such as α-linolenic acid (ALA) can be endogenously converted to DHA. As n-3 and n-6 PUFA share the same chain elongation/desaturation enzymes, the balanced intake of both families of fatty acids is crucial [11, 12]. However, DHA synthesis was also observed in infancy [13]. Certain vegetable oils contain relevant amounts of ALA and possess a favourable (i.e. low) n-6/n-3 ratio, i.e. linseed oil (52.8 % ALA of total fatty acids, n-6/n-3 ratio 0.3:1), walnut oil (12.2 % ALA; ratio 4.3:1) and rapeseed oil (8.6 % ALA; ratio 1.7:1) [14], although only the latter is presently consumed in Germany in similar quantities as the common n-6-rich sunflower oil [15]. Rapeseed oil is currently recommended in the Food Based Dietary Guidelines for infant nutrition in Germany [9] and already often used for commercial baby jars [10].

Taken together, fish and rapeseed oil may improve the DHA supply of infants, but little is known about their current consumption in complementary foods in Germany. The representative VELS study conducted in 2001/2002 showed a mean intake of <4 g fish/week (median 0 g/week) in the age group 6–12 months, with more than 90 % of the infants receiving no fish at all [16]. More recent data on German toddlers revealed a mean fish consumption far below the reference value in the optimised mixed diet [17, 18].

Although purchasing data of private households indicated that rapeseed oil has the highest market share in vegetable oils in Germany since 2010 [15], there is currently no general information on rapeseed oil usage for infant nutrition in Germany.

Based on a nationwide online survey on mothers of infants and toddlers concerning complementary feeding in Germany in 2010, the following key issues were examined in this data analysis:

  1. 1.

    (a) Assessment of consumption habits regarding fish and rapeseed oil in infants

    (b) Identification of potential determinants of infant fish and rapeseed oil consumption in order to reveal potential groups for targeted health promotion activities

  2. 2.

    (a) Assessment of consumption habits regarding fish and rapeseed oil in the mothers, as there is evidence that maternal dietary behaviour is one of the main determinants of complementary feeding practice [19]

    (b) Identification of potential determinants of maternal dietary behaviour.

Methods

The online nationwide survey was conducted from November 26th until December 16th 2010 and preceded the partially blinded randomised controlled intervention trial PINGU (polyunsaturated fatty acids in child nutrition—a German multimodal optimisation study) conducted in 2010–2013 in families in the region of Dortmund. The primary objective of PINGU was to examine the effects of an optimisation of the n-3 PUFA supply via:

  • salmon, rich in preformed DHA, or

  • rapeseed oil, rich in ALA, the precursor of endogenous DHA synthesis,

on endogenous DHA status in the second 6 months after delivery.

In contrast to the local PINGU intervention study, the online survey was designed to get an overall insight into habits and attitudes towards n-3 PUFA rich foods in young families in Germany. A particular objective of this data analysis was to examine the common intake and its determining factors of fish in general and rapeseed oil during the period of complementary feeding. The data collection was carried out by Kantar Health GmbH (Munich, Germany), an experienced market research company, which also conducted the National Nutrition Survey II. Data were derived from an existing access panel which is regularly contacted to participate in consumer surveys. For the present study, mothers with children aged between 5 and 36 months were selected for the online interview. Regarding mothers with more than one child between 5 and 36 months of age, responses were based on the youngest child. The interview retrospectively assessed feeding practices during infancy. Of 1,804 available mothers, 1,013 mothers who had already started with complementary feeding participated in the online interview (i.e. 56 % response rate). Of those, 28 had to be excluded due to implausible or conflicting answers (e.g. child’s current age lower than age at commencement of complementary feeding). Consequently, 985 interviews were analysed. To approach a true reflection of German mothers, data were weighed for mothers’ age, maternal level of school and professional education, household incomes, number of persons in the households, size of residences and federal states.

Outcome variables and potential determinants

Of primary interest was the fish and rapeseed oil consumption of the infants and its determinants. However, as mothers are usually the main caretaker [20], we also examined the maternal consumption of these foods and potential influencing factors. In general, the online survey was designed to examine the connection between dietary behaviour with specific respect to omega 3-rich foods for the period of complementary feeding and sociodemographic data and dietary knowledge.

The online interview consisted of 48 questions covering four main topics:

  1. 1.

    dietary behaviour of the infant and mother in terms of fish, rapeseed oil, breastfeeding and convenience levels of food consumption

  2. 2.

    sociodemographic data

  3. 3.

    maternal attitudes towards rapeseed oil and fish as a part of complementary food (not considered in the present article), and

  4. 4.

    maternal nutritional knowledge with respect to n-3 PUFA.

With regard to topic 1, the survey focused on fish and rapeseed oil during the period of complementary feeding. As even lean sea fish contributes to DHA supply and large variations in fatty fish exist (even within the species), we simplified our data collection and asked about consumption frequency of fish in general only.

The same applies for rapeseed oil, where our main interest was the individual usage (yes/no), but not absolute intake (e.g. in g per day).

Besides from fish and rapeseed oil further potential n-3 PUFA sources exist, e.g. egg yolk in the case of DHA [21, 22], and other plant oils in the case of ALA such as linseed, walnut and soybean oil [23]. However, these n-3 sources are neither explicitly recommended nor common in infant nutrition in Germany and were therefore not considered in the survey.

We defined the following outcome variables: “high fish consumption”, i.e. eating fish at least once per week; “rapeseed oil consumption”, i.e. that self-prepared and/or commercial complementary meals contained rapeseed oil; and, in the case of maternal nutrition, that rapeseed oil was used for cooking and/or for salads in the mother’s own nutrition.

Answers to topic 2 were used to define the following potential determinants of fish and rapeseed oil consumption: mother’s age (median cut; ≤32 or >32); number of children (firstborn vs. not firstborn); residence in West or East (former GDR) Germany, residence in North (coastal region: Lower Saxony, Schleswig–Holstein, Mecklenburg-Western Pomerania, Hamburg, Bremen) or South Germany; size of residence (small: ≤2,000 citizens, medium: 2,001–100,000 citizens, high: >100,000 citizens); and social classes (low, medium, high; calculated by maternal level of school education, maternal level of professional education and household income). Further potential determinants were derived from topic 1. These were “breastfeeding initiation” (child had been breastfed to some extent) and “≥4 months exclusive breastfeeding” (child received only breast milk, no complementary food or formula). Additionally, three “convenience degrees” of feeding the vegetable–potato–meat meal (which plays a central part in the German complementary feeding guideline) were defined: the mother used “mainly baby jars”, “mainly self-prepared” meals or made use of “both equally”.

Topic 4 included five questions, which were used for a rough classification of mothers’ levels of knowledge concerning n-3 PUFA (knowledge score). The first question asked whether the mother had ever heard the term “omega-3 fatty acids”. If so, the mother’s knowledge about occurrence of n-3 PUFA in different plant oils and other selected foods (different fish species and kinds of cheese) was tested. The last two questions dealt with several (both correct and incorrect) health claims which could be attributed to n-3 PUFA. If a mother answered none of the five questions correctly or negated question 1, she was defined as having “low omega-3 knowledge”, in the case of 1–3 right answers (in addition to knowing the term “omega-3 fatty acids”), the mother was defined as having “medium omega-3 knowledge”, and if all questions were answered correctly, they were to have “high omega-3 knowledge”.

Statistical analysis

All statistical tests were performed using SAS procedures (version 9.2, Statistical Analysis Systems, Cary, NC, USA). A p value <0.05 was considered as significant.

The procedure FREQ was used for the description of mothers’ and infants’ fish consumption frequency and for the usage of different types of oil. The same procedure was also used to perform the unifactorial analysis of potential determinants of high fish consumption and rapeseed oil usage by chi-square tests or Cochran–Armitage trend tests. Furthermore, logistic models (PROC Logistic) considered all potential determinants for multifactorial analyses. In infants, model 1 comprised mother’s age, social class, part of Germany, knowledge, breastfeeding behaviour, number of children and convenience degree. As maternal behaviour could be associated with these explanatory variables, we included maternal fish and rapeseed oil consumption in the extended model 2.

The maternal nutrition model 1 contained the mother’s age, social class, part of Germany and knowledge. Breastfeeding as an indicator of individual behaviour was incorporated in a separate step (model 2) to illustrate the raw impacts of the variables in model 1.

Results

Sample characteristics and consumption habits

The study population was equally distributed in terms of the children’s age strata (≤12, 13–≤24 and 25–≤36 months) and socioeconomic status (Table 1).

Table 1 Characteristics of mothers and their children who participated in a nationwide online survey in Germany (n = 985)
Full size table

Frequency of fish and rapeseed oil consumption in infants and their mothers

Fish was part of the diet for most infants and mothers to some extent (Fig. 1). However, only 25 % of infants and 41 % of mothers met the recommendation to eat fish at least once per week; 31 % of all infants never ate fish, even though only 3 % of the mothers of those children reported that fish was never part of their own diet.

Fig. 1
figure 1

Frequencies of fish (in general) consumption of infants and mothers (n = 985)

Full size image

Rapeseed oil was the most frequently used oil for vegetable–potato–meat meals without major differences between self-prepared and commercial meals (49 and 38  %, respectively), followed by sunflower oil and olive oil (Fig. 2). Overall, about half of the infants received meals containing rapeseed oil. In the case of commercial jars, 40 % of the mothers were unable to remember which oil was contained therein.

Fig. 2
figure 2

Predominantly used oils in a self-prepared vegetable–potato–meat meals (n = 731, mothers who only used baby jars not included) and b in commercial vegetable–potato–meat meals (n = 820, mothers who only used self-prepared menus not included)

Full size image

In contrast, sunflower oil (37 %) was most frequently used for cooking in maternal nutrition, followed by rapeseed oil (30 %) and olive oil (23 %). For salads, olive oil (46 %) was preferred, followed by sunflower oil (24 %) and rapeseed oil (18 %) (Fig. 3).

Fig. 3
figure 3

Predominantly used oils for mother’s own nutrition for a cooking and frying and b salads

Full size image

Determinants of infants’ fish consumption

Unifactorial analysis showed no associations between an infant’s fish consumption and the number of children in the family, social class, home region of Germany, or the mother’s omega-3 knowledge (data not shown). Only a mother’s fish intake was strongly associated with an infant’s fish intake: 44 % of those infants whose mothers ate fish at least once per week were high fish consumers, but only 11 % of those infants whose mothers ate fish less frequently had high fish consumption.

Results from the multifactorial analyses supported the hypothesis that maternal fish intake is a major determinant of an infant’s high fish consumption (Table 2). The odds ratio for infants’ high fish consumption at a low maternal fish consumption compared with high fish consuming mothers was 0.16 (95 % CI 0.12–0.23) in model 2. Although significant associations were observed in model 1 between an infant’s fish intake and both maternal nutritional knowledge and breastfeeding status, the extended model 2 did not indicate associations between those factors when the maternal dietary behaviour was considered. Besides maternal dietary behaviour, only the mother’s age seemed to determine infants’ fish consumption, as younger mothers (≤32 years) fed fish to their infants more frequently (model 2).

Table 2 Results from the multifactorial analysis of potential determinants of infant’s fish and rapeseed oil consumption in a nationwide sample of children and their mothers in Germany (n = 985)
Full size table

Determinants of infants’ rapeseed oil consumption

In the case of rapeseed oil usage in vegetable–potato–meat meals, unifactorial analyses revealed more potential determinants in comparison with an infant’s fish consumption (Fig. 4). Rapeseed oil consumption was more common in firstborns and children from higher social classes. Furthermore, regional differences were observed between East and West Germany but not between North and South Germany. A mother’s knowledge about omega-3 fatty acids was also strongly associated with an infant’s rapeseed oil consumption: 26, 47 and 73 % of mothers with low, medium and high knowledge, respectively, usually fed their infants rapeseed oil containing meals. Analogous to fish consumption, a mother’s nutritional behaviour was also a key indicator for offering rapeseed oil in complementary food: 72 % of those infants whose mothers used rapeseed oil for their own nutrition received this oil, but only 38 % of those infants whose mother used another oil were given rapeseed oil.

Fig. 4
figure 4

Unifactorial analyses of potential determinants of infantile rapeseed oil consumption in a nationwide sample of children and their mothers in Germany (n = 985) a tested by chi-square test and b tested by Cochran–Armitage trend test. The continuous line represents the mean value in the total sample. Social class was calculated by maternal level of school education, maternal level of professional education and household income. Knowledge (i.e. maternal knowledge concerning omega-3 fatty acids) was calculated by a score of five food and health-related questions

Full size image

In Model 1 of the multifactorial analyses, infants’ rapeseed oil consumption was significantly associated with social class, East/West Germany, maternal knowledge and the number of children in the family (Table 2, model 1). In contrast to an infant’s fish consumption, most associations remained unchanged when maternal dietary behaviour was considered (Table 2, model 2). Analogous to infants’ fish consumption, maternal behaviour was strongly associated with an infant’s rapeseed oil consumption. Model 2 also indicated that besides the mother’s dietary behaviour, nutritional knowledge and social class were major predictors of an infant’s rapeseed oil consumption.

Determinants of mothers’ fish and rapeseed oil consumption

Multivariate analyses (Table 3) almost completely confirmed the findings from the univariate analyses (Fig. 5), indicating that mothers aged >32 years, residing in South Germany (Table 4) and with higher omega-3 knowledge, were associated with a high maternal fish consumption (Table 3, model 1). The extended model 2 showed that ≥4 months of exclusive breastfeeding seems to be a further important indicator (Table 3, model 2).

Table 3 Multifactorial analysis of potential determinants of maternal fish and rapeseed oil consumption in a nationwide sample of mothers in Germany (n = 985)
Full size table
Fig. 5
figure 5

Unifactorial analyses of potential determinants of maternal fish consumption in a nationwide sample of mothers in Germany (n = 985) a tested by chi-square test and b tested by Cochran–Armitage trend test; 1at least 1×/week. The continuous line represents the mean value in the total sample. Social class was calculated by maternal level of school education, maternal level of professional education and household income. Knowledge, i.e. maternal knowledge concerning omega-3 fatty acids, was calculated by a score of five food and health-related questions

Full size image
Table 4 Fish consumption in the present nationwide sample of mothers in Germany (n = 985) compared to women in the National Nutrition Survey II (NVS II)
Full size table

In the case of maternal rapeseed oil consumption, maternal nutritional knowledge of omega-3 fatty acids was the only significant predictor in the unifactorial (data not shown) and multifactiorial (Table 3) analyses.

Discussion

The main finding of the present nationwide survey was that only one-fourth of German infants ate fish as often as recommended (at least once per week), whereas rapeseed oil usage in infant nutrition seemed to be more established. Half of the mothers stated that they mainly used this oil for self-prepared and/or commercial vegetable–potato–meat meals for their babies. In contrast, mothers themselves more frequently met the recommendations for fish consumption (41 %), but used rapeseed oil less often (34 %) for their own nutrition. However, maternal eating behaviour was the most important predictor for both of these n-3 PUFA rich foods in an infant’s nutrition. In contrast to infants’ fish consumption, rapeseed oil intake in infancy might be influenced by further factors, i.e. social class and omega-3 knowledge (which was also the key determinant of mothers’ own fish and rapeseed oil consumption).

Infants’ fish consumption

Infants’ fish consumption has rarely been examined in Germany and other industrialised countries, and when they have been, it has often been done using three-day (3-d) weighed dietary records. Although 3-d weighed dietary records are considered to be the gold standard of dietary assessments methods, they are not appropriate in assessing a rarely eaten food like fish. Thus, results from other studies are difficult to compare with our survey. The only nationwide German study [16], conducted in 2001/2002 using 3-d weighed dietary records, showed that more than 90 % of the infants never ate fish. Mesch et al. [10] recently assessed the supply of commercial fish meals on the German baby food market as well as use in self-prepared meals. In 2012, only nine fish-containing meals of four brands were offered. Regarding self-made complementary meals assessed by 3-d weighed dietary records in participants of the DOrtmund Nutritional and Anthropometric Longitudinally Designed (DONALD) study, only 2 % of the considered 649 records included fish. Taken together, the results indicate that fish consumption is still uncommon among German infants.

Studies in other industrialised countries revealed divergent results on fish consumption in infancy. In Stockholm (Sweden) in the 1990s, more than 60 % of the 3,619 children had fish at least once a week at the age of 12 months [24]. In Rotterdam (Netherlands) between 2002 and 2006, more than 90 % of 7,210 infants were given fish in the first year of life [25]. In contrast, in south-west England, only 25 % of 8-month-old infants born in 1992 ate fish at all [26]. In Canadian infants (n = 2,663), the fish-eating proportion of infants increased with increasing age from only 1 % at 6-months, to 12 % at 9-months and 22 % at 12-months old [27]. Overall, data suggest that lower than recommended fish consumption in infancy is not limited to German infants.

Awareness of potential determinants is, therefore, important to effectively promote healthy eating patterns including fish consumption. In the present study, maternal fish consumption was the main predictor of infants’ fish consumption. Most previous studies of resemblance in children’s and parental dietary patterns have focused on preschool and school children and showed that the association is weak, but the variation in results depended on study design, nutrient/food and mother–child dyads [28]. To the best of our knowledge, no study has specifically examined similarities between infants’ and mothers’ fish consumption. However, in a dietary pattern study in Southampton (UK), besides educational level, a mother’s diet quality was the most important determinant of an infant’s compliance with dietary guidelines, including fish consumption [19].

An important issue, we did not consider in the present analysis, is the attitude of German mothers concerning food variety in infancy. Maier et al. [29] compared French and German mothers and revealed that large differences as German mothers are more cautious and the infant’s taste development was not of high importance. In contrast, concern about allergy manifestation in German mothers seemed to inhibit a varied nutrition, including fish, while French mothers introduced fish earlier (median 12 vs. 7 months).

Mothers’ fish consumption

Although the relationship between mothers’ and infants’ fish consumption needs to be further studied, it seems to be reasonable that interventions to increase fish consumption in infancy should primarily target those families where mothers themselves have a low fish intake. As comparative data of German and international studies showed, low fish consumption in women [30–34] seems to prevail in a large proportion of young mothers. For instance, in German women mean weekly fish intake (72–98 g) was well below the recommendation (150–220 g) [31]. In a subgroup of the European Investigation into Cancer and Nutrition (EPIC) study, women in Germany and the Netherlands showed the lowest fish consumption (weekly 111 g resp. 139 g), while highest consumption was found in Spain (weekly 386 g) [32]. In the UK, median daily white and oily fish consumption for women was zero [34].

With regard to determinants for maternal fish intake, primarily nutritional knowledge impartment regarding omega-3 could be helpful to promote mothers’ fish consumption as a preceding step to improve the intention to feed fish to their infants. Comparisons with other studies are hampered due to differing definitions of nutritional knowledge—which was very specific in our study—but several studies also suggest an association between nutritional knowledge and maternal consumption of certain food groups, although not specifically fish [35, 36].

Our result that mothers in North Germany eat fish less frequently in general than in South Germany was surprising at first view. However, the NVS II (Table 4) showed an almost consistent distribution of fish consumption (g/d, determined by 24 h recalls) for women among the Northern and Southern federal states with the exception that the highest fish intake in Germany was found in Hamburg. The intake of fish was low in Mecklenburg-Western Pomerania and Bremen compared with other federal states.

Infants’ rapeseed oil consumption

An alternative approach besides fish consumption to ensure an adequate n-3 LC-PUFA supply for a large proportion of the population is the consumption of rapeseed oil, which is rich in essential ALA, the precursor to endogenous synthesis of n-3 LC-PUFA. Data on nationwide rapeseed oil consumption for children and adults are rare. The purchasing data of German private households in general showed that rapeseed oil reached the highest market share initially in 2010 [15]. Moreover, a market analysis of commercial complementary vegetable–potato–meat meals revealed that rapeseed oil was the predominantly used oil [10]. Both investigations indicate that rapeseed oil has become more prevalent in recent years.

In line with our findings on fish consumption in infancy, the usage of rapeseed oil in infant nutrition also depends on maternal behaviour, which strengthens the hypothesis of a strong relationship between general eating habits of mothers and infants [19, 28]. As illustrated, for a mother’s fish consumption, nutritional knowledge concerning omega-3 fatty acids also seemed to be an important determinant of infants’ rapeseed oil consumption. In general, the strength of influence of nutritional knowledge on dietary behaviour is debatable and depends on the observed food group and other factors like the age of the children. Effects of maternal nutritional knowledge on a child’s food consumption seem to decrease with the age of the child. In Belgian preschool children (mean age 3.5, SD = 0.4), compliance with dietary guidelines was significantly associated with maternal nutritional knowledge [37]. In an Australian socioeconomically disadvantaged sample, William et al. [38] found an association between maternal nutritional knowledge and mothers’ own diets for three out of seven food groups (including vegetables), but not for their primary school-aged children. In a US sample, stronger effects of a high maternal nutritional knowledge on the children’s diets (e.g. less sodium and more fibre) were observed in 2–5 year-olds than in 6–17 year-olds [39]. Therefore, maternal knowledge might be most important during a child’s infancy.

Strengths and limitations

A strength of the present survey is that the data collection was conducted nationwide with almost 1,000 mother-infant dyads. By data weighing, an almost complete adjustment (99.99–100) to the German population was reached according to mothers’ ages, maternal levels of school and professional education, household incomes, number of persons in the household, size of residences and federal states. The participation via internet allowed a comfortable questionnaire completion, although the results cannot be considered as representative per se as persons without internet access could not participate. To approach a true reflection of German mothers, data were weighed.

A limitation of this study is its cross-sectional design, which does not allow the identification of causalities between potential determinants and dietary behaviour. However, results allow the identification of vulnerable groups with low fish and rapeseed oil consumption in infancy. As we only assessed single aspects of complementary feeding, energy adjustment was not possible. Conversely, this allowed us to design a timesaving questionnaire in order to achieve high participation rates. Furthermore, only maternal characteristics could be considered, but no further data on the family environment (e.g. on the father) were collected. As mothers are customarily the main caregivers for their babies [20], focusing on maternal characteristics should be reasonable. One further limitation is the possibility of a recall bias that might have influenced results, especially about older children, due to the longer recall period. Therefore, further prospectively designed studies are needed to examine the impact of the potential determinants identified in the present study.

It must be noted, that the promotion of single foods should be part of a complex strategy using an adequately mixed nutrition to improve n-3 (LC-) PUFA supply as well as supply of all other nutrients [40]. As no representative data on PUFA supply in German infants exist and up to now it is just possible to define a lower level (acceptable intake, AI) for LA (AI: 4 Energy%) and ALA (AI: 0.5 Energy%) instead of quantitative dietary reference values [41], it is difficult to assess if the current situation in Germany is critical. At least data from a 6- and 9-month-old DONALD sample do not indicate an ALA deficit or an unfavourable LA/ALA ratio [4]. As the endogenous DHA synthesis is generally considered to be low [13, 42], the recommendation to prefer rapeseed oil is reasonable as the replacement of corn oil with rapeseed oil in commercial food was observed to have beneficial effects on the n-6/n-3 ratio and the n-3 LC-PUFA status in infants [43].

The former dietary guidelines for infant nutrition without consideration of fish, theoretically supplied preformed DHA (34 (19) mg/day in 6–(9) month-old infants) [4] well below the EFSA Adequate Intake (100 mg DHA per day (EFSA). Nowadays, the guidelines comprise 1–3 portions of fish to ensure an adequate intake of DHA in infancy [9, 41, 44]. Egg yolk could be an alternative, but its usage is not common in infant nutrition in Germany because of the risk of salmonella infection if not well cooked.

Taken together, rapeseed oil and (fatty) fish as strategies to improve the n-3 fatty acid supply and the n-6/n-3 ratio is state of the art in the latest Food Based Dietary Guidelines for infant nutrition in Germany [9, 44].

Effectiveness of campaigns has yet to be proven, but there are indications that knowledge improvement could have at least a supportive function, as shown in an intervention trial to promote fish consumption in pregnancy [45]. Ideally, information campaigns could be combined with environmental improvements to facilitate food/nutrient availability, as it has been started already by exchanging corn oil by rapeseed oil in commercial complementary food in Germany after the demonstration of positive effects in an intervention study under real-life conditions [43].

Conclusion

Only one in four infants in Germany met the recommendation to eat fish at least once a week. Since the recommendation to feed infants fish has been established only quite recently in Germany, it is possible that mothers are still cautious with fish for reasons like allergy prevention [8], fear of incorrect preparation or contaminants. As maternal fish consumption is associated with infants’ consumption, health promotion programs should focus on families whose mothers rarely eat fish—ideally beginning in the postnatal period when women are more interested in nutritional information [46]. In a recent randomised controlled trial on pregnant women, a positive effect of an educational intervention was observed for DHA intake from fish [45].

Although rapeseed oil has been recommended for infant nutrition in Germany since the end of the 1990s [47], only every second infant consumed this oil as part of the complementary vegetable–potato–meat meal. Rapeseed oil, containing ALA, the precursor of endogenous n-3 LC-PUFA synthesis, is a feasible means to improve the n-3 LC-PUFA supply and might be easier to implement in an infant’s nutrition than fish (because in contrast to fish no special cooking skills or a change of the dietary pattern are required, it has a neutral taste and it is inexpensive).

The revealed cross-sectional association between mothers’ omega-3 knowledge and social class and infants’ rapeseed oil consumption does not prove a causal relationship between both factors, as these determinants might also be a proxy for a general dietary attitude. Therefore, intervention studies need to examine whether an improvement of knowledge—especially focusing on lower social classes—leads to increasing rapeseed oil consumption although these programs should be combined with environmental improvements as it has been already started through the use of rapeseed oil in commercial baby jars.