TY - JOUR
AU - Monnery-Patris,, Sandrine
AB - Abstract Infants’ olfactory experience begins before birth and extends after birth through milk and complementary foods. Until now, studies on the effects of chemosensory experience in utero and/or through human milk focused on experimentally controlled exposure to only 1 target food bearing a specific odor quality and administered in sizeable amounts. This study aimed to assess whether early olfactory experience effect was measurable in “everyday conditions” of maternal food intake during pregnancy and lactation, and of infant intake at weaning, leading to expose the infant to corresponding odors as fetus, neonate, and infant up to 8 and 12 months of age. Infants’ early food exposures were assessed by asking mothers to fill out diaries about their food consumption during pregnancy and breastfeeding, and about their infant’s consumption during complementary feeding. To test odor liking, odorants representing a priori pleasant and unpleasant food odors, as well as odorless stimuli, were presented. The infant’s exploratory behavior toward odorized bottles and nonodorized control bottles was measured in terms of mouthing duration, which is thought to reflect attraction and/or appetence. At age 8 months only, positive correlations were found between liking of some unpleasant odors and early exposure to these odors through mother’s diet. No correlations were found between infants’ liking of the pleasant odors and early exposures to the foods bearing these odors. This study highlights that early exposure to unpleasant food odors may increase subsequent liking (or reduce subsequent dislike) of these food odors at least until the age of 8 months. amniotic fluid, breast milk, complementary feeding, liking, olfaction, perinatal exposure Introduction In human infants, as in other mammalian infants, olfactory experience begins well before direct exposure to solid foods and related odorous constituents (Schaal 2015, 2016). Some odor-active compounds from the pregnant and lactating mother’s diet are indeed transferred into the prenatal environment and into mother’s milk (Hauser et al. 1985; Mennella 1995; Mennella et al. 1995; Mennella and Beauchamp 1999). Such olfactory cues passing the placenta are detectable by the fetus (Marlier et al. 2001; Schaal 2015) as they subsequently influence the newborns’ reactions to corresponding odor in aerial conditions of stimulation (Hepper 1995; Schaal et al. 2000). For instance, human fetuses born to mothers who have eaten anise-flavored foods exhibit after birth higher appetence and attraction to the odor of anethole than nonexposed infants (Schaal et al. 2000). Similarly, odorants transferred from the mothers’ diet to her breast milk are detected by infants and influence their suckling behavior (Mennella and Beauchamp 1991a, 1991b, 1993a, 1993b, 1996, 1999). The effect of early experience also extends to the period of weaning or complementary feeding when infants begin to be directly exposed to solid foods. For example, infants whose mothers drank carrot juice during pregnancy or breastfeeding exhibit fewer negative facial responses than unexposed infants when they were fed cereals prepared with carrot juice than when they were fed cereals prepared with water (Mennella et al. 2001). Prenatal odor experience has been shown to influence behavior even after several years; thus, 8- to 9-year-old children whose mothers consumed garlic during pregnancy eat significantly more garlic-flavored potato puree than nonexposed children (Hepper et al. 2013). Early odor experience during milk feeding has also been shown to influence odor selectivity for more than 1 year (Delaunay-El Allam et al. 2010) and even into adulthood (Haller et al. 1999). Food-related odors are known to be important drivers of food liking and consumption in adults (Engel et al. 2006; Font-i-Furnols 2012; Jaeger et al. 2012) and, before, during, and after actual ingestion, they provoke a cascade of psychophysiological and metabolic responses (Mattes 1997; Yeomans 2006). Thus, olfaction may have a rather essential role in the control of the appetence and intake dimensions of eating behavior. Converging results exist in children and infants. For example, food neophobia is more a matter of olfactory reactivity than of taste reactivity, suggesting that the appreciation and rejection of unfamiliar foods, which is common in children and infants, may be prevalently caused by odor cues (Monnery-Patris et al. 2015). Therefore, it is relevant to better understand how and when early experience with food-related odorants in the preweaning (amniotic and lacteal) environment influences the appetitive value of odors conveyed in solid foods. The aim of this study was to investigate the relative influence of exposure in the early succession of niches constituted by the amniotic, lacteal, and the initial post-lacteal food environments on the liking of food-related odor in 8- and 12-month-old French infants. In France, 100% of infants aged on average 8 months have been repeatedly exposed to foods other than human and/or formula milks (Turberg-Romain et al. 2007). By age 12 months, the infants’ food repertoire has begun to progressively shift from baby foods to table foods (Briefel et al. 2004). Thus, 8 and 12 months can be considered as key ages for the establishment of the food repertoire. The couple of previous studies focusing on the prenatal shaping of odor-liking assessed subsequent effects in the very short term of several hours or days (Hepper 1995; Schaal et al. 2000). This study aimed to investigate longer-term effects in attempting to assess whether such early exposure effects occur in conditions where the infants are introduced to a wide panel of foods that characterizes the culture of their parents. Most studies investigating early odor or flavor exposure required indeed that participant pregnant or lactating mothers ingest one specific food in consequent amounts [e.g., 300 mL of carrot juice 4 days/week during 3 weeks (Mennella et al. 2001); 75 g of hummus with 30 mg d-carvone every third day over a period of 28 days (Hausner et al. 2010)]. The originality of this study was thus to evaluate the consequence of early odor exposure on the infant through the mothers’ spontaneous consumption of a wide spectrum of foods, without inducing them to consume any specific foods bearing a given dominant odor. Spontaneous diets were recorded during late pregnancy and lactation and for the infant, during the first months of diversified feeding. Early exposure to given odor qualities was inferred from the foods eaten by mothers (late fetal and lacteal periods) or infants (early post-lacteal period). When the infants were 8 and 12 months of age, their responses to different odor qualities conveyed by several categories of foods were measured to evaluate the magnitude of odor exposure on their hedonic responses to related odors. Because previous studies suggest positive effects of early odor exposure on later odor liking in infants (Hepper 1995; Schaal et al. 2000; Mennella et al. 2001; Schaal 2016), we hypothesized that the more frequently infants were exposed to the target olfactory stimuli during the 3 periods described earlier, the higher their subsequent liking of corresponding odors. Experimental Participants The present data were collected in the context of the “Observatory of Food preferences in Infants and Children” (OPALINE study), which aims to understand the formation of food preferences from birth up to 2 years of age. Participating mothers were recruited before or during the last trimester of pregnancy using leaflets left at health professionals and in day-care centers. To be included in the cohort, both parents had to be older than 18 years (legal age of majority), and their infant had to be void of any previous medical complication. The aims and methods of the study were explained to both parents in great detail, after which they gave their written informed consent. The study was conducted in accordance with the Declaration of Helsinki for human experimentation and was approved by the local ethics committee (Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale de Bourgogne). Mother–infant dyads (n = 235) participated in olfactory testing when the infants were 8 and 12 months old (at the mean ages [±standard deviation, SD] of 239 ± 13 days, and 372 ± 12 days, respectively). The mothers had a mean age (±SD) of 32 ± 4 years when infants where 8 months old. The infants were devoid of any oronasal affections (cold or allergy) at the time of the tests. Only the infants for whom mothers provided complete information on their diet during pregnancy and breastfeeding, and on their infant’s diet during complementary feeding, were entered into the analyses. Thus, the data set included data of 100 mother–infant dyads at the infants’ age of 8 months. Their demographic characteristics are described in Table 1. Complete data were available at 12 months for 71 of these 100 mother–infant dyads. Table 1. Characteristics of mothers and infants (n = 100) Infants Gender (N, girl/N, boy) 42/58 Feeding mode at birth (N)a  Exclusive bottle-feeding 12  Exclusive breastfeeding 81  Bottle- and breastfeeding 5 Duration of exclusive breastfeeding (days; mean ± SD) 90 ± 60 Duration of total breastfeeding (days; mean ± SD) 170 ± 152 Age at onset of complementary feeding (days; mean ± SD) 164 ± 27 Infants Gender (N, girl/N, boy) 42/58 Feeding mode at birth (N)a  Exclusive bottle-feeding 12  Exclusive breastfeeding 81  Bottle- and breastfeeding 5 Duration of exclusive breastfeeding (days; mean ± SD) 90 ± 60 Duration of total breastfeeding (days; mean ± SD) 170 ± 152 Age at onset of complementary feeding (days; mean ± SD) 164 ± 27 aThis information is missing for 2 infants. View Large Table 1. Characteristics of mothers and infants (n = 100) Infants Gender (N, girl/N, boy) 42/58 Feeding mode at birth (N)a  Exclusive bottle-feeding 12  Exclusive breastfeeding 81  Bottle- and breastfeeding 5 Duration of exclusive breastfeeding (days; mean ± SD) 90 ± 60 Duration of total breastfeeding (days; mean ± SD) 170 ± 152 Age at onset of complementary feeding (days; mean ± SD) 164 ± 27 Infants Gender (N, girl/N, boy) 42/58 Feeding mode at birth (N)a  Exclusive bottle-feeding 12  Exclusive breastfeeding 81  Bottle- and breastfeeding 5 Duration of exclusive breastfeeding (days; mean ± SD) 90 ± 60 Duration of total breastfeeding (days; mean ± SD) 170 ± 152 Age at onset of complementary feeding (days; mean ± SD) 164 ± 27 aThis information is missing for 2 infants. View Large Measurement of infants’ odor liking Odor stimuli The stimuli consisted of 8 odorants selected to represent diverse food qualities. On the basis of previous literature on early odor preferences (Wagner et al. 2013), 4 odorants were selected to represent foods considered a priori as pleasant (strawberry, peach/apricot, apple, and vanilla), and 4 odorants represented foods considered a priori as unpleasant (2-isobutyl-3-methoxypyrazine, trimethylamine, dimethyl disulfide, butyric acid; see Table 2 for the correspondence between odorants and associated foods). Control stimuli consisted of mineral oil. All the stimuli were soaked in 11 × 5 cm strips of scentless absorbent (3M) placed in the bottom of nipple-less, transparent infant-ergonomic bottles (12 × 6 cm, opening diameter of 2.3 cm; Tex, Carrefour). Table 2. Correspondences between the qualities of the test odorants and the foods conveying similar qualities Test odorant Food category Foods given to infants Vanillina Vanilla desserts Food items containing vanillin, such as vanilla cakes or cookies, vanilla dairy products Appleb Apple products Apple, food items containing apple, such as apple compote or apple pie, apple juice Peach/apricotc Peach or apricot products Peach, apricot, food items containing peach or apricot, such as peach cake or apricot pie, peach/apricot juice Strawberryd Strawberry products Strawberry, food items containing strawberry, such as strawberry yoghurt or pie, strawberry syrup Trimethylamined Fish Dried cod, whitefish, herring in oil, dog fish, fresh salmon, tuna, sardine, mackerel, fish soup, shellfish, sea shell, breaded fish Butyric acida Butyric cheeses Raclette, Emmental, Gruyère, Comté, Beaufort, Mont d’or, Saint Marcellin, Camembert, Brie, Cantal, Morbier, Tomme, Saint Nectaire, Munster, Epoisses, blue cheese, Cancoillotte, sheep cheese, Bonbel, Babybel, Port Salut Dimethyl disulfided Sulfurous cheeses Garlic and chives cheese, Cancoillotte, Camembert, Munster, Epoisses, blue cheese Dimethyl disulfided Sulfurous vegetables Broccoli, cabbage, cauliflower, Brussels sprouts, leek, chive, garlic, onion, shallot, capers 2-Isobutyl-3-methoxypyrazinea Green vegetables Green pepper, green bean, salad, peas Test odorant Food category Foods given to infants Vanillina Vanilla desserts Food items containing vanillin, such as vanilla cakes or cookies, vanilla dairy products Appleb Apple products Apple, food items containing apple, such as apple compote or apple pie, apple juice Peach/apricotc Peach or apricot products Peach, apricot, food items containing peach or apricot, such as peach cake or apricot pie, peach/apricot juice Strawberryd Strawberry products Strawberry, food items containing strawberry, such as strawberry yoghurt or pie, strawberry syrup Trimethylamined Fish Dried cod, whitefish, herring in oil, dog fish, fresh salmon, tuna, sardine, mackerel, fish soup, shellfish, sea shell, breaded fish Butyric acida Butyric cheeses Raclette, Emmental, Gruyère, Comté, Beaufort, Mont d’or, Saint Marcellin, Camembert, Brie, Cantal, Morbier, Tomme, Saint Nectaire, Munster, Epoisses, blue cheese, Cancoillotte, sheep cheese, Bonbel, Babybel, Port Salut Dimethyl disulfided Sulfurous cheeses Garlic and chives cheese, Cancoillotte, Camembert, Munster, Epoisses, blue cheese Dimethyl disulfided Sulfurous vegetables Broccoli, cabbage, cauliflower, Brussels sprouts, leek, chive, garlic, onion, shallot, capers 2-Isobutyl-3-methoxypyrazinea Green vegetables Green pepper, green bean, salad, peas aPurchased from Sigma–Aldrich. bProvided by Firmenich. cProvided by IFF. dProvided by Symrise. View Large Table 2. Correspondences between the qualities of the test odorants and the foods conveying similar qualities Test odorant Food category Foods given to infants Vanillina Vanilla desserts Food items containing vanillin, such as vanilla cakes or cookies, vanilla dairy products Appleb Apple products Apple, food items containing apple, such as apple compote or apple pie, apple juice Peach/apricotc Peach or apricot products Peach, apricot, food items containing peach or apricot, such as peach cake or apricot pie, peach/apricot juice Strawberryd Strawberry products Strawberry, food items containing strawberry, such as strawberry yoghurt or pie, strawberry syrup Trimethylamined Fish Dried cod, whitefish, herring in oil, dog fish, fresh salmon, tuna, sardine, mackerel, fish soup, shellfish, sea shell, breaded fish Butyric acida Butyric cheeses Raclette, Emmental, Gruyère, Comté, Beaufort, Mont d’or, Saint Marcellin, Camembert, Brie, Cantal, Morbier, Tomme, Saint Nectaire, Munster, Epoisses, blue cheese, Cancoillotte, sheep cheese, Bonbel, Babybel, Port Salut Dimethyl disulfided Sulfurous cheeses Garlic and chives cheese, Cancoillotte, Camembert, Munster, Epoisses, blue cheese Dimethyl disulfided Sulfurous vegetables Broccoli, cabbage, cauliflower, Brussels sprouts, leek, chive, garlic, onion, shallot, capers 2-Isobutyl-3-methoxypyrazinea Green vegetables Green pepper, green bean, salad, peas Test odorant Food category Foods given to infants Vanillina Vanilla desserts Food items containing vanillin, such as vanilla cakes or cookies, vanilla dairy products Appleb Apple products Apple, food items containing apple, such as apple compote or apple pie, apple juice Peach/apricotc Peach or apricot products Peach, apricot, food items containing peach or apricot, such as peach cake or apricot pie, peach/apricot juice Strawberryd Strawberry products Strawberry, food items containing strawberry, such as strawberry yoghurt or pie, strawberry syrup Trimethylamined Fish Dried cod, whitefish, herring in oil, dog fish, fresh salmon, tuna, sardine, mackerel, fish soup, shellfish, sea shell, breaded fish Butyric acida Butyric cheeses Raclette, Emmental, Gruyère, Comté, Beaufort, Mont d’or, Saint Marcellin, Camembert, Brie, Cantal, Morbier, Tomme, Saint Nectaire, Munster, Epoisses, blue cheese, Cancoillotte, sheep cheese, Bonbel, Babybel, Port Salut Dimethyl disulfided Sulfurous cheeses Garlic and chives cheese, Cancoillotte, Camembert, Munster, Epoisses, blue cheese Dimethyl disulfided Sulfurous vegetables Broccoli, cabbage, cauliflower, Brussels sprouts, leek, chive, garlic, onion, shallot, capers 2-Isobutyl-3-methoxypyrazinea Green vegetables Green pepper, green bean, salad, peas aPurchased from Sigma–Aldrich. bProvided by Firmenich. cProvided by IFF. dProvided by Symrise. View Large Procedure of the olfactory tests The procedure of the olfactory tests, which is described in full details elsewhere (Wagner et al. 2013), is only summarized here. Infants were individually welcomed in the presence of one parent in a laboratory room dedicated to infant testing. To control for the infants’ hunger state, parents were asked not to feed them 1.5 h before the visit to the laboratory. Parents’ compliance with this instruction was checked before the beginning of the testing session by asking when they last fed their infant. To accustom the infants to the room and experimenters, a familiarization phase took place before engaging the tests. For the tests themselves, the infants were exposed to 4 series of 3 test-bottles, sequentially ordered to present them a scentless control, a pleasant, and an unpleasant odor (Wagner et al. 2013). This order of presentation was chosen to avoid the infant refusing to pursue the test after smelling first an unpleasant odor as was noted in previous studies (Schmidt and Beauchamp 1988) and in our own pilot tests. Exposure to a bottle meant that a given bottle was put under the nose of the infant during 5 s and then placed on the table in front of the infant who was left free to explore it by any means. Within series of bottle triplets, bottles were presented separately during 60 s with a 15-s interval between bottles. Between each series of bottle triplets, we managed a 5- to 15-min break. Behavioral analyses and odor-liking scores All odor tests were videotaped for subsequent analysis of the infants’ spontaneous exploratory behavior toward the bottles. We focused on “mouthing” behavior, which was previously shown to be a reliable indicator of hedonic discrimination in infants within the age range considered (Soussignan et al. 1997; Mennella and Beauchamp 1998; Delaunay-El Allam et al. 2010). Mouthing behavior was defined as direct contact between the infant’s perioral and/or perinasal areas with the opening of the bottle. The videos were analyzed frame-by-frame with the Observer software (Noldus) by trained observers who were blind to the nature of the odor conveyed by the bottle. Ten video sequences were randomly selected to check interobserver reliability in coding mouthing behavior. The average percentage of agreement was higher than 0.90. When relevant, mouthing duration was corrected by the duration of inaccessibility of the bottle (e.g., if dropped onto the floor). The analysis was restricted to the infants who mouthed the bottles for a combined total duration of at least 10 s. Thus, the data corresponding to 10 infants at 8 months of age and to 7 infants at 12 months of age were not considered. For each test odorant, mouthing duration was converted into an odor-liking score. These scores were defined as the ratio between mouthing duration for the odorized bottle and the sum of mouthing durations for the odorized bottle and corresponding control bottle in a triplet. Although a score of 0.5 indicates indifference for the odorant, a score higher than 0.5 indicates attraction, and a score lower than 0.5 indicates avoidance of the odorant compared to the control. Recording of dietary data for deriving of odor exposure data Mothers were asked to fill out food diaries during pregnancy, breastfeeding, and the period of infant complementary feeding. The diaries were organized by day (for 7 consecutive days) and by meal. A very detailed description of the meal was required to allow a precise characterization of odor, taste, and flavor of the food items composing each meal. Mothers were also asked to give details about the ingredients and seasonings for homemade preparations and to provide the label with the list of ingredients for commercial dishes (Lange et al. 2007). The instructions were detailed to the mothers during individual sessions, and the diaries included written instructions with examples. Maternal diet during pregnancy and breastfeeding Mothers had to indicate in a qualitative way all the food or beverage items they consumed each day during 1 week of the 8th gestational month, and during the 4 weeks of the last gestational month (depending on the infant’s gestational age at birth, there were records for 1–4 weeks). The mothers also completed such diaries daily during 1 week per month during the breastfeeding period. For the analyses, we considered first the breastfeeding period from birth to 7 months in order to take into account all exposures up to the first olfactory test at 8 months; then we considered the breastfeeding period from birth to 11 months included, to take into account all exposures up to the second olfactory test at 12 months. For the infants who were not breastfed, the breastfeeding exposure was set to 0. Infant diet during complementary feeding The mothers were asked to complete diaries daily about their infants’ qualitative consumption of foods during 1 week per month. As the olfactory tests occurred at 8 and 12 months of age, we derived the food exposure data from 2 periods: the target infant’s beginning of complementary feeding (as decided and reported by the mother) to 7 months, and then from the beginning of complementary feeding to 11 months. Selection of food items for deriving infant odor exposure The frequency at which each infant was exposed to given odorants over the 3 periods studied was derived from the mothers’ and infants’ dietary diaries. We extracted from these diaries all food items carrying dominant odor qualities corresponding to the test odorants. The number of consumption occurrences was calculated from the mothers’ diaries for each food category and each period (pregnancy and breastfeeding). The number of occurrences of infants’ direct exposure to given food odor stimuli was inferred for each food category from the infants’ ingestion of non-milk foods. For the odorants characteristic of apple, strawberry, peach/apricot, and vanilla, all food items composed of apple, strawberry, peach, apricot or vanilla, or related artificial aromas were selected from the diaries. For the other odorants, 3 chemical databases of volatile compounds were systematically screened to assess their presence in food items reported in the diaries: the Flavor-Base 2010 (http://www.leffingwell.com), the VCF Database 14.1 available through TNO Triskelion (VCF Volatile Compounds in Food), and the CSGA odor database. Each test odorant was associated with 1 food category, except for dimethyl disulfide that was associated with 2 food categories. The correspondences between the test odorants and the food items containing them used for infant exposure inference are presented in Table 2. Statistical analyses The statistical analyses were carried out using the R software, version 3.0.1. Odor liking At each time point, we calculated the individual median odor-liking scores for pleasant odors and for unpleasant odors. A paired Wilcoxon test was performed to assess whether the median odor-liking scores were significantly different between a priori unpleasant and a priori pleasant odors. Kendall correlations were calculated to assess whether odor-liking scores at 8 and 12 months were correlated. As positive correlations were expected based on previous studies, unilateral tests were performed. Mothers’ food consumption during pregnancy and breastfeeding Friedman tests were performed to assess whether there were significant differences of consumptions between the food categories; when the Friedman test was significant Wilcoxon test for paired data with Bonferroni corrections were performed to compare the different food categories 2 by 2. To assess the continuity of food consumption during pregnancy and breastfeeding, Kendall correlations were calculated between the consumption of each food category during pregnancy and during breastfeeding. As positive correlations were expected, unilateral tests were performed. Relationships between odor liking and early exposures The relationships between early exposures during each period and odor-liking scores of the related odorant at 8 months (and 12 months) were assessed by calculating Kendall correlations. The relationships were also assessed for the cumulative periods of exposure, that is, prenatal plus lactational and, prenatal plus lactational and complementary feeding. According to our hypothesis, positive correlations were expected between early exposure and odor-liking score, justifying the use of unilateral tests. Results Liking of food odors The bottles containing odorants considered as unpleasant were mouthed for marginally shorter durations than the bottles containing pleasant odorants at 8 months [P = 0.09—median (Q1–Q3) for pleasant/unpleasant stimuli were: 0.41 (0.26–0.55)/0.39 (0.08–0.51)] and at 12 months, even if the difference was not significant (P = 0.11), the duration of mouthing [median (Q1–Q3)] for the a priori pleasant odors was 0.45 (0.18–0.53) versus 0.35 (0.24–0.47) for the a priori unpleasant odors. For the 71 infants for which data were available at 8 and 12 months, their individual liking scores for all odorants at the 2 age points were not correlated (all Ps > 0.11). Food items consumed during pregnancy and breastfeeding The occurrences of mothers’ consumption differed between the target food categories during pregnancy (Friedman = 452.7, P < 0.0001) and during breastfeeding (Friedman = 452.8, P < 0.0001). The mothers’ most consumed food category (Figure 1A) during pregnancy was “apple products” (occurrence: median, Q1, Q3 = 33.5, 21, 47), whereas “vanilla desserts” was one of the least consumed food categories (occurrence: median, Q1, Q3 = 6, 2, 11). During breastfeeding, a similar pattern was found (Figure 1B): the most consumed food category were “apple products” (occurrence: median, Q1, Q3 = 20.5, 4, 48.5), whereas “strawberry products” (occurrence: median, Q1, Q3 = 3, 0, 9.5), “vanilla desserts” (occurrence: median, Q1, Q3 = 4, 0, 9.25), and “sulfurous cheeses” (occurrence: median, Q1, Q3 = 4, 0, 10) were amongst the least consumed food categories. It is worth noting that consumption patterns during pregnancy and breastfeeding were weakly but significantly correlated for each food category (Kendall correlations between 0.16 and 0.34, all Ps < 0.01), suggesting a qualitative consistency in mothers’ consumption during pregnancy and the postpartum period. Figure 1. View largeDownload slide  Box-plots depicting the number of occurrences of mothers’ consumption for the different target food categories during the 2 last months of pregnancy from which infants’ in utero exposure to odorant stimuli were inferred (A) and during breastfeeding from which infants’ exposure to odorant stimuli through human milk were inferred (B). For each box-plot, the bottom and the top of the box represent respectively the 25th and 75th percentiles and the line within the box the median. The 95% confidence intervals of the medians are represented. The food categories with a different letter have significant different medians (paired Wilcoxon tests with Bonferroni correction, P < 0.05). Figure 1. View largeDownload slide  Box-plots depicting the number of occurrences of mothers’ consumption for the different target food categories during the 2 last months of pregnancy from which infants’ in utero exposure to odorant stimuli were inferred (A) and during breastfeeding from which infants’ exposure to odorant stimuli through human milk were inferred (B). For each box-plot, the bottom and the top of the box represent respectively the 25th and 75th percentiles and the line within the box the median. The 95% confidence intervals of the medians are represented. The food categories with a different letter have significant different medians (paired Wilcoxon tests with Bonferroni correction, P < 0.05). Food exposures during complementary feeding The occurrences of infants’ consumption differed between the target food categories from the beginning of complementary feeding (which is variable and depends on the mother’s decision) to 7 months (Friedman = 392.9, P < 0.0001) and to 11 months (Friedman = 408.0, P < 0.0001). From the beginning of complementary feeding to 7 months, the food categories studied were introduced to at least 25% of the infants, except for fish, sulfurous cheeses, and butyric cheeses, which were only introduced to 9, 2 and 6 infants, respectively. Among the other food categories considered here (Figure 2A), the infants were the most exposed to “apple products” [occurrence: median (Q1–Q3) = 6 (3–12)], and the least exposed to “vanilla desserts” [occurrence: median (Q1–Q3) = 0 (0–1)]. From the beginning of complementary feeding to 11 months (Figure 2B), all the food categories studied were introduced to at least 89% of infants. Infants were the most exposed to “apple products” [occurrence: median (Q1–Q3) = 27 (17.5–32)], and the least exposed to sulfurous and butyric cheeses [occurrence: median (Q1–Q3) = 0 (0–0) for both categories) and to fish [occurrence: median (Q1–Q3) = 0 (0–0.5)]. Figure 2. View largeDownload slide  Box-plots depicting food exposures between the beginning of complementary feeding and 7 months (A) and between the beginning of complementary feeding and 11 months (B). The 95% confidence intervals of the medians are represented. The food categories with a different letter have significant different medians (paired Wilcoxon tests with Bonferroni correction, P < 0.05). Figure 2. View largeDownload slide  Box-plots depicting food exposures between the beginning of complementary feeding and 7 months (A) and between the beginning of complementary feeding and 11 months (B). The 95% confidence intervals of the medians are represented. The food categories with a different letter have significant different medians (paired Wilcoxon tests with Bonferroni correction, P < 0.05). Relationships between early exposures and subsequent odor liking Relationships between in utero exposure and subsequent odor liking At 8 months, infants whose mothers reported eating more green vegetables during pregnancy displayed higher liking scores for the corresponding odorant (Table 3). In addition, in utero exposure to sulfurous cheese odor led to a marginally significant correlation with the odor-liking score for dimethyl disulfide odor at 8 months. No significant correlations were found between in utero exposures and the corresponding odor-liking scores at 12 months (Table 4). Table 3. Kendall correlations (unilateral tests) between odor-liking scores for each odorant at 8 months and early exposures to the corresponding odors Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 72 0.51 (0.33–0.67) −0.09 0.85 0.12 0.07 −0.15 0.93  Butyric acid Cheeses 79 0.39 (0–0.71) 0.04 0.31 −0.02 0.61 −0.07 0.76  Dimethyl disulfide Sulfurous cheeses 81 0.41 (0.13–0.64) 0.12 0.06 −0.04 0.71 −0.06 0.75  Dimethyl disulfide Sulfurous vegetables −0.01 0.54 −0.12 0.94 −0.12 0.91  2-Isobutyl-3-methoxypyrazine Green vegetables 79 0.51 (0.33–0.81) 0.30 <0.001 0.16 0.02 −0.07 0.79 Odors considered a priori as pleasant  Vanillin Vanilla 86 0.51 (0.28–0.72) 0.09 0.11 −0.05 0.73 −0.03 0.64  Apple Apple 77 0.54 (0.32–0.74) 0.07 0.18 0.05 0.28 −0.06 0.76  Peach/apricot Peach/apricot 75 0.54 (0.28–0.75) −0.04 0.69 −0.05 0.72 0.06 0.23  Strawberry Strawberry 76 0.48 (0.19–0.67) 0.07 0.18 0.03 0.36 0.002 0.49 Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 72 0.51 (0.33–0.67) −0.09 0.85 0.12 0.07 −0.15 0.93  Butyric acid Cheeses 79 0.39 (0–0.71) 0.04 0.31 −0.02 0.61 −0.07 0.76  Dimethyl disulfide Sulfurous cheeses 81 0.41 (0.13–0.64) 0.12 0.06 −0.04 0.71 −0.06 0.75  Dimethyl disulfide Sulfurous vegetables −0.01 0.54 −0.12 0.94 −0.12 0.91  2-Isobutyl-3-methoxypyrazine Green vegetables 79 0.51 (0.33–0.81) 0.30 <0.001 0.16 0.02 −0.07 0.79 Odors considered a priori as pleasant  Vanillin Vanilla 86 0.51 (0.28–0.72) 0.09 0.11 −0.05 0.73 −0.03 0.64  Apple Apple 77 0.54 (0.32–0.74) 0.07 0.18 0.05 0.28 −0.06 0.76  Peach/apricot Peach/apricot 75 0.54 (0.28–0.75) −0.04 0.69 −0.05 0.72 0.06 0.23  Strawberry Strawberry 76 0.48 (0.19–0.67) 0.07 0.18 0.03 0.36 0.002 0.49 aThe number of infants varied between 72 and 86 because, in some cases, infants did not complete the olfactory test because they became fussy during the session, thus data were missing for some odorants. bA score of 0.5 indicates indifference for the odorant, a score higher than 0.5 indicates attraction and a score lower than 0.5 indicates avoidance of the odorant compared to the control. Here, none of liking scores was significantly different from 0.5. View Large Table 3. Kendall correlations (unilateral tests) between odor-liking scores for each odorant at 8 months and early exposures to the corresponding odors Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 72 0.51 (0.33–0.67) −0.09 0.85 0.12 0.07 −0.15 0.93  Butyric acid Cheeses 79 0.39 (0–0.71) 0.04 0.31 −0.02 0.61 −0.07 0.76  Dimethyl disulfide Sulfurous cheeses 81 0.41 (0.13–0.64) 0.12 0.06 −0.04 0.71 −0.06 0.75  Dimethyl disulfide Sulfurous vegetables −0.01 0.54 −0.12 0.94 −0.12 0.91  2-Isobutyl-3-methoxypyrazine Green vegetables 79 0.51 (0.33–0.81) 0.30 <0.001 0.16 0.02 −0.07 0.79 Odors considered a priori as pleasant  Vanillin Vanilla 86 0.51 (0.28–0.72) 0.09 0.11 −0.05 0.73 −0.03 0.64  Apple Apple 77 0.54 (0.32–0.74) 0.07 0.18 0.05 0.28 −0.06 0.76  Peach/apricot Peach/apricot 75 0.54 (0.28–0.75) −0.04 0.69 −0.05 0.72 0.06 0.23  Strawberry Strawberry 76 0.48 (0.19–0.67) 0.07 0.18 0.03 0.36 0.002 0.49 Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 72 0.51 (0.33–0.67) −0.09 0.85 0.12 0.07 −0.15 0.93  Butyric acid Cheeses 79 0.39 (0–0.71) 0.04 0.31 −0.02 0.61 −0.07 0.76  Dimethyl disulfide Sulfurous cheeses 81 0.41 (0.13–0.64) 0.12 0.06 −0.04 0.71 −0.06 0.75  Dimethyl disulfide Sulfurous vegetables −0.01 0.54 −0.12 0.94 −0.12 0.91  2-Isobutyl-3-methoxypyrazine Green vegetables 79 0.51 (0.33–0.81) 0.30 <0.001 0.16 0.02 −0.07 0.79 Odors considered a priori as pleasant  Vanillin Vanilla 86 0.51 (0.28–0.72) 0.09 0.11 −0.05 0.73 −0.03 0.64  Apple Apple 77 0.54 (0.32–0.74) 0.07 0.18 0.05 0.28 −0.06 0.76  Peach/apricot Peach/apricot 75 0.54 (0.28–0.75) −0.04 0.69 −0.05 0.72 0.06 0.23  Strawberry Strawberry 76 0.48 (0.19–0.67) 0.07 0.18 0.03 0.36 0.002 0.49 aThe number of infants varied between 72 and 86 because, in some cases, infants did not complete the olfactory test because they became fussy during the session, thus data were missing for some odorants. bA score of 0.5 indicates indifference for the odorant, a score higher than 0.5 indicates attraction and a score lower than 0.5 indicates avoidance of the odorant compared to the control. Here, none of liking scores was significantly different from 0.5. View Large Table 4. Kendall correlations (unilateral tests) between odor-liking scores for each odorant at 12 months and early exposures to the corresponding odors (statistically significant results are shown in bold) Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 52 0.43 (0.12–0.57) 0.11 0.12 −0.04 0.67 0.08 0.23  Butyric acid Cheeses 58 0.42 (0.25–0.64) 0.07 0.23 −0.001 0.50 −0.11 0.85  Dimethyl disulfide Sulfurous cheeses 58 0.40 (0.12–0.64) −0.13 0.92 −0.17 0.96 −0.10 0.83  Dimethyl disulfide Sulfurous vegetables −0.22 0.99 −0.26 0.99 −0.05 0.69  2-Isobutyl-3-methoxypyrazine Green vegetables 60 0.50 (0.18–0.82) −0.17 0.97 −0.01 0.56 −0.17 0.96 Odors considered a priori as pleasant  Vanillin Vanilla 58 0.48 (0.26–0.69) 0.07 0.22 −0.16 0.95 −0.04 0.65  Apple Apple 60 0.49 (0.32–0.79) 0.02 0.40 −0.09 0.85 −0.02 0.59  Peach/apricot Peach/apricot 60 0.46 (0.21–0.67) 0.07 0.21 −0.01 0.54 0.08 0.20  Strawberry Strawberry 59 0.49 (0.29–0.74) −0.11 0.89 0.06 0.26 −0.06 0.75 Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 52 0.43 (0.12–0.57) 0.11 0.12 −0.04 0.67 0.08 0.23  Butyric acid Cheeses 58 0.42 (0.25–0.64) 0.07 0.23 −0.001 0.50 −0.11 0.85  Dimethyl disulfide Sulfurous cheeses 58 0.40 (0.12–0.64) −0.13 0.92 −0.17 0.96 −0.10 0.83  Dimethyl disulfide Sulfurous vegetables −0.22 0.99 −0.26 0.99 −0.05 0.69  2-Isobutyl-3-methoxypyrazine Green vegetables 60 0.50 (0.18–0.82) −0.17 0.97 −0.01 0.56 −0.17 0.96 Odors considered a priori as pleasant  Vanillin Vanilla 58 0.48 (0.26–0.69) 0.07 0.22 −0.16 0.95 −0.04 0.65  Apple Apple 60 0.49 (0.32–0.79) 0.02 0.40 −0.09 0.85 −0.02 0.59  Peach/apricot Peach/apricot 60 0.46 (0.21–0.67) 0.07 0.21 −0.01 0.54 0.08 0.20  Strawberry Strawberry 59 0.49 (0.29–0.74) −0.11 0.89 0.06 0.26 −0.06 0.75 aThe number of infants varied between 52 and 60 because, in some cases, infants did not complete the olfactory test because they became fussy during the session, thus data were missing for some odorants. bA score of 0.5 indicates indifference for the odorant, a score higher than 0.5 indicates attraction and a score lower than 0.5 indicates avoidance of the odorant compared to the control. Here, none of liking scores was significantly different from 0.5. View Large Table 4. Kendall correlations (unilateral tests) between odor-liking scores for each odorant at 12 months and early exposures to the corresponding odors (statistically significant results are shown in bold) Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 52 0.43 (0.12–0.57) 0.11 0.12 −0.04 0.67 0.08 0.23  Butyric acid Cheeses 58 0.42 (0.25–0.64) 0.07 0.23 −0.001 0.50 −0.11 0.85  Dimethyl disulfide Sulfurous cheeses 58 0.40 (0.12–0.64) −0.13 0.92 −0.17 0.96 −0.10 0.83  Dimethyl disulfide Sulfurous vegetables −0.22 0.99 −0.26 0.99 −0.05 0.69  2-Isobutyl-3-methoxypyrazine Green vegetables 60 0.50 (0.18–0.82) −0.17 0.97 −0.01 0.56 −0.17 0.96 Odors considered a priori as pleasant  Vanillin Vanilla 58 0.48 (0.26–0.69) 0.07 0.22 −0.16 0.95 −0.04 0.65  Apple Apple 60 0.49 (0.32–0.79) 0.02 0.40 −0.09 0.85 −0.02 0.59  Peach/apricot Peach/apricot 60 0.46 (0.21–0.67) 0.07 0.21 −0.01 0.54 0.08 0.20  Strawberry Strawberry 59 0.49 (0.29–0.74) −0.11 0.89 0.06 0.26 −0.06 0.75 Odorants Food category Infantsa Liking scoreb Pregnancy Breastfeeding Complementary feeding Medians (Q1–Q3) τ P value τ P value τ P value Odors considered a priori as unpleasant  Trimethylamine Fish 52 0.43 (0.12–0.57) 0.11 0.12 −0.04 0.67 0.08 0.23  Butyric acid Cheeses 58 0.42 (0.25–0.64) 0.07 0.23 −0.001 0.50 −0.11 0.85  Dimethyl disulfide Sulfurous cheeses 58 0.40 (0.12–0.64) −0.13 0.92 −0.17 0.96 −0.10 0.83  Dimethyl disulfide Sulfurous vegetables −0.22 0.99 −0.26 0.99 −0.05 0.69  2-Isobutyl-3-methoxypyrazine Green vegetables 60 0.50 (0.18–0.82) −0.17 0.97 −0.01 0.56 −0.17 0.96 Odors considered a priori as pleasant  Vanillin Vanilla 58 0.48 (0.26–0.69) 0.07 0.22 −0.16 0.95 −0.04 0.65  Apple Apple 60 0.49 (0.32–0.79) 0.02 0.40 −0.09 0.85 −0.02 0.59  Peach/apricot Peach/apricot 60 0.46 (0.21–0.67) 0.07 0.21 −0.01 0.54 0.08 0.20  Strawberry Strawberry 59 0.49 (0.29–0.74) −0.11 0.89 0.06 0.26 −0.06 0.75 aThe number of infants varied between 52 and 60 because, in some cases, infants did not complete the olfactory test because they became fussy during the session, thus data were missing for some odorants. bA score of 0.5 indicates indifference for the odorant, a score higher than 0.5 indicates attraction and a score lower than 0.5 indicates avoidance of the odorant compared to the control. Here, none of liking scores was significantly different from 0.5. View Large Relationships between breastfeeding exposure and subsequent odor liking Infants whose mothers reported eating more green vegetables during their breastfeeding period displayed a higher liking of the corresponding odor at 8 months (Table 3). In addition, exposure to fish odor during breastfeeding led to a marginally significant correlation with the odor-liking score for trimethylamine odor at 8 months. No significant correlations were found between exposures during breastfeeding and the corresponding odor-liking scores at 12 months (Table 4). Relationships between complementary feeding exposure and subsequent odor liking Direct exposures of infants to the real foods during complementary feeding were not correlated with odor-liking scores at 8 or 12 months (Tables 3 and 4). Relationships between the cumulative periods of exposure and subsequent odor liking Exposure to green vegetables odor during the perinatal period (i.e., cumulating exposures in utero and during breastfeeding) led to a significant correlation with green vegetables odor-liking score at 8 months (τ = 0.28, P < 0.001). A significant correlation was also found for this odor when we considered cumulative exposures during the perinatal period and complementary feeding (τ = 0.27, P < 0.001). Regarding the other odors, no significant correlations were found between the perinatal exposure or the cumulated exposures over the perinatal and complementary feeding periods and the corresponding odor-liking scores at 12 months. Discussion This study aimed to assess longitudinally the association between early exposure to specific dietary odorants, inferred from the diet consumed spontaneously by the pregnant-lactating mother and by the weanling infant, and subsequent liking of these odors. The relationship between maternal food consumption during pregnancy and breastfeeding and her infant’s subsequent liking for corresponding odors was based on the assumption that odorous compounds of the mother’s diet are transferred to the fetus through the placenta and to the nursling through milk. On the basis of previous studies on human fetuses and neonates (Hepper 1995; Schaal et al. 2000; Mennella et al. 2001), we hypothesized that the perinatal olfactory experience linked with the mother’s ingestion of a given food leads to higher liking for related food odors at a later age. In the conditions of this study, such exposure-liking relationships were found significant only for unpleasant odor qualities, and only at 8 months. Prenatal or lactational exposure to foods bearing unpleasant odors was associated with increased liking (or reduced disliking) responses of related odorants. However, similar odor exposure-liking relationships did not reach significance when the food exposure occurred during complementary feeding. Importantly, no relationships were found between exposures to foods bearing pleasant odor qualities and the liking of related odors. Thus, this study brought only partial support to the hypothesis that early olfactory experience would lead to an enhanced liking of a related odor separately from the context of the entire food flavor. In utero and breastfeeding exposures in relation with odor liking at 8 months No significant relationships were uncovered between the mothers’ gestational and lactational consumption frequency of foods carrying pleasant odors and the infants’ later liking for the pleasant odors at 8 and 12 months of age. Of the 4 pleasant odorants considered in this study, 3 conveyed fruity odors. The rate of transplacental and mammary transfer of compounds carrying fruity notes has not yet been systematically analyzed, although isolated evidence indicates that a fruity-smelling ester is only weakly traced into human milk (Hausner et al. 2008). Nevertheless, 1 study indicated that 4- to 8-month-old infants express higher acceptance of peaches when they have been breastfed (as opposed to formula-fed), the authors explaining this result by the fruit-rich diet of breastfeeding mothers in their study (Forestell and Mennella 2007). By contrast, prenatal exposure to odorants characteristic of green vegetables positively influenced the infants’ liking of the corresponding unpleasant odor at 8 months. A similar significant positive correlation was noted with lactational exposure to green vegetable odorants, and so was it when the cumulative perinatal exposure (in utero and during breastfeeding) was considered. The compound used here to represent green vegetable odor, 2-isobutyl-3-methoxypyrazine, has been found in human milk although without controlling whether the mothers effectively ingested green vegetables (Buettner 2007). Our results are inconsistent with the results of Forestell and Mennella (2007). They found no clear evidence for any influence of lactating mothers’ consumption of green beans on infants’ acceptance of this vegetable at the time of complementary feeding and explained their results by the low rate of consumption of green beans by the mothers and the fact that being exposed to the odor is not sufficient. Other significantly positive exposure/liking relationships were reached in this study for sulfurous cheeses and for fish, but only when infants’ liking of the corresponding odors was assessed at 8 months. The exposure to cheese was linked with ensuing liking for the cheesy odor dimethyl disulfide only when the mother consumed cheese while pregnant. Similarly, a liking for the fishy odor trimethylamine emerged only after maternal ingestion of fish during the breastfeeding period. To our knowledge, the odorant related to sulfurous cheeses, namely dimethyl disulfide, has not been identified in amniotic fluid or in human milk after the ingestion of raw garlic by mothers (Scheffler et al. 2016). However, these authors identified several garlic-derived metabolites, one of which (allyl methyl sulfide) having a pronounced garlic-like odor. Trimethylamine, the compound used here to represent fish odor, was found in both amniotic fluid and milk (Lichtenberger et al. 1991). Despite the fact that food intakes by the mothers during pregnancy and breastfeeding were correlated, the highlighted relationships regarding the 2 periods of early exposures do not concern the same foods. Moreover, when considering the cumulative exposure over the in utero and breastfeeding periods, no relationships were noted for fish or sulfurous cheeses. This raises the question of potentially variable rates of transfer of dietary odorants or their metabolites through the placenta and the mammary systems, of differential solvent properties of amniotic fluid and milk for various odor-active compounds, and/or of heterogeneous consequences of chemosensory exposure during earliest phases of development of the olfactory tract. Relationships between exposures during complementary feeding and odor liking at 8 months In the present conditions, no significant associations were found between food exposures during complementary feeding and odor liking at 8 months. Thus, the most recent olfactory exposure through direct feeding does not appear to be the strongest determinant of infants’ odor liking. For fish, sulfurous cheeses and butyric cheeses, it may be explained by the fact that these 3 food categories were introduced to less than 10 infants. For the other food categories, one possibility might be that the time elapsed between the last food exposures and the olfactory test was too short. Previous studies suggest indeed that a time interval of less than 4 days between the end of exposure and the test can affect food acceptance score negatively, by the hypothetical effect of sensory specific satiety or sensory boredom (Mennella and Beauchamp 1999; Mennella et al. 2006). Absence of relationships between early exposures and odor liking at 12 months At 12 months, the relationships between food exposures during the different periods considered in our study (prenatal, breastfeeding, complementary feeding) and ensuing odor liking by infants appeared to weaken or to vanish. Several explanations may be proposed for this result. First, the method used to gauge odor liking may be less adequate in 12-month-olds than in 8-month-olds who just begun the transition to post-lacteal foods and, hence, remain more reliably demonstrative in their behavior toward the test-bottles. Nevertheless, this explanation is unlikely as the infants differentiated pleasant and unpleasant stimuli similarly at 12 months and at 8 months. Second, the effect of exposures during the prenatal period or during the breastfeeding period may be limited in time. This is supported by studies showing that breastfeeding exposure confers an initial advantage on food acceptance during complementary feeding, but this advantage wanes when other ways enabling infants to experience flavors operate, such as the direct, repeated exposure to the dietary flavors (Maier et al. 2007; Hausner et al. 2010) without presumable attenuation by maternal and/or fetal metabolism. Moreover, there is another study finding that the effect of early exposure on infants’ preferences for basic tastes was evident in younger children still eating baby foods but was no longer evident in older children who have started to be offered table foods (Mennella et al. 2009). Third, as noted earlier, phenomena of sensory-specific satiety or of sensory boredom could explain the absence of effect between the most recent exposure to complementary feeding foods and liking responses to corresponding odors. Finally, in studies demonstrating lengthily persisting effects of olfactory experience in utero or during breastfeeding (Mennella et al. 2001; Delaunay-El Allam et al. 2010; Hepper et al. 2013), the amount of exposure by the mothers was controlled and relatively high (Mennella et al. 2001). Therefore, it is possible that the persistent effects of early exposure are found only after a long and repeated exposure to specific and relatively intense odorants. The fact that significant relationships were found only between very early exposure to given food odorants, namely, in utero and breastfeeding, and odor liking at 8 months may point to a sensitive period for the encoding of concurrent odors or flavors. These results are in line with previous experimental studies that demonstrated that the last weeks of pregnancy, as well as the first weeks of the breastfeeding period (Mennella et al. 2001), are optimal for the perception and long-term memorization of chemosensory cues (Hepper 1995; Schaal et al. 2000; Delaunay-El Allam et al. 2010; Hepper et al. 2013). Limitations of the study The findings of this study should be interpreted in light of potential limitations. The prenatal and breast milk exposures to the different target odorants were inferred from the spontaneous food intake reported by mothers during pregnancy and breastfeeding, respectively. This study is composed of a first phase of unsupervised odor/flavor exposure recording by mothers and a second phase of measurement of ensuing effects in infants. Thus, no causality can be inferred and other noncontrolled factors might account for the observed relationships. For example, a genetic predisposition of the parents could have been transmitted to their infant and could have influenced infant odor liking. It is also known that the transfer of volatile compounds of mothers’ diet to the amniotic fluid or breast milk may differ between compounds and between mothers (Schaal 2005; Hausner et al. 2008) and this could explain that different associations were found for the exposure in utero and during breastfeeding. However, considering the number of mothers who participated in the study, it would not have been possible to analyze the composition in volatile compounds in the breast milk of each mother during the whole breastfeeding period, and analyzing the amniotic fluid composition was not at all possible. During the complementary feeding period, odor exposures were inferred from the occurrences of consumption because mothers were not asked to report the quantities of the different foods eaten by the infants. However, one may wonder which parameter (number of occurrences, frequency, intensity of each exposure) is the most significant to describe the effect of olfactory exposure on odor liking. Otherwise, despite some limitations of the olfactory test, such as presentation in triplet that are discussed elsewhere (Wagner et al. 2013), the strength of this study lays in the recording of the usual food intake of a large range of foodstuffs by mothers in the last months of pregnancy or while breastfeeding, and for the infants during the complementary feeding period, followed by an experiment conducted with a quite large number of infants. In conclusion, this study highlights a positive association of prenatal and postnatal olfactory exposure and enhanced liking, or reduced disliking, of odors later in life. Specifically, early chemosensory exposures to green vegetables, sulfurous cheeses, and fish in utero and/or during breastfeeding correlate positively with responses indicating acceptance of related dominant odors at the beginning of complementary feeding. From a practical point of view, our work is important regarding the long-term persistence of perinatal odor impressions and the transmission of food culture. Further, it gives new highlights to promote the early exposure and acceptance of foods recommended for their health benefits (e.g., Anzman-Frasca et al. 2018). Funding This work was supported by the Conseil Régional Bourgogne, Franche-Comté (PARI grant) and the FEDER (European Funding for Regional Economic Development), the Institut Fédératif de Recherche 92, and the Agence Nationale de la Recherche under the “Programme National de Recherche en Alimentation et Nutrition Humaine” (project ANR-06-PNRA-028, OPALINE). The present study, labeled by Vitagora (The Taste, Nutrition and Health Innovation Pole), was also supported by grants from Blédina, Nestlé, Symrise and Cedus. The present work was part of a PhD work (S.W.) supported by a grant funded by the Nutrition, Chemical Food Safety and Consumer Behavior Division of INRA (the French National Institute for Agronomical Research) and the Conseil Régional Bourgogne, Franche-Comté. Conflicts of Interest The authors had no conflicts of interest to declare. Acknowledgements The authors wish to thank the infants and parents who took part in the present study, and also C. Laval for recruitment, A. Vincent, A. Fornerol, R. Bouhalassa, E. Szleper, J. Pierard, V. Feyen, F. Durey; V. De Anfrasio for the odor-liking data collection; C. Gulluscio for behavioral analyses; S. Jacob and S. Crevoisier for the food-liking data collection; and the whole OPALINE team for support and advice. The authors thank Symrise, IFF, and Firmenich for graciously providing the odorants. References Anzman-Frasca S , Ventura AK , Ehrenberg S , Myers KP . 2018 . Promoting healthy food preferences from the start: a narrative review of food preference learning from the prenatal period through early childhood . Obes Rev . 19 : 576 – 604 . Google Scholar Crossref Search ADS PubMed Briefel RR , Reidy K , Karwe V , Jankowski L , Hendricks K . 2004 . Toddlers’ transition to table foods: impact on nutrient intakes and food patterns . J Am Diet Assoc . 104 : s38 – s44 . Google Scholar Crossref Search ADS PubMed Buettner A . 2007 . A selective and sensitive approach to characterize odour-active and volatile constituents in small-scale human milk samples . Flavour Fragr J . 22 : 465 – 473 . Google Scholar Crossref Search ADS Delaunay-El Allam M , Soussignan R , Patris B , Marlier L , Schaal B . 2010 . Long-lasting memory for an odor acquired at the mother’s breast . Dev Sci . 13 : 849 – 863 . Google Scholar Crossref Search ADS PubMed Engel E , Martin N , Issanchou S . 2006 . Sensitivity to allyl isothiocyanate, dimethyl trisulfide, sinigrin, and cooked cauliflower consumption . Appetite . 46 : 263 – 269 . Google Scholar Crossref Search ADS PubMed Font-i-Furnols M . 2012 . Consumer studies on sensory acceptability of boar taint: a review . Meat Sci . 92 : 319 – 329 . Google Scholar Crossref Search ADS PubMed Forestell CA , Mennella JA . 2007 . Early determinants of fruit and vegetable acceptance . Pediatrics . 120 : 1247 – 1254 . Google Scholar Crossref Search ADS PubMed Haller R , Rummel C , Henneberg S , Pollmer U , Köster EP . 1999 . The influence of early experience with vanillin on food preference later in life . Chem Senses . 24 : 465 – 467 . Google Scholar Crossref Search ADS PubMed Hauser GJ , Chitayat D , Berns L , Braver D , Muhlbauer B . 1985 . Peculiar odours in newborns and maternal prenatal ingestion of spicy food . Eur J Pediatr . 144 : 403 . Google Scholar Crossref Search ADS PubMed Hausner H , Bredie WL , Mølgaard C , Petersen MA , Møller P . 2008 . Differential transfer of dietary flavour compounds into human breast milk . Physiol Behav . 95 : 118 – 124 . Google Scholar Crossref Search ADS PubMed Hausner H , Nicklaus S , Issanchou S , Mølgaard C , Møller P . 2010 . Breastfeeding facilitates acceptance of a novel dietary flavour compound . Clin Nutr . 29 : 141 – 148 . Google Scholar Crossref Search ADS PubMed Hepper PG . 1995 . Human fetal “olfactory” learning . Int J Prenat Perinat Psychol Med . 7 : 147 – 151 . Hepper PG , Wells DL , Dornan JC , Lynch C . 2013 . Long-term flavor recognition in humans with prenatal garlic experience . Dev Psychobiol . 55 : 568 – 574 . Google Scholar Crossref Search ADS PubMed Jaeger SR , Pineau B , Bava CM , Atkinson KR , McRae JF , Axten LG , Chheang SL , Beresford MK , Peng M , Paisley AG , et al. 2012 . Investigation of the impact of sensitivity to cis-3-hexen-1-ol (green/grassy) on food acceptability and selection . Food Qual Pref . 24 : 230 – 242 . Google Scholar Crossref Search ADS Lange C , Laval C , Boggio V , Nicklaus S , Issanchou S , Chabanet C , Monnery-Patris S , Schlich P . 2007 . A tool for characterizing a diet in terms of sensory variety and flavour exposure . Poster presented at the 7th Pangborn Sensory Science Symposium ; 2007 Aug 12–16 ; Minneapolis (MN) . Lichtenberger LM , Gardner JW , Barreto JC , Morriss FH Jr . 1991 . Evidence for a role of volatile amines in the development of neonatal hypergastrinemia . J Pediatr Gastroenterol Nutr . 13 : 342 – 346 . Google Scholar Crossref Search ADS PubMed Maier A , Chabanet C , Schaal B , Issanchou S , Leathwood P . 2007 . Effects of repeated exposure on acceptance of initially disliked vegetables in 7-month old infants . Food Qual Pref . 18 : 1023 – 1032 . Google Scholar Crossref Search ADS Marlier L , Schaal B , Gaugler C , Messer J . 2001 . Olfaction in premature human newborns: detection and discrimination abilities two months before gestational term . In: Marchlewska-Koj A , Lepri JJ , Müller-Schwarze D , editors. Chemical signals in vertebrates 9 . New York : Kluwer Academic . p. 205 – 209 . Mattes RD . 1997 . Physiologic responses to sensory stimulation by food: nutritional implications . J Am Diet Assoc . 97 : 406 – 413 . Google Scholar Crossref Search ADS PubMed Mennella JA . 1995 . Mother’s milk: a medium for early flavor experiences . J Hum Lact . 11 : 39 – 45 . Google Scholar Crossref Search ADS PubMed Mennella JA , Beauchamp GK . 1991a . The transfer of alcohol to human milk. Effects on flavor and the infant’s behavior . N Engl J Med . 325 : 981 – 985 . Google Scholar Crossref Search ADS Mennella JA , Beauchamp GK . 1991b . Maternal diet alters the sensory qualities of human milk and the nursling’s behavior . Pediatrics . 88 : 737 – 744 . Mennella JA , Beauchamp GK . 1993a . The effects of repeated exposure to garlic-flavored milk on the nursling’s behavior . Pediatr Res . 34 : 805 – 808 . Google Scholar Crossref Search ADS Mennella JA , Beauchamp GK . 1993b . Beer, breast feeding, and folklore . Dev Psychobiol . 26 : 459 – 466 . Google Scholar Crossref Search ADS Mennella JA , Beauchamp GK . 1996 . The human infants’ responses to vanilla flavors in human milk and formula . Infant Behav Dev . 19 : 13 – 19 . Google Scholar Crossref Search ADS Mennella JA , Beauchamp GK . 1998 . Infants’ exploration of scented toys: effects of prior experiences . Chem Senses . 23 : 11 – 17 . Google Scholar Crossref Search ADS PubMed Mennella JA , Beauchamp GK . 1999 . Experience with a flavor in mother’s milk modifies the infant’s acceptance of flavored cereal . Dev Psychobiol . 35 : 197 – 203 . Google Scholar Crossref Search ADS PubMed Mennella JA , Forestell CA , Morgan LK , Beauchamp GK . 2009 . Early milk feeding influences taste acceptance and liking during infancy . Am J Clin Nutr . 90 : 780S – 788S . Google Scholar Crossref Search ADS PubMed Mennella JA , Jagnow CP , Beauchamp GK . 2001 . Prenatal and postnatal flavor learning by human infants . Pediatrics . 107 : E88 . Google Scholar Crossref Search ADS PubMed Mennella JA , Johnson A , Beauchamp GK . 1995 . Garlic ingestion by pregnant women alters the odor of amniotic fluid . Chem Senses . 20 : 207 – 209 . Google Scholar Crossref Search ADS PubMed Mennella JA , Kennedy JM , Beauchamp GK . 2006 . Vegetable acceptance by infants: effects of formula flavors . Early Hum Dev . 82 : 463 – 468 . Google Scholar Crossref Search ADS PubMed Monnery-Patris S , Wagner S , Rigal N , Schwartz C , Chabanet C , Issanchou S , Nicklaus S . 2015 . Smell differential reactivity, but not taste differential reactivity, is related to food neophobia in toddlers . Appetite . 95 : 303 – 309 . Google Scholar Crossref Search ADS PubMed Schaal B . 2005 . From amnion to colostrum to milk. Odour bridging in early developmental transitions . In: Hopkins B , Johnson S , editors. Prenatal development of postnatal functions Advances in infancy research . New York : Praeger . p. 51 – 102 . Schaal B . 2015 . Prenatal and postnatal human olfactory development: influences on cognition and behavior . In: Doty RL , editor. Handbook of olfaction and gustation . Hoboken, NJ : John Wiley . p. 305 – 336 . Schaal B . 2016 . How amniotic fluid shapes early odor-guided responses to colostrum and milk (and more) . In: Etiévant P , Guichard E , Salles C , Voilley A , editors. Flavor: from food to behaviors, wellbeing and health ( 460 p). Cambridge, UK : Elsevier. , p. 52 – 72 . Schaal B , Marlier L , Soussignan R . 2000 . Human foetuses learn odours from their pregnant mother’s diet . Chem Senses . 25 : 729 – 737 . Google Scholar Crossref Search ADS PubMed Scheffler L , Sauermann Y , Zeh G , Hauf K , Heinlein A , Sharapa C , Buettner A . 2016 . Detection of volatile metabolites of garlic in human breast milk . Metabolites . 6 : 18 . Google Scholar Crossref Search ADS Schmidt HJ , Beauchamp GK . 1988 . Adult-like odor preferences and aversions in three-year-old children . Child Dev . 59 : 1136 – 1143 . Google Scholar Crossref Search ADS PubMed Soussignan R , Schaal B , Marlier L , Jiang T . 1997 . Facial and autonomic responses to biological and artificial olfactory stimuli in human neonates: re-examining early hedonic discrimination of odors . Physiol Behav . 62 : 745 – 758 . Google Scholar Crossref Search ADS PubMed Turberg-Romain C , Lelièvre B , Le Heuzey MF . 2007 . [Evolution of feeding behavior in mothers of infants and young children from 1 to 36 months old in France] . Arch Pediatr . 14 : 1250 – 1258 . Google Scholar Crossref Search ADS PubMed Wagner S , Issanchou S , Chabanet C , Marlier L , Schaal B , Monnery-Patris S . 2013 . Infants’ hedonic responsiveness to food odours: a longitudinal study during and after weaning (8, 12 and 22 months). Flavour. 2:19. Yeomans MR . 2006 . Olfactory influences on appetite and satiety in humans . Physiol Behav . 87 : 800 – 804 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2019. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Weanling Infants Prefer the Odors of Green Vegetables, Cheese, and Fish When Their Mothers Consumed These Foods During Pregnancy and/or Lactation
JF - Chemical Senses
DO - 10.1093/chemse/bjz011
DA - 2019-04-15
UR - https://www.deepdyve.com/lp/oxford-university-press/weanling-infants-prefer-the-odors-of-green-vegetables-cheese-and-fish-fLyyeG836P
SP - 257
VL - 44
IS - 4
DP - DeepDyve
ER -