Alterations in adolescents' sensory taste preferences during and after pregnancy

JOURNAL OF ADOLESCENTHEALTH1998;22:43-49

ORIGINAL ARTICLE

Alterations in Adolescents' Sensory Taste Preferences During and After Pregnancy JEAN D. SKINNER, Ph.D., JANET F. POPE, Ph.D., AND BETTY RUTH CARRUTH, Ph.D.

Purpose: The purpose of this study was to determine adolescents" preferences during and after pregnancy for foods that differed in fat, sugar, and salt content, components that are associated with m u c h of the flavor in food and, hence, are related to food acceptance. Methods: Sensory taste preferences for 16 foods were assessed by 52 white adolescents during the third trimester of pregnancy and again 11-13 months postpartum. The study design was a randomized, incomplete block design. Each subject evaluated the same eight foods at each interview using a 10-cm hedonic scale, which was scored 0 = "dislike extremely" to 10 = "like extremely." Three pairs of foods differed in fat content (e.g., ice cream vs. ice milk), two pairs differed in sugar content, but neither contained a sugar substitute (e.g., peaches in heavy syrup vs. natural juices), and three pairs differed in salt content (e.g., salted vs. unsalted peanuts). Differences in preferences were determined with least-squares analysis of variance. Results: Although preferences for 13 of the 16 foods did not differ b e t w e e n pregnancy and postpartum assessments, w h o l e milk, s k i m milk, and salted peanuts were ranked significantly higher (p -- 0.02, 0.007, and 0.05) during pregnancy. Hedonic scores were higher (p = 0.002 and 0.005) for w h o l e m i l k over s k i m at both assessments, but ice cream was not preferred significantly over lowerfat frozen products. During pregnancy, but not after, salted peanuts, chips, and crackers were preferred (p = 0.0003, 0.0001, and 0.03) over the unsalted/low-salt products. Conclusions: The adolescents" increased sensory taste preferences for m i l k and salty snack foods during pregnancy, as w e l l as personal taste preferences, should

From the Nutrition Department, the University of Tennessee, Knoxville, Tennessee Address reprint requests to: Jean D. Skinner, Nutrition Department, the University of Tennessee, Knoxville, TN 37996-1900. Janet F. Pope's current address is: College of Human Ecology, Louisiana Tech University, Ruston, LA 71272. Manuscript accepted February 19, 1997.

be recognized by clinicians, and nutrition recommendations should be individualized to enhance compliance. © Society for Adolescent Medicine, 1998 KEY WORDS;

Pregnant adolescents Sensory preference taste testing Food preferences

Many adolescents begin pregnancy with poor nutritional status (1-7) owing to eating patterns of adolescents (7-10) and delayed initiation of prenatal care (1,3). Teen pregnancy may interrupt formal education, and the pregnant adolescent often lacks the support of an established family unit (3). Given these many stresses, pregnant teens may be inclined to eat those foods that provide them pleasure and those that are familiar to them, not necessarily foods that enhance the nutritional quality of their diets. Pregnant women frequently report that their preferences for certain foods change during pregnancy and that they either like a food that was previously disliked or they dislike a food that was previously enjoyed (11,12). Some of these preference changes may result from physiological changes in the taste mechanism (13-15). However, some also report psychological factors affecting preference changes (11,16). Moreover, preference changes differ among pregnant women, and taste sensitivities differ among pregnancies in the same woman (14). The taste preferences of pregnant adolescents have not, to our knowledge, been studied. Though not synonymous, taste sensitivity is assumed to be related to preference and acceptance of food. Food preference is a complex reaction affected by numerous psychological and physiological vari-

© Societyfor AdolescentMedicine,1998 Publishedby ElsevierScienceInc.,655Avenueof the Americas,NewYork,NY10010

1054-139X/98/$19.00 PII $1054-139X(97)00207-3

44

SKINNERET AL.

ables (17,18). Preference testing of foods is done by sensory preference (hedonic) ratings with actual tasting of foods or by reporting preferences when the subject is given a list of foods. The sensory ratings are more valid assessments of preference because all subjects are responding to the same food product under similar conditions. Taste is one of the physiological mechanisms that can influence preference, and taste is only one of several sensory perceptions. In the real world, each time a food is eaten a sensory evaluation is made, using the senses of sight, taste, smell, touch, and hearing. Taste sensitivity generally is measured by threshold tests in which subjects are presented increasing concentrations of a substance in solution; the threshold is the concentration at which they first detect the substance. Thus, lower thresholds indicate greater sensitivity. Although results of several studies confirm that changes in threshold (i.e., sensitivity for detecting a taste) and in sensory preferences for solutions of the four primary tastes (sweet, sour, bitter, and salty) occur during pregnancy in adults (14,15,19), results often are inconsistent and none have been reported in adolescents. Brown and Toma (14) reported no differences in sensitivity for sweet tastes or sucrose preferences between pregnant and nonpregnant adult women, whereas Dipple and Elias (19) suggested that weaker solutions of sucrose are preferred during pregnancy. Similar conclusions about sucrose solutions were reached in an animal study (20). Both increased sensitivity (15) and decreased sensitivity to salt solutions (14) have been reported. In addition, Brown and Toma (14) reported that pregnant women preferred saltier solutions than did nonpregnant women. Animal studies indicated that pregnant rats have increased appetite for salt (21). Takahashi et al. (15) reported increased thresholds (i.e., decreased sensitivity) for bitter and sour tastes in pregnant women. This seems contradictory to some frequently reported food aversions during pregnancy (11) (i.e., coffee and tea), which suggest increased sensitivity. It is clear that threshold tests and other assessments of solutions with a single primary taste may not accurately reflect preferences for real foods, for which there are many potential determinants. The purpose of this study was to determine adolescents' preferences during and after pregnancy for foods that differed in fat, sugar, and salt content. Fat, sugar, and salt are responsible for much of the flavor associated with food. Although adolescents may make some dietary changes for the health of the fetus (22), they are more likely to do so if they can make

JOURNALOF ADOLESCENTHEALTHVol. 22, No. 1

dietary changes without sacrificing taste and other sensory characteristics.

Methods Subjects Subjects in this study were a subgroup of participants in a longitudinal study of factors influencing the nutritional health of pregnant adolescents (2325). Criteria for inclusion in the sensory preference testing portion of the study included: (a) being between 11 and 17 years of age at conception; (b) being in the third trimester of pregnancy during the initial 11-month period of data collection; (c) not being pregnant at 11-13 months postpartum (selfreport), and (d) residing within a 120-mile radius of the research university to facilitate personal interviews. Pregnant adolescents were recruited from health clinics, schools, community programs, physicians' practices, and residential group homes. Following approval from the university's committee on research participation and similar approvals from schools and agencies as required, professionals employed in the various settings requested permission from potential participants to release the girl's name to the research team. The girl was contacted by a member of the research team who described the study to her and answered her questions, and if she was willing to participate, the first interview was scheduled. The technique of oversampling was used with the expectation that some subjects would be unavailable for the postpartum data collection or ineligible owing to another pregnancy. The initial sample included 81 adolescents who were in the third trimester of pregnancy at the time of the first interview.

Data Collection Eight pairs of food were selected for sensory preference testing; food pairs were selected for their visual similarity despite varying degrees of fat, salt, or sugar. Three pairs of food differed in their fat content: whole milk versus skim milk, vanilla ice cream versus vanilla ice milk, and strawberry ice cream versus low-fat strawberry frozen yogurt. Two food pairs differed in sugar content: canned sliced peaches packed in heavy syrup versus canned sliced peaches packed in natural fruit juices, and ready-toeat chocolate pudding versus the same product with added sugar; the chocolate pudding with added

January 1998

PREGNANTADOLESCENTS'SENSORYPREFERENCES

45

sugar was the only product altered by the researchers. Three pairs of foods differed in salt content: salted versus unsalted peanuts, salted versus unsalted potato chips, and regular salted crackers versus low-salt crackers. To avoid sensory differences based on brand characteristics, each pair represented a single brand used consistently throughout the study. Because of the assessments during pregnancy and about 12 months postpartum, it was not possible to both retain food quality and purchase all foods initially; thus, products were purchased fresh as needed and stored according to manufacturers' directions. To provide consistent environmental conditions for sensory evaluation, interviews were conducted in a mobile laboratory that had complete kitchen facilities, heat, and air-conditioning. The mobile laboratory was driven to a site convenient to the participant. Each subject was asked to evaluate 8 of the 16 foods; the number of foods evaluated was limited to avoid tasting fatigue a n d / o r decreased sensitivity during the sensory preference testing, which can decrease the reliability of the data obtained (26). A computer-generated, randomized, incomplete block design was used to assign foods and the sequence of tasting foods for each subject. This procedure enhanced the amount of information obtained while reducing the burden on each subject. Tasting fatigue was further avoided by dividing the sensory preference testing into two time periods during the interview. Collection of demographic data intentionally interrupted the sensory testing periods. Girls evaluated the same eight foods during and after pregnancy. They each were assigned 4 of the 6 foods that differed in fat content and 4 of the 10 remaining foods that differed in sugar or salt content. Greater emphasis was placed on foods in the fat category because of the absence of data on sensory preference testing for fat in foods for pregnant subjects, whereas salt and sugar perceptions had been reported for pregnant adults (14,15,19) but not for pregnant adolescents. Thus, each of the foods that differed in fat were initially evaluated by 54 girls; 33 girls evaluated each food that differed in sugar, and 32 girls evaluated each salty food. Products were served to subjects at temperatures at which these foods normally are eaten. Ice cream, ice milk, and frozen yogurt were solidly frozen; the milks, puddings, and canned peaches had been chilled in the refrigerator; and nuts, chips, and crackers were served at room temperature. Preparation of food samples for presentation was completed prior

to the subject entering the mobile laboratory so that food products could not be identified by brand or type (e.g., ice cream vs. ice milk, salted vs. unsalted chips). Subjects were presented food samples in white plastic or paper containers; portion sizes for liquid and solid foods were approximately 30 mL and 60 g, respectively. Food samples were not identified for the subjects, but were labeled with a randomly selected three-digit number. Cool water, carrot strips, and apple wedges were provided for subjects to cleanse the palate prior to and between tasting food samples. Subjects evaluated each food product by marking on a 10-cm unstructured hedonic scale with the anchors "dislike extremely" (scored 0) and "like extremely" (scored 10). The scale responses were measured for scoring and recorded to the nearest 0.5 ram. Statistical Analysis Demographic data were summarized with the Statistical Analysis System programs PROC MEANS and PROC FREQ (27). Socioeconomic status (SES) was determined with the Four-Factor Hollingshead Index, which uses education and occupation of the family's primary wage earner(s) to yield a score ranging from 8 (low SES) to 66 (high SES) (28). Mean hedonic scores and standard errors of the mean (SEM) were determined twice (pregnancy and postpartum) for each of the 16 foods using the USSES general linear mixed model (GLMM) procedure least-squares means (LS means) (29). GLMM was used instead of the traditional general linear model (GLM) because the effects being evaluated were neither fixed nor completely random. GLMM assumes that all effects are in fact "mixed effects." LS means are appropriate when cell sizes differ, as was the case with food data. LS means were used to determine the group's extent of "dislike" or "like" for each food product. Using the 0-10 hedonic scale, we interpreted means of 4.0-6.0 as neutral and categorized foods above 6.0 as liked and those below 4.0 as disliked by the adolescents. The differences in hedonic scores between food pairs (e.g., whole milk vs. skim milk) were tested with least-squares analysis of variance using the GLMM procedure CONTRAST (30), with food and subject in the model. These analyses showed whether the subjects had equal preferences for foods that differed in fat, sugar, and salt. The GLMM CONTRAST (30) procedure was also used to test for differences in hedonic scores for each

46

SKINNER ET AL.

food during and after pregnancy with food, status (i.e., pregnant/postpartum), and subject in the model. Because these analyses included the same subjects interviewed during the third trimester of pregnancy and again at 11-13 months postpartum, and "subject" was included in the model, the effects of individual preferences were controlled.

Results Characteristics of Subjects Of the 81 girls who completed the sensory preference (hedonic) testing during pregnancy, 52 (64.2%) were eligible and completed the second sensory testing 11-13 months postpartum. The oversampling of pregnant adolescents was warranted because almost 20% (n = 15) of the girls could not be located in the postpartum period a n d / o r had moved out of state. In addition, seven girls were pregnant again and thus were dropped from the analyses, and an equal number (n = 7) of postpartum girls declined to participate or failed to keep scheduled appointments. Comparisons of demographic data for girls who completed both the pregnancy and postpartum assessments and those who completed only the interviews during pregnancy indicated that the groups were similar. The initial (n = 81) and final groups (n = 52) were similar in age at first interview (X = 16.3 years) parity (~90% primiparous), and parents' education level. Seventy-three percent of the family wage earners were machine operators or in semiskilled or unskilled occupations; 27% were in clerical, sales, or skilled occupations. As would be expected, some changes in the girls' education and living arrangements occurred between the pregnancy and postpartum assessments. At 11-13 months postpartum, 48% of the girls had completed high school and only 25% were in school, compared to < 20% of pregnant girls who had completed high school and 67% who were enrolled in school during pregnancy. Living arrangements of some girls had changed, with more living independently with their spouse (40%) than during pregnancy (17%). SES of the postpartum group was slightly lower (Hollingshead Index Y( = 24.8 + 9.2) than the initial group (X = 26.2 + 9.2), possibly because some girls' SES after pregnancy was calculated based on the education and occupation of their spouse, rather than their father who was established in the workforce.

JOURNAL OF ADOLESCENT HEALTH Vol. 22, No. 1

Hedonic Ratings Initial hedonic ratings of the 81 girls interviewed only once and the initial ratings of the 52 girls assessed twice were very similar. Most foods (66%) evaluated during and after pregnancy (Table 1) had mean scores above 6.0 on the 11-point hedonic scale, and thus were categorized as foods liked by the adolescents. Scores for several foods were in the 4.0-6.0 neutral category (e.g., unsalted nuts, chips, crackers, whole milk). The only disliked food was skim milk. This occurred at the postpartum evaluation but not during pregnancy.

Differences Between Food Pairs Among the food pairs that differed in fat, only whole milk was preferred over skim milk (Table 1). Hedonic scores were not significantly different for ice cream versus ice milk or ice cream versus low-fat frozen yogurt. Similarly, there were no significant differences in preference for pairs that differed in sugar. However, during pregnancy but not after, the adolescents preferred the salted food products over the unsalted or low-salt alternatives.

Hedonic Ratings During versus After Pregnancy Preferences for 13 of the 16 foods did not differ significantly by pregnancy status (Table 1). However, whole milk, skim milk, and salted peanuts were preferred less after pregnancy compared to during pregnancy, p = 0.02, 0.007, and 0.05, respectively.

Discussion This study is unique because actual foods, rather than solutions with a single, specific taste (e.g., salty, bitter), were used with sensory preference (hedonic) methodology to assess food preferences of adolescents during and after pregnancy. It also is unique in the assessment of pregnant adolescents' preferences related to fat content of similar foods. Adolescents liked most of the foods included in the sensory preference analysis because foods were selected as those reportedly popular with teens, such as salty snack foods (8,9) a n d / o r those craved during pregnancy, such as milk and chocolate (11,12). A major finding of this study was that the teenagers preferred the higher-fat whole milk over skim milk both during and after pregnancy. However, they did not significantly prefer ice cream over either ice milk or frozen yogurt. These data indicate that some

January 1998

PREGNANT ADOLESCENTS' SENSORY PREFERENCES

47

Table 1. Adolescents' Sensory Preference (Hedonic) Scores* on Paired Foods That Differ in Fat, Sugar, or Salt Content Pregnancy* Food Pairs

Pregnancy vs. Postpartum

Postpartum *

Nt

Mean (SEM)

F (p)§

Mean (SEM)

F (p)§

F (p)ll

Whole milk Skim milk

33 36

5.70 (0.37) 4.16 (0.37)

9.56 (0.002)

4.41 (0.48) 2.67 (0.46)

7.82 (0.005)

5.22 (0.02) 7.25 (0.007)

Vanilla ice cream Vanilla ice milk

38 36

7.84 (0.37) 7.40 (0.37)

0.06

7.71 (0.45) 7.91 (0.46)

0.11

0.84 0.08

Strawberry ice cream Strawberry frozen yogurt

34 31

7.36 (0.37) 6.85 (0.37)

1.07

6.59 (0.47) 6.73 (0.49)

1.07

0.04 2.06

Canned peaches in syrup Canned peaches in natural juice

18 22

6.84 (0.47) 6.28 (0.47)

0.74

7.50 (0.64) 7.21 (0.58)

0.13

0.88 1.48

Chocolate pudding plus sugar Chocolate pudding

16 25

7.67 (0.47) 7.01 (0.47)

1.04

7.44 (0.67) 6.83 (0.54)

0.54

0.08 0.07

Salted peanuts Unsalted peanuts

21 23

7.42 (0.48) 4.57 (0.48)

13.26 (0.0003)

5.99 (0.59) 5.40 (0.56)

0.57

4.00 (0.05) 1.51

Salted potato chips Unsalted potato chips

21 19

6.83 (0.48) 4.45 (0.48)

18.94 (0.0001)

7.19 (0.59) 5.74 (0.62)

3.08

0.18 2.92

Salted crackers Low-salt crackers

21 22

6.84 (0.48) 5.42 (0.48)

4.73 (0.03)

6.34 (0.59) 5.90 (0.58)

0.29

0.42 0.35

* Scores were derived from sensory preference ratings on a 10-cm unstructured hedonic scale scored as 0 = "dislike extremely," 10 = "like extremely." +Total sample size was 52; using a random, incomplete block design, each subject evaluated 4 of the 6 foods that differed in fat content and 4 of the 10 remaining foods. * Sensory preference evaluations were done during the third trimester of pregnancy and 11-13 months postpartum. § F statistic used to test for differences in mean hedonic scores between food pairs. I F statistic used to test for differences between mean hedonic scores during pregnancy versus postpartum.

lower-fat p r o d u c t s s u c h as f r o z e n y o g u r t m a y be e q u a l l y acceptable to adolescents, while certain o t h e r l o w fat p r o d u c t s are n o t liked. For e x a m p l e , cons u m p t i o n of milk is especially i m p o r t a n t d u r i n g a d o l e s c e n c e a n d p r e g n a n c y b e c a u s e of its n u t r i e n t d e n s e content, especially as the best s o u r c e of certain n u t r i e n t s p e r u n i t of s e r v i n g (i.e., calcium, v i t a m i n s A a n d D, a n d protein). In a d d i t i o n , the lactic acid f o r m e d f r o m lactose e n h a n c e s the a b s o r p t i o n of c a l c i u m f r o m milk. S o m e p r e g n a n t a d o l e s c e n t s m a y r e c o g n i z e that the a d d i t i o n of 300 calories d u r i n g p r e g n a n c y is n e e d e d , a n d t h e y can increase their e n e r g y intake b y c o n s u m i n g m o r e milk. H o w e v e r , others m a y fear " g e t t i n g fat" a n d look for lower-fat a n d lower-calorie f o o d s d u r i n g p r e g n a n c y . Skim milk is o f t e n r e c o m m e n d e d b e c a u s e of its n u t r i e n t c o n t e n t a n d l o w - f a t c o m p o s i t i o n . H o w e v e r , the results of this s t u d y s u g g e s t that for s o m e adolescents, the r e c o m m e n d a tion of s k i m milk m a y lead to a v o i d a n c e of the p r o d u c t or to d e c r e a s e d milk intake d u r i n g p r e g n a n c y , a l t h o u g h f r e q u e n c y of intake w a s n o t assessed in this s t u d y . O t h e r c a l c i u m sources, s u c h as spinach, cottage cheese, or tofu, p r o v i d e less c a l c i u m p e r s e r v i n g t h a n

milk, a n d in a d d i t i o n d o n o t c o n t a i n v i t a m i n D, w h i c h e n h a n c e s c a l c i u m utilization. A l t e r n a t e calc i u m s o u r c e s in u s u a l a m o u n t s c o n s u m e d w o u l d n o t m e e t the 1200-1500 m g of c a l c i u m r e c o m m e n d e d in the N a t i o n a l Institutes of H e a l t h C o n s e n s u s Statem e n t (31). A n a d d i t i o n a l s e q u e l a e of p r e g n a n c y is the o b l i g a t o r y excretion of c a l c i u m t h a t s h o u l d be cons i d e r e d in t e e n a g e p r e g n a n c i e s . F o o d c o n s u m p t i o n d a t a (32) indicate that a d o l e s c e n t f e m a l e s t e n d to h a v e diets h i g h in p r o t e i n a n d s o d i u m , f o o d c o m p o n e n t s that n e g a t i v e l y influence c a l c i u m utilization. Thus, nutritionally, the p r e f e r e n c e for w h o l e milk d u r i n g p r e g n a n c y is a p o s i t i v e f i n d i n g that can p o t e n t i a t e calorie a n d n u t r i e n t intake. The increased preference for milk (both w h o l e a n d skim) d u r i n g p r e g n a n c y c o m p a r e d to the p o s t p a r t u m p e r i o d could reflect taste alterations d u r i n g p r e g n a n c y , as f o u n d in adult p r e g n a n t w o m e n . The possibility also exists that p r e g n a n t girls in this s t u d y rated milk higher d u r i n g p r e g n a n c y because they k n e w , b a s e d on materials a n d information received in clinics, that while p r e g n a n t they w e r e s u p p o s e d to increase their intake of milk a n d milk p r o d u c t s as sources of calcium. P r e g n a n t a d o l e s c e n t s ' s t r o n g p r e f e r e n c e s for salted chips, crackers, a n d n u t s c o u l d reflect a p h y s -

48

SKINNERET AL.

iological need. That is, Brown and Toma (4) posited that the physiological need for sodium may increase during pregnancy, and the increase in sodium and fluid retention during pregnancy is well documented. In this study, preferences for salty foods over less salty foods were more intense during pregnancy. Current recommendations for sodium intake during pregnancy do not support a sodium restriction in healthy pregnant females (33). This study supports the finding of alteration of taste for salt during pregnancy previously reported in adults (14). If salty foods are selected on a consistent basis and as a major portion of the calories per day, then the satiety value may lead the adolescent to exclude or avoid cereal grains, fruits and vegetables, and other nutrient-dense foods that provide essential nutrients for pregnancy. However, counseling a pregnant adolescent to exclude these food products runs counter to alterations in taste for salt and, possibly, to physiological needs for sodium. Using moderation in daily intake of these foods would probably satisfy the altered tastes and add calories to the diet. Although no empirical assessment was made, it is possible that the sodium in milk also enhanced the preference for milk as reported in food comparisons done during and after pregnancy. With regard to fat intake during pregnancy, the clinician needs to take an individualized approach with pregnant teens. The importance of an adequate energy intake and an adequate weight gain is well documented (33). However, in the face of concern with weight gain during pregnancy, adolescents may need encouragement that some dietary fat is necessary and desirable to meet energy and nutrient needs. Conversely, for adolescents whose fat intake or weight intake is excessive, the clinician needs to explore the sources of fat and total calories in the diet. Using an example from this study, if the adolescent prefers whole milk over skim or low fat, the clinician may recommend restriction of other highfat foods (e.g., ice cream) and substitution of frozen yogurt or ice milk rather than focusing on whole milk, which also provides nutrients essential for pregnancy. Moreover, the clinician should recognize that the ideal percentage of calories from dietary fat during pregnancy is not known. The same approach should be used for sugarcontaining foods; individualized recommendations are needed. Sugar and fat are undesirable dietary components when they preclude dietary adequacy from essential nutrients a n d / o r contribute to excessive calorie intake. The altered sensory preferences during pregnancy

JOURNAL OF ADOLESCENTHEALTHVol. 22, No. 1

found in this study have several implications. Use of commonly eaten foods in the sensory evaluations allowed individuals to reveal their current sensory preferences without introducing products that were unfamiliar or were not relevant to adolescents' lifestyle. Results support the notion that adolescents may alter their food consumption patterns because of taste changes during pregnancy and that these alterations can be used in counseling. The sensory alterations can serve as enhancers for some adolescents to consume more of a desirable food, to try new foods during pregnancy, and to modify the amount of other foods that are less nutrient dense. Counseling about foods consumed during pregnancy often emphasizes the nutrients needed and the welfare of the fetus. An appeal to change food patterns that is based on sensory changes of the mother and is beneficial to the baby may achieve some modicum of dietary changes not achieved by the more conventional approaches. Because individuals are more likely to eat foods that taste good and are less likely to consume those that do not, nutrition recommendations that account for pregnancy-related taste alterations as well as individual taste preferences are more likely to be followed than recommendations based only on what is needed. The primary limitation of this study was the restricted number of foods evaluated by the adolescents. Results should not be extrapolated to other foods that were not evaluated. Another limitation is the potential for regional a n d / o r racial bias, because the sample included only white girls living in a southern state. Because whites are less prone to lactose intolerance than other races, preferences for milk and other dairy products most likely would differ among other racial groups. Although none of the foods evaluated in this study were regional favorites or specialties, regional differences in food patterns do exist. Additional research should address these limitations. Another limitation was that testing time in relation to meals was not formally standardized. However, most pregnant girls were still in school, and thus interviews were performed after school hours. For interviews in the postpartum period, girls were asked to restrict food and fluid consumption for 2 h prior to the interview for body composition assessment by bioelectrical impedance. Thus, at neither interview were the sensory assessments done immediately after a meal, although the timing was not standardized. This study contributes to the literature about pregnant adolescents and altered sensory prefer-

January 1998

PREGNANT ADOLESCENTS' SENSORY PREFERENCES

49

ences using foods rather than solutions of sweet, sour, bitter, and salt. In addition, these sensory changes are reflected in better nutrient intakes, as shown in dietary data of this group published elsewhere (23). Future research should focus on sensory testing of additional foods that are consumed by adolescents and those that contribute in a major way to meeting the nutrient allowances for pregnancy and the future health status of postpartum adolescents.

14. Brown JE, Toma RB. Taste changes during pregnancy. Am J Clin Nutr 1986;43:414-8.

18. Rozin P, Volemecke TA. Foods likes and dislikes. Ann Rev Nutr 1986;6:433-56.

This research was supported by the Tennessee Agricultural Experiment Station, Project AES-TN 860.

19. Dipple RL, Elias JW. Preferences for sweet in relationship to use of oral contraceptives and pregnancy. Horrn Behav 1980; 14:1-6.

15. Takahashi K, Yanomine K, Hamada H. Taste sensibility in pregnant women for the four basic tastes. Chem Senses 1986;11:277. 16. Dickens G, Trethowan WH. Cravings and aversions during pregnancy. J Psychosom Res 1971;15:259-68. 17. Pilgrim FJ. The components of food acceptance and their measurement. Am J Clin Nutr 1957;5:171-5.

20. Wade GN, Zucker I. Hormonal and developmental influences on rat saccharin preferences. J Comp Phys 1969;60:291-300.

References 1. Elster AB, McAnarney ER. Medical and psychosocial risks of pregnancy and childbearing during adolescence. Pediatr Ann 1980;3:89-94. 2. American Dietetic Association. Position of the American Dietetic Association: Nutrition management of adolescent pregnancy. J Am Diet Assoc 1989;89:104-9. 3. Mulchahey KM. Adolescent pregnancy: prevalence, health and psychosocial risks. In: Story M, ed. Washington, D.C.: Nutrition Management of the Pregnant Adolescent. National Clearing House, 1990:1-7. 4. Schneck ME, Sideras KS, Fox RA, et al. Low-income pregnant adolescents and their infants: Dietary findings and health outcomes. J Am Diet Assoc 1990;90:555-8. 5. Frisancho AR, Matos J, Leonard WR, et al. Developmental and nutritional determinants of pregnancy outcome among teenagers. Am J Phys Anthropol 1985;66:247-61. 6. Stevens-Simon C, McAnarney ER. Adolescent maternal weight gain and low birth weight: A rnultifactorial model. Am J Clin Nutr 1988;47:948-53. 7. Skinner JD, Carruth BR. Dietary quality of pregnant and nonpregnant adolescents. J Am Diet Assoc 1991;91:718-20. 8. McCoy H, Moak S, Kenney MA, et al. Snacking patterns and nutrient density of snacks consumed by southern girls. J Nutr Educ 1986;18:61-6. 9. Ezell JM, Skinner JD, Penfield MP. Appalachian adolescents' snack patterns: Morning, afternoon, and evening snacks. J Am Diet Assoc 1985;85:1450-4. 10. Story M. Eating behaviors and nutritional implications. In: Story M, ed. Nutrition Management of the Pregnant Adolescent, Washington, D.C.: National Clearing House, 1990:29-35. 11. Schwab RB, Axelson ML. Dietary changes of pregnant women: Compulsions and modifications. Ecol Food Nutr 1984;14: 143-53. 12. Tierson FD, Olsen CL, Hook EB. Influence of cravings and aversions on diet in pregnancy. Ecol Food Nutr 1985;17:11729. 13. Worthington-Roberts B, Little RE, Lambert MD, et al. Dietary cravings and aversions in the postpartum period. J Am Diet Assoc 1989;89:647-51.

21. Pike RL, Yao C. Increased sodium chloride appetite during pregnancy in the rat. J Nutr 1971;101:169-76. 22. Skinner JD, Carruth BR, Ezell JM, Shaw A. How and what do pregnant adolescents want to learn about nutrition? J Nutr Educ 1996;28:266-71. 23. Skinner JD, Carruth BR, Pope J, et al. Food and nutrient intake of white, pregnant adolescents. J Am Diet Assoc 1992;92: 1127-9. 24. Pope JF, Skinner JD, Carruth BR. Cravings and aversions of pregnant adolescents. J Am Diet Assoc 1992;92:1479-82. 25. Goldberg D, Carruth BR, Skinner J. Television viewing and dietary intake of pregnant adolescents. Nutr Res 1993;13:62132. 26. Campbell AM, Penfield MP, Griswold RM. Evaluating foods by sensory methods. In: The Experimental Study of Food, 2nd ed. Boston, MA: Houghton Mifflin Company, 1979:433-50. 27. SAS Institute, Inc. SAS User's Guide: Basics. Cary, NC: SAS Institute Inc., 1988. 28. Hollingshead AB. Four Factor Index of Social Status. New Haven, CT: Yale University Press, 1976. 29. Stroupe WC. Why mixed models? In: Applications of Mixed Models in Agriculture and Related Disciplines. Southern Cooperative Ser. Bull. No. 343. Baton Rouge, LA: Louisiana Agricultural Experiment Station, 1989:1-8. 30. General Linear Mixed Models GLMM. Users' Manual, Ver. 1.0. Department of Experimental Statistics, Louisiana State University and Agricultural and Mechanical College, Baton Rouge, LA: Louisiana Agricultural Experiment Station, 1990: 1-38. 31. Optimal Calcium Intake. National Institutes of Health Consensus Statement 1994. June 6-8;12:1-31. 32. Tippett KS, Mickle SJ, Goldman JD, et al. Food and Nutrient Intakes by Individuals in the United States, 1 Day, 1989-91. United States Department of Agriculture. Agricultural Research Service, NFS, Rep. No. 91-2, September 1995. 33. Food and Nutrition Board, Institute of Medicine, National Academy of Sciences. Nutrition During Pregnancy. Washington, D.C.: National Academy Press, 1990.