Customize your JAMA Network experience by selecting one or more topics from the list below.
To examine whether adolescent vegetarians were more likely than nonvegetarian peers to meet the dietary recommendations of the Healthy People 2010 objectives and to examine differences in other nutrients between these 2 groups.
A total of 4746 adolescents from 31 middle and high schools in the Twin Cities area of Minnesota. Data were collected via self-report surveys, with a student response rate of 81.5%.
Urban secondary schools.
Participants were equally divided by sex. The mean age was 14.9 years; 34.3% were in junior high school and 65.7% in high school. The racial/ethnic distribution was 48.5% white, 19.0% African American, 19.2% Asian American, 5.8% Hispanic, 3.5% American Indian, and 3.9% mixed or other.
Main Outcome Measures
Questions on vegetarian status and whether particular foods (eggs, dairy foods, chicken, fish) were excluded. Dietary intake was assessed using the Youth and Adolescent Food Frequency Questionnaire.
Vegetarian adolescents were significantly more likely than nonvegetarian adolescents to meet the Healthy People 2010 objectives. This was particularly noteworthy for total fat (70% vs 48%), saturated fat (65% vs 39%), daily servings of vegetables (26% vs 14%), and 5 or more servings of fruits and vegetables (39% vs 28%). Vegetarians were also less likely to eat fast food or drink regular soda and fruit drinks. Vegetarians consumed less vitamin B12, more diet soda, more caffeine, and more iron.
Adolescent vegetarians have a dietary pattern that is more likely than nonvegetarians to meet the Healthy People 2010 objectives.
VEGETARIANS, for a variety of reasons, choose to limit their intake of meat and other animal products.1 Some vegetarians eat no animal products (vegans), whereas others include milk products (lactovegetarians) or eggs (ovovegetarians). Many self-identified vegetarians also eat some meat, particularly fish and chicken (semivegetarians).1 Adult vegetarianism has been studied for the relationships among a plant-based diet, chronic diseases, and length of life.2 Overall, adult vegetarians tend to live significantly longer lives than nonvegetarians, with substantially lower rates of cardiovascular diseases, hypertension, type 2 diabetes mellitus, obesity, and some types of cancer.3-7
Vegetarian diets, like all diets, need to be planned appropriately to be nutritionally adequate. Of concern with adult vegetarians is sufficient consumption of nutrients often found in animal products.8 Nutrients of most concern include iron, calcium, zinc, and vitamin B12.7-9 White and Frank1 point out that clinical nutrition deficiencies are uncommon even among vegans, although they suggest that careful planning and, in some cases, vitamin supplementation may be warranted. The intake of these nutrients has not been examined in a large adolescent vegetarian population.
Given that dietary patterns are adopted and appear to be maintained during adolescence and young adulthood,10-12 an interesting question is whether younger vegetarians report healthier dietary patterns than nonvegetarian counterparts. Specifically, what percentage of adolescent vegetarians (vs nonvegetarians) comply with the dietary recommendations outlined in the Healthy People 2010 objectives? Do adolescent vegetarians consume less fat and saturated fat, consume more fruits and vegetables and fiber, and maintain healthier intake levels of other nutrients, such as calcium, than do adolescent nonvegetarians? Or, alternatively, are adolescent vegetarians at increased risk for inadequate intake of any of these foods or nutrients? These questions are important since dietary behaviors have implications for adolescents' current and future health.11,13-16 In a prior study, Neumark-Sztainer et al17 found that adolescent vegetarians were twice as likely to consume fruits and vegetables, one third as likely to consume sweets, less likely to consume dairy products, and one fourth as likely to consume salty snack foods compared with nonvegetarians. Donovan and Gibson18 found that mean daily intakes of energy and most nutrients were comparable for vegetarians and nonvegetarian adolescents. Adolescent vegetarians in their study, however, consumed fewer dairy products, meat, and sweets and more legumes, nuts, and vegetables than did nonvegetarians.18 The studies with adolescent vegetarians, however, have been limited in their assessment of diet, had small samples of vegetarians, and have not included a multiracial, multiethnic group of subjects.
Data for the current study were drawn from Project EAT (Eating Among Teens), a comprehensive study of adolescent eating patterns and weight concerns.19 A large group of urban adolescents from multiple ethnic groups in this study provided the opportunity to examine the eating patterns and nutrient intake of a diverse population of adolescents. We hypothesized that adolescents who reported to be vegetarians, compared with nonvegetarian peers, would be more likely to meet the dietary recommendations of the Healthy People 2010 objectives but might eat less of some nutrients, particularly calcium, iron, vitamin B6, and possibly vitamin B12. Finally, we hypothesized that more stringent vegetarians, those who did not eat fish and chicken, would be more likely to meet the dietary recommendations of Healthy People 2010 related to fat, fruits, and vegetables.
The study population included 4746 adolescents from 31 public middle schools and high schools in the Twin Cities area of Minnesota. Participants were equally divided by sex (50.2% male, 49.8% female). The mean age of the study population was 14.9 years (range, 11-18 years); 34.3% were in junior high school and 65.7% in high school. The racial/ethnic backgrounds of the participants were as follows: 48.5% white, 19.0% African American, 19.2% Asian American, 5.8% Hispanic, 3.5% American Indian, and 3.9% mixed or other. Most of the Asian American population was from Southeast Asia.
Data for this study were collected via self-report surveys and anthropometric assessments in the 1998-1999 school year. The data were collected in schools within health, physical education, and science classrooms in one 90-minute period or two 50-minute periods. Trained research staff administered the surveys in the classrooms and measured height and weight in a private area. Study procedures were approved by the University of Minnesota Human Subjects' Committee and by research boards of the participating school districts. Consent procedures also followed the requirements of the participating school districts. In some schools, passive consent procedures were used, whereas in others active consent procedures were required. The response rate for student participation was 81.5%. The main reasons for lack of participation were absenteeism and failure to return consent forms within schools requiring active consent.
To identify vegetarians, all students were asked on the survey to respond yes or no to the question, "Are you a vegetarian?" Those who answered yes were asked to respond to additional questions. The first was "As a vegetarian, do you eat any of the following? (1) Eggs; (2) Dairy food (such as milk, cheese); (3) Chicken; (4) Fish." Students who checked no to chicken and fish were labeled lacto-ovo-vegans. Students who checked yes to either chicken or fish were labeled semivegetarians.
Demographic factors, including sex, ethnicity/race, school level, and socioeconomic status (SES), were based on self-report. Ethnicity/race was assessed with the question, "Do you think of yourself as . . . (1) White; (2) Black or African American; (3) Hispanic or Latino; (4) Asian American; (5) Hawaiian or Pacific Islander; or (6) American Indian or Native American?" Youth were given the option of choosing multiple responses, and those reporting more than one response (other than white) were coded as mixed or other. School level was divided into middle school (7th through 8th grades) and high school (9th through 12th grades). The prime determinant of SES was parental educational level, defined by the higher level of either parent. Response categories for questions on parental educational level were as follows: (1) did not finish high school; (2) finished high school or received general equivalency diploma; (3) some college; (4) finished college; (5) master's or PhD degree; and (6) don't know. Other variables used to assess SES included the following: family eligibility for public assistance (yes, no, or don't know), eligibility for free or reduced-cost school meals (yes, no, or don't know), and employment status of mother and father (full-time, part-time, networking, or don't know). An algorithm was developed to avoid classifying adolescents as having high SES, based on parental education levels, if they were receiving public assistance, eligible for free or reduced-cost school meals, or had 2 unemployed parents (or 1 unemployed parent if from a single-parent household). Students were classified as having low SES, middle SES, or high SES.
Assessment of dietary intake was performed with the 149-item Youth and Adolescent Food Frequency Questionnaire (YAQ). The YAQ asks about specific food items, identifies amounts of each food, and asks for frequency of consumption. For example, one question asks about how frequently 2 pieces of pizza are consumed (never, 1-3 times per month, once a week, 2-4 times per week, ≥5 times per week). Validity and reliability of the YAQ have been tested among a random sample of children (aged 9-18 years) of participants in the Nurse's Health Study and found to be within acceptable ranges for dietary assessment tools.20,21 Mean correlation for energy-adjusted nutrients between two YAQs and three 24-hour recalls (implemented in 3 seasons) was 0.45. The mean energy intake (in calories) for the YAQ was higher than for the recalls but within 1% of them. Test-retest correlations between 2 YAQs during a 1-year period were 0.49 for fruit and 0.48 for vegetables. Responses to questions on the frequency of intake of the following fruits and vegetables were summed to assess average total daily intake of fruits and vegetables: apples, apple juice, bananas, beets, broccoli, carrots, celery, coleslaw, corn, grapes, greens or kale, lettuce, melon, mixed vegetables, oranges, orange juice, peaches or plums, pears, peas, peppers, potatoes (not including French fries), raisins, spinach, strawberries, string beans, tomatoes, tomato or spaghetti sauce, yams, and zucchini or squash. Consumption of fruits and vegetables was generally reported using 5-point scales (eg, never or less than once a month, 1-3 times per month, once a week, 2-4 times per week, ≥5 times per week).
Dietary intake in the present study was evaluated in 2 ways: relative to the Healthy People 2010 objectives (for relevant nutrients and foods) and as mean daily intake. In the current study, nutrient and food behaviors examined relative to the Healthy People 2010 objectives included the following: total fat (percentage of total energy), saturated fat (percentage of total energy), calcium (milligrams), fruits (servings), vegetables (servings), and grains (servings). These were converted to the percentage of students who complied with the Healthy People 2010 objectives. Other daily nutrient intakes included the following: calories (kilocalories), protein (grams), calcium (milligrams), zinc (milligrams), iron (milligrams), vitamin A (units), vitamin B6 (milligrams), vitamin B12 (micrograms), vitamin C (milligrams), folate (micrograms), linoleic acid (grams), cholesterol (milligrams), caffeine (milligrams), and fiber (grams). Other daily food intakes included soda (daily servings), diet soda (daily servings), fruit drink (noncarbonated, daily servings), and fast food (frequency in the past week). We also examined the percentage of calories from protein, carbohydrates, total fat, saturated fat, polyunsaturated fat, and monounsaturated fat, as well as the average daily servings of fruit and vegetables. These were selected to assess whether vegetarians were at greater or lesser risk than nonvegetarians for nutrient inadequacies of particular concern during adolescence.
The present study assessed dietary intake associated with adolescent vegetarians. In the first set of analyses, the percentages of adolescents who met the Healthy People 2010 dietary objectives were compared between vegetarians and nonvegetarians. These analyses also assessed whether dietary intake was moderated by sex (ie, sex-by-vegetarian interactions) or race/ethnicity (ie, race/ethnicity–by–vegetarian interactions). Daily intake of specific nutrients and foods was also compared between vegetarians and nonvegetarians. In the second set of analyses, the percentages of adolescents who met Healthy People 2010 objectives were compared between specific vegetarian groups. For these analyses, vegetarians who were lacto-ovo-vegans were compared with semivegetarians. Logistic and linear regression analyses were conducted to test for statistical differences between the comparison groups when the dependent variables were either dichotomous or linear, respectively. All analyses controlled for sex and race since there were differences between groups in sex and racial make-up. The interaction analyses by sex and race were controlled for race or sex, respectively. P<.05 was considered significant. P values were not adjusted for multiple testing because false-positive differences were unlikely when the P values are small (<.01), which was the case for most of the comparisons. All analyses were conducted using the Statistical Analysis System.22
From the total sample of students, 4521 adolescents had completed the YAQ and had usable data, and 262 (5.8%) reported being vegetarian. Among the vegetarians, nearly three fourths (73.7%) were female. Nearly half (47.5%) of the vegetarians were white, 26.8% were Asian, 11.1% were African American, 5.8% were Hispanic, 5.0% were American Indian, and the remaining 4% were Hawaiian or Pacific Islander or other. Because there were so few Hispanic, American Indian, and mixed or other vegetarians (because of the sample sizes of those race/ethnicities in the entire sample), these groups were combined and referred to as other in the race/ethnicity analyses. Among the vegetarians, 38.6% were in high school (vs middle or junior high school). More than half of the vegetarians (62%) reported eating chicken and/or fish and were semivegetarians; 32% ate eggs and/or dairy products (but not chicken or fish), and 6% were vegans, so that 38% were classified as lacto-ovo-vegans for these analyses. As given in Table 1, there were significant differences between vegetarians and nonvegetarians in the sex and racial makeup of the samples.
Overall, adolescent vegetarians were significantly more likely to meet the dietary recommendations of Healthy People 2010 as given in Table 2. Vegetarian adolescents were more than twice as likely to eat less than 30% of their calories from fat and nearly 3 times more likely to eat less than 10% of their calories from saturated fat. They were also 1.4 to 2 times more likely to eat 2 or more servings of fruit, 3 or more servings of vegetables, 3 or more servings of vegetables including 1 that is dark yellow or green, and 5 or more servings of fruits and vegetables daily. There were no significant differences in daily calcium intake or servings of grains. There were also no significant sex–by–vegetarian status or race–by–vegetarian status interactions.
Differences between vegetarians and nonvegetarians for a variety of other food intake variables are given in Table 3. Vegetarian adolescents, compared with nonvegetarian adolescents, consumed significantly less vitamin B12, cholesterol, regular soda, fruit drink, and fast food, as well as the percentage of calories from protein and all types of fat. Vegetarians were significantly more likely than nonvegetarians to consume more iron, vitamin A, folate, caffeine, fiber, and diet soda, as well as a greater percentage of calories from carbohydrates and more servings of fruit and vegetables. There were nonsignificant differences between the 2 groups for calories, protein, calcium, zinc, vitamin B6, vitamin C, and linoleic acid.
Among the vegetarians, differences between the lacto-ovo-vegans and semivegetarians (who ate chicken and/or fish) were examined. The results for the percentage of each group who met the dietary guidelines of the Healthy People 2010 objectives are given in Table 4. Lacto-ovo-vegans were more than twice as likely as semivegetarians to eat less than 30% of the calories in their diet from fat, more than 2 servings of fruit per day, and 5 or more servings of fruits and vegetables per day.
Comparing lacto-ovo-vegans with semivegetarians, there were no interactions between adherence to the Healthy People 2010 objectives by sex, but there were 5 significant interactions by race. White semivegetarians were significantly less likely than white lacto-ovo-vegans to meet the daily requirement for calcium (22.6% vs 43.9%; odds ratio [OR], 0.37; 95% confidence interval [CI], 0.16-0.82), to eat more than 2 servings of fruit each day (34.5% vs 68.6%; OR, 0.24; 95% CI, 0.11-0.54), to eat more than 3 servings of vegetables each day (8.6% vs 35.3%; OR, 0.17; 95% CI, 0.06-0.49), to eat more than 3 vegetables each day, including 1 deep yellow or green vegetable (6.2% vs 22.8%; OR, 0.22; 95% CI, 0.07-0.74), and to eat 6 or more servings of grains each day (30.0% vs 53.7%; OR, 0.36; 95% CI, 0.17-0.79). There were no significant differences between lacto-ovo-vegans and semivegetarians on these items among African Americans, Asians, or others.
Overall, vegetarian adolescents were much more likely than nonvegetarians to meet the Healthy People 2010 objectives. The percentage of vegetarians vs nonvegetarians who met the objectives was particularly noteworthy for fat consumption (70% vs 48%), saturated fat consumption (65% vs 39%), daily servings of vegetables (26% vs 14%), and 5 or more daily servings of fruits and vegetables (39% vs 28%). These are substantive differences in the major targeted dietary behaviors for cardiovascular disease (fat intake) and cancer (fruit and vegetable intake) and suggest that vegetarian adolescents, similar to their adult counterparts, have dietary patterns that, if maintained, could significantly lower their risk of the leading causes of death as adults.23-26
The results from the present study on dietary fat were corroborated by comparing vegetarians' and nonvegetarians' average daily intake of fat. On average, 26% of the vegetarians' total calories were from fat, compared with 30% of nonvegetarians' calories, a 13% reduction in fat consumption. More notably,27 vegetarians' intake of saturated fat and cholesterol was 14% and 22% lower, respectively, than nonvegetarians.
Similar results were found for fruit and vegetable intake. Vegetarians, compared with nonvegetarians, consumed 0.4 more servings of fruit and 0.5 more servings of vegetables daily. In fact, the lacto-ovo-vegans consumed 5.8 servings of fruits and vegetables per day. This can be compared with 4.1 servings for nonvegetarians and 4.7 servings for semivegetarians. Only the lacto-ovo-vegans, those who did not eat chicken and/or fish, consumed, on average, more than the recommended 5 servings of fruits and vegetables each day.
The magnitude of the differences between vegetarians and nonvegetarians for fat consumption and fruit and vegetable intake was substantial. For example, the differences between vegetarians and nonvegetarians were greater for fat consumption and vegetable intake than differences between male and female adolescents, those with low SES and high SES, and white and black populations, as reported in a prior study19 with this same population. Although these are not ideal comparisons, they do provide population groups that suggest how sizable the differences were between vegetarian and nonvegetarian dietary patterns.
There were positive and negative aspects of the adolescent vegetarian dietary patterns when examining other aspects of their diet. For example, adolescent vegetarians were less likely than nonvegetarian peers to eat fast food or drink soda and fruit drinks and more likely to consume more carbohydrates, iron, fiber, and vitamin A. These dietary factors add to the picture of a healthier overall dietary pattern for the vegetarians and are consistent with the data on fat, fruits, and vegetables. It is particularly important that iron intake was higher among vegetarians since the absorption of iron from vegetable sources is generally lower than from meat sources.28 However, vegetarian adolescents, compared with nonvegetarians, consumed less vitamin B12, which may be important for future health. Vegetarians also consumed marginally less protein. They consumed, on average, 70 g of protein (62 g for the lacto-ovo-vegans), which is sufficient for growth and development.29 The recommended daily allowance for protein for 15- to 18-year-olds is 44 g/d for girls and 59 g/d for boys.28 Although adequate protein is essential for growth, plant sources of protein alone can provide adequate amounts of essential amino acids if a variety of plant foods are consumed and energy needs are met.7
Most vegetarian (70%) and nonvegetarian (65%) adolescents did not meet the recommended intake (1300 mg/d) for calcium. There were no significant differences between vegetarian and nonvegetarian calcium intakes or between lacto-ovo-vegans' and semivegetarians' calcium intakes. Other studies have found that lacto-ovo-vegetarians have calcium intakes that are comparable to or higher than those of nonvegetarians.30,31 To meet calcium requirements, adolescent vegetarians and nonvegetarians should consume calcium-rich or calcium-fortified foods, or if they are unwilling or unable to do this, they should take supplemental calcium.
There were few differences between lacto-ovo-vegans and semivegetarians. As hypothesized, the lacto-ovo-vegans were more likely to meet the Healthy People 2010 objectives for total fat, fruit, and 5 or more daily servings of fruits and vegetables compared with semivegetarians.
It is important to address substantive differences between vegetarians and nonvegetarians that might account for the dietary differences observed in this study. These demographic and psychosocial differences were the focus of a prior publication with this same population of students.32 The adolescent vegetarians were more likely than nonvegetarians to be female, not African American, in middle school (rather than high school), weight and body conscious, dissatisfied with their bodies, and involved in a variety of healthy and unhealthy weight control behaviors. In addition, vegetarians more often reported having been told by a physician that they had an eating disorder and were more likely to have contemplated and attempted suicide. Other behaviors, such as substance use and physical activity, were not different between vegetarians and nonvegetarians. Thus, the adolescent vegetarians differ from the nonvegetarians in being female (although it is noteworthy that the relationship between vegetarian status and psychosocial factors did not differ by sex) and in their focus on weight and diet and eating behaviors; this focus on weight and diet seems to be the primary motivation for adopting a vegetarian diet, rather than its long-term health benefits.17,32,33 Health care practitioners should be aware of adolescents who greatly limit food choices and who exhibit symptoms of eating disorders. However, data suggest that vegetarian diets do not lead to eating disorders.7 Adolescents who adopt vegetarian eating habits without adequate nutrition knowledge and with a desire to achieve weight loss are most vulnerable to malnutrition and growth failure34 and should receive added attention. Thus, although the adoption of a vegetarian-eating pattern may have benefits in terms of nutritional intake for many adolescents, for others it may be a method of unhealthy food restriction. Therefore, adolescents who choose to become vegetarians should be monitored for adequate intake and questioned about their motivations by a health care practitioner.
This study had several strengths and limitations. The questions on vegetarianism and dietary factors were more comprehensive than prior population-based studies. We chose a representative sample of adolescents from the major urban school districts in Minnesota and thus had sufficient data to examine ethnic group and sex differences. The study was limited by the self-reporting of vegetarianism and the cross-sectional design of the study. Furthermore, although the YAQ has been found to have acceptable reliability and validity,20,21 as with any dietary assessment tool, it has limitations. Specifically, it is limited by its reliance on adolescents' recall, the lack of specificity of portion sizes, and the inability to assess nutrient content of nonstandard mixed dishes (such as homemade spinach lasagna). More work is needed on a dietary measure than can be self-administered in a school setting and yet can also provide more reliable and valid data.
In the current study, adolescent vegetarians were more likely than nonvegetarians to be adhering to the Healthy People 2010 objectives relevant to nutritional health. This was even more the case among the lacto-ovo-vegans. Thus, it seems that rather than viewing adolescent vegetarianism as a difficult phase or fad, the dietary pattern could be viewed as a healthy alternative to the traditional American meat-based diet. With careful planning, using the vegetarian food guide pyramid as a guide,35 vegetarian adolescents could learn proper nutritional patterns and practices that could lead to a lifelong dietary practice that might be salutary for themselves and their families in the future.
Accepted for publication January 1, 2002.
This study was supported by grant MCJ-270834 (Dr Neumark-Sztainer) from the Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Service Administration, US Department of Health and Human Services, Washington, DC.
Few adolescents meet the dietary guidelines of the Healthy People 2010 objectives. Prior research has indicated that adolescent vegetarians may eat more fruits and vegetables, fewer sweets, and fewer salty snack foods than nonvegetarian peers. The previous studies of vegetarians were limited in their assessment of diet, small samples of adolescents, and lack of representation of multiple racial/ethnic groups. The present study examines the dietary patterns of adolescent vegetarians and compares them with nonvegetarian adolescents and the Healthy People 2010 objectives. The study is performed with a large, multiethnic population in an urban area in the Upper Midwest. Adolescent vegetarians were significantly more likely to meet the Healthy People 2010 objectives, especially for fat intake and fruits and vegetables. This suggests that vegetarianism may provide a healthy dietary pattern for adolescents, particularly if it is well planned to address potential deficiencies.
Corresponding author and reprints: Cheryl L. Perry, PhD, Division of Epidemiology, School of Public Health, University of Minnesota, Minneapolis, MN 55454 (e-mail: firstname.lastname@example.org).
Perry CL, McGuire MT, Neumark-Sztainer D, Story M. Adolescent Vegetarians: How Well Do Their Dietary Patterns Meet the Healthy People 2010 Objectives? Arch Pediatr Adolesc Med. 2002;156(5):431–437. doi:10.1001/archpedi.156.5.431
Create a personal account or sign in to: