Factors and common conditions associated with adolescent dietary supplement use: an analysis of the National Health and Nutrition Examination Survey (NHANES)

The 1994 Dietary Supplement Health and Education Act (DSHEA) defined dietary supplements (DS) as vitamin, mineral, herb/botanical, amino acid, enzyme, protein, probiotic, glandular, or hormone-like substances [1]. Dietary supplements are used by more than 50 percent of American adults [2]. Use in adults is most common in those with higher income and education, women, and those with chronic health conditions. Clinical, regional, and internet-based surveys of adolescents have shown higher DS use among athletes and teens with chronic illnesses, but these associations need to be confirmed in population-based national surveys [3–9]. In small clinical surveys, pediatric chronic conditions for which herbal therapies and dietary supplements were reported include: attention deficit hyperactivity disorder (ADHD), asthma, atopic dermatitis, allergic rhinitis, cancer, inflammatory bowel disease, headache, and cystic fibrosis [3, 10–18].

The 1988–94 National Health and Nutrition Survey (NHANES) found that approximately one fourth of adolescents used DS [19]. The 1999–2000 NHANES reported similar rates [20, 21]. However, these analyses did not explore self-reported clinical conditions associated with teens' use of DS. [20, 21]

We conducted this study to describe recent rates of DS use among US adolescents and factors associated with that use. Specifically, we wished to describe the prevalence of DS use among US adolescents age 11 to 19 and to analyze socio-demographic and clinical factors associated with DS use. Based on previous research, we hypothesized that DS use would be higher among: girls, older adolescents, non-Hispanic whites, those who diagnosed with common outpatient conditions, and those who reported using prescription medications.

We used data from the 1999–2000 and 2000–2001 National Health and Nutrition Examination Surveys (NHANES) [22]. NHANES is a series of nationally representative surveys of the civilian, non-institutionalized population in the United States administered by the National Center for Health Statistics. The NHANES data are collected through household interviews and physical examinations at mobile examination centers. The sampling technique follows a complex, stratified, multi-stage probability design that over-samples adolescents, African Americans, Mexican Americans, and persons of low-income. NHANES has been conducted as a two year data collection with data released for public use in two-year increments. In 1999–2000 the household interview response rate was 82 percent, and in 2001–2002 the household interview response rate was 84 percent. Adolescents younger than 16 years old had a proxy (family member or guardian) answer for them unless they were an independent minor. Those 16 years and older answered the questions for themselves or had a proxy.

DS use was assessed from responses to questions asked during the household interview: "Have you used or taken any vitamins, minerals, or other dietary supplements in the past month?" Those who said "yes" were asked to provide dose, frequency, and duration of use and were asked to show the supplement containers to the interviewer so that supplement details could be recorded. If the container was not available (about 25 percent of the time), participants were asked for the exact name of the product, or to provide information in as much detail as possible. Additionally, respondents were shown a card during the interview with examples of dietary supplements including herbs, vitamins, minerals, fiber supplements, amino acids, fish oils, sport supplements, memory and energy promoting products. The survey did not ask about medicinal herbal teas, medicinal foods (nutraceuticals), or folk remedies.

Age, sex, race/ethnicity, total family income, physical activity, and health status were generally obtained from the head of household. We dichotomized the age variable into two age groups 11–15 years old versus 16–19 years old. Race/ethnicity was classified as non-Hispanic black, non-Hispanic white, Mexican American, and all others. Health status was assessed with the question "would you say your health in general is... excellent, very good, good, fair, or poor?" Respondents were asked if they used any prescription medications in past month.

Finally, to address the question as to whether adolescents with common out-patient conditions use more dietary supplements, we extracted data on chronic non-life threatening conditions frequently seen in out-patient practices asked about during the NHANES interview. Questions were asked about asthma, being overweight, and ADHD with the phrasing, "Has the doctor or other health professional ever told you that (you/he/she has) X ?". The conditions "hay fever in past year" and "frequent severe headaches in past year" were based on self report, not physician diagnosis. Many potentially clinically relevant conditions such as acne, depression, cystic fibrosis, and inflammatory bowel disease were not included in the survey.

Overall, 27% of youth reported using one or more DS in the month prior to the survey. Non-vitamin/mineral DS were used by fewer than 5% of youth (Table 1). Older teens used more prenatal vitamins (p = 0.003) and iron (p = 0.004) than younger teens.

Similarly, in the adjusted model, some health factors remained associated with DS use while others did not. Compared with very good/excellent health, poor or fair health status was associated with a lower overall DS use (0.59 [0.40–0.88] p = 0.011). Conversely, adolescents reporting use of a prescription medication had higher overall rates of any DS (1.37 [1.10–1.72] p = 0.007) and vitamin/mineral supplements (1.48 [1.16–1.90] p = 0.003) use than those who did not use prescription medications.

After adjustment, the only specific condition that remained associated with a greater overall DS use was headache (1.25 [1.04–1.50] p = 0.018). Use of non-vitamin/mineral DS was more common among obese respondents (1.51 [1.07–2.13] p = 0.021).

This study describes the relationship between several varieties of DS use and clinical and socio-demographic factors in a nationally representative sample of American adolescents. Overall, 27% of adolescents reported using a DS in the last month. DS use was higher among non-Hispanic whites, those who reported better health status, and those who used prescription medications. However, we did not find the expected association between DS use and gender, nor after controlling for other factors, between DS use and any specific health condition other than headache. The findings that the most commonly used DS by teens were multivitamins (16%) and vitamin C (6%) is reassuring in terms of potential health risks. The higher use of non-vitamin/mineral supplements among obese teens raises questions about the types and safety of the specific DS used in this growing population.

The prevalence of DS use in our sample was consistent with that found in other large studies that used different sampling techniques [4, 19, 26]. The use of some non-vitamin/mineral supplements (e.g. sport and herbal performance enhancers) in this sample was lower than in studies of teen athletes [27–32]. It was also lower than rates reported in clinically based surveys [3, 11, 12, 14, 33]. Differences in rates might be due to different usage in different groups, differences in the time period of use, and by different survey methods such as self report using the Internet vs. parent or adolescent response in a face to face interview [4].

In the adjusted model, certain demographic factors were associated with supplement use, while others were not. We confirmed findings of another large survey in which older adolescents were more likely than younger teens to use non-vitamin/mineral DS such as herbals, sports and weight loss products [4]. Similarly, we confirmed earlier reports that non-Hispanic blacks and Mexican American adolescents use fewer vitamin/mineral DS than non-Hispanic white youth [4, 19]. On the other hand, unlike earlier studies, after adjusting for other factors, female gender was not significantly associated with DS use [4, 19]. Further studies are needed to understand how race, culture, income and health care disparities affect DS use and the impact of this use on health status.

As expected, certain clinical factors were also associated with DS use. For example, those reporting very good/excellent health status reported higher rates of DS use, confirming an earlier study [34]. The direction of this association (Does use of DS cause good health or are healthy teens more likely to use DS?) is a worthy question for future research.

As expected, we found a higher rate of DS use among teens who reported using prescription medications than those who did not, even after adjusting for potential covariates [23, 24]. The etiology and clinical implications of the relationship between DS and prescription medication use requires additional exploration. Given reports about clinically serious interactions between prescription drugs and herbals/supplements, physicians and pharmacists need to be mindful of the higher use of DS among prescription users, ask prescription medication users about their use of DS, and monitor them for adverse effects or interactions.

The analysis of DS use by specific health conditions suggested mixed relationships. For example, obese youth were more likely to use non-vitamin/mineral DS than their peers. The higher use among obese teens is likely due to the inclusion of weight loss products in the non-vitamin/mineral DS variable; this finding deserves more specific scrutiny in future research for two reasons: a) rates of obesity are rising; b) weight loss products such as those containing ephedra can present risks of serious toxicity or drug-DS interaction. Also, adolescents with headaches were more likely to use DS, confirming earlier reports [35, 36]. Future analyses are needed to look specifically at different types of DS with different disorders and the impact of DS use on disease-specific outcomes including use of professional health care, health status, quality of life, and ability to work or attend school.

Despite its large size, national scope, and extensive, detailed ascertainment of DS use, this analysis has several limitations. First, proxy information was used for respondents less than 16 years old. Proxies (e.g. parents) could have overestimated multivitamin use and under estimated use of products teens may have purchased on their own such as weight loss and performance-enhancing supplements. Second, questions about dietary supplements may have missed products considered home remedies, folk remedies, or parts of the diet, such as green tea or foods containing supplements. Furthermore, the limited data on specific health conditions reduced our ability to assess the complex relationships between diseases and DS use. Finally, since NHANES did not ask about reasons for supplement use or the impact of this use on health status, we do not know why teens used certain supplements or how it affected their health [37].

DS use is most common among non-Hispanic white youth and those taking prescription medications; use of non-vitamin/mineral DS is most common among older teens and those who are overweight. Therefore, physicians and pharmacists should regularly ask patients about DS use, and check for potential interactions. To better understand use among culturally diverse groups and those with different clinical conditions, future studies should include a broader range of DS (such as those used in folk remedies, foods and medicinal teas) and ask about common health conditions. Additional studies are needed to determine the impact of DS use on health care use, health status and quality of life.