Copyright 2004 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2004
Data on the prevalence and co-occurrence of multiple somatic symptoms among US adolescent females as they are influenced by sociodemographic, behavioral, and environmental factors is limited.
To describe the health status of adolescent US females measured by the prevalence, frequency, and co-occurrence of headache, stomachache, backache, and morning fatigue and to investigate associations between selected risk and protective factors.
Design, Setting, and Participants
School-based, cross-sectional, nationally representative survey of adolescents in the 6th through 10th grades in the US. Data collected between 1997 and 1998.
Main Outcome Measures
Prevalence of headache, stomachache, backache, and morning fatigue.
Among US adolescent girls, 29.1% experience headaches, 20.7% report stomachaches, 23.6% experience back pain, and 30.6% report morning fatigue at the rate of more than once a week. Co-occurrence of somatic complaints is common. Among girls who experienced headaches more than once a week, 3.2 million (53.3%) also reported stomach pain more than once a week and 4.1 million (74.3%) reported morning fatigue more than once a week. Heavy alcohol use, high caffeine intake, and smoking cigarettes every day were strongly associated with all symptoms, while parent and teacher support served as protective factors.
Somatic complaints of headache, stomachache, backache, and morning fatigue are common among US adolescent girls and co-occur often. Effective clinical treatment of this population requires comprehensive assessment of all female adolescents presenting with seemingly isolated somatic complaints.
Headache, stomachache, backache, and morning fatigue are frequently reported somatic complaints among children and adolescents.1- 7 Studies of this population in the United States, Israel, and Scandinavia show that, compared with boys, girls report higher prevalence of these symptoms.2- 9 Furthermore, at least 1 nonrepresentative US study and 1 Icelandic survey indicate that girls are at greater risk for experiencing symptoms concurrently.4,6 For all adolescents, the experience of chronic pain may have long-term effects and contribute to functional limitations, adversely affecting school attendance, development of healthy social relationships, and participation in a variety of developmental experiences.10- 16
Factors associated with somatic symptoms include the following: sociodemographic characteristics, social and institutional support, and health behaviors.2,3,17- 21 Social and institutional support measures may capture the protective role of peer, teacher, and parental support as well as risks associated with being bullied.3,19,20,22 Research has also shown potential biological links between some health behaviors and the 4 symptoms in question. For example, the relationships between headache, stomachache, and back pain and health behaviors such as caffeine consumption, heavy alcohol use, and smoking have been documented, but associations with other symptoms and behaviors need further investigation.3,19,23,24
The disproportionate rates of symptoms experienced by adolescent girls, in conjunction with the potentially detrimental effects of these symptoms on social development, including social phobia, separation anxiety,25 and social withdrawal14 suggests the need for a better understanding of the nature, extent, and possible causes for frequent somatic symptoms among this population. This investigation describes the health status of US adolescent females as measured by the prevalence, frequency, and co-occurrence of 4 somatic symptoms and explores associations with psychosocial and environmental factors.
Between 1997 and 1998, the National Institutes of Health, National Institute of Child Health and Human Development conducted a school-based, cross-sectional, nationally representative survey of adolescents in grades 6 through 10 as part of the multicountry Health Behavior of School-aged Children study. Study design and protocols are discussed elsewhere.26 Schools were selected to ensure nationally representative estimates for each grade and classes were selected from the target grade using simple random sampling. All students in a selected class were asked to participate with active parental and student consent. Seventeen thousand students responded, resulting in a participation rate of 83%. If records were missing for key variables (eg, age or sex) or had more than 75% missing responses, the records were dropped according to the study protocol requirements (n = 835). Records were excluded if the participant's grade was out of the target range, if the participant's age was outside of the 99th percentile for the grade (n = 440 students), or if either the participant's grade or age was unknown (n = 39). The resulting analytic sample was 15 686 students in grades 6 through 10, including 8370 girls. For the 4 dependent variables, headache, stomachache, backache, and morning fatigue, a range of 13 to 120 records were excluded for each variable owing to missing data resulting in a final analytic sample of 8250 girls, representing 10 360 601 US girls in grades 6 through 10.
The 4 dependent variables were measured by asking how often a student experienced any of the symptoms in the last 6 months: rarely or never, about once every month, about once every week, more than once a week, or about every day. The 5 categories were compressed to 4 levels for this analysis with the responses for more than once a week and about every day being combined; the resulting composite category was labeled "more than once a week." Social and institutional support was assessed using 3 composite variables: parent support, teacher support, and student support, and frequency of bullying experienced during the past school term. Health behavior measures included the frequency of alcohol use in the past 30 days, particularly binge drinking defined as 5 drinks or more consumed consecutively on more than 3 days in the past month; smoking; caffeine intake (including both coffee and soda); and physical activity as measured by participation in nonorganized sports. Sociodemographic characteristics included grade, race/ethnicity, and mother's educational attainment.
Statistical analyses were performed using SUDAAN and SPSS software.27,28 Principal components analysis and correlation matrices (including varimax rotations29 and scree plots30) were used to determine the selection of variables for inclusion in the composite parent, teacher, and student support variables. SUDAAN was used in all statistical analyses to obtain standard errors and 95% confidence intervals (CIs) that account for the complex multistage design. Methods of analysis included univariate descriptive analyses, bivariate comparisons, and multiple logistic regression controlling for all predictors and grade level, race/ethnicity, and mother's educational attainment. The significance of the bivariate relationships between the 4 somatic complaints and the predictors were tested using the χ2 test. The independent variables in the logistic regression models were examined for evidence of multicollinearity and interactions with grade, race/ethnicity, and mother's educational attainment. All final analyses were weighted to be nationally representative.
Table 1 presents the overall prevalence of headache, stomachache, backache, and morning fatigue among US girls in grades 6 through 10. Within this population, about one third experienced headaches or morning fatigue more than once a week, one fifth reported stomachache more than once a week, and almost one quarter experienced backache more than once a week. Almost 40% of the girls reported stomachache about once a month, while more than one third reported morning fatigue about once a week. Only one fifth of the respondents reported rarely or never experiencing headaches or stomachaches and only 7% reported rarely or never experiencing morning fatigue.
Race and ethnicity was not significantly associated with the prevalence of reported symptoms, with one exception. Non-Hispanic black girls were significantly more likely than other girls to report stomachaches more than once a week. Girls whose mothers had not graduated from high school were also at higher risk for frequent headaches (34.5%; 95% CI, 31.5-37.5) and stomachaches (27.6%; 95% CI, 25.6-29.6) compared with those whose mothers had at least completed high school (Table 2).
There was a consistent and significant relationship between grade and symptom prevalence. Morning fatigue 4 times or more a week increased from 26.9% among sixth graders (95% CI, 24.5-28.3) to 36.5% among 10th graders (95% CI, 34.3-38.6). More girls also reported experiencing backache and stomachache once a month with each increasing grade level, with the exception of the 10th grade. The prevalence of stomachache reported once a month increased from 30.5% (95% CI, 29.0-32.0) in sixth grade to 40.8% (95% CI, 39.1-42.4) in 10th grade (Table 2).
Co-occurring symptoms were common. About one third of the girls who reported experiencing headaches more than once a week also reported experiencing at least 1 of the other 3 symptoms at least once a week or more. Among girls reporting stomachaches more than once a week, about 40% also reported experiencing backaches or morning fatigue once a week or more while more than 40% of those with backaches more than once a week reported either headaches or morning fatigue once a week or more (Table 3).
Multivariate analyses in Table 4 indicated that very low levels of support, particularly from parents and teachers, were associated with higher risk for the 4 symptoms studied. Students who reported very low levels of parental support were 1.4 times (95% CI, 1.0-2.0) more likely to experience headaches and 1.5 times (95% CI, 1.1-2.0) more likely to report stomachaches once a week or more while students reporting very low teacher support were 1.9 times (95% CI, 1.3-2.9) and 1.8 times (95% CI, 1.3-2.5) more likely to experience stomach and back pain, respectively, compared with students with high support. Prevalence data support these findings, with higher levels of perceived support from peers, parents, and teachers generally associated with reduced incidence of frequent symptoms, particularly for stomachaches and backaches. The proportion of girls reporting a backache once a week or more decreased from more than 50% with very low levels of teacher or parent support to fewer than one third with high levels of perceived support (data available on request). Girls who drank alcohol were more likely to experience stomachaches (adjusted odds ratio, 1.2; 95% CI, 1.0-1.5) than nondrinkers, while girls who reported high caffeine consumption were 1.4 times more likely to report headaches, stomachaches, and backaches compared with those who consumed lower levels. Teens who smoked were about 1.5 times more likely to report all symptoms compared with nonsmokers. A dose-response was also observed between the prevalence of frequent symptoms and the increasing levels of substance use and caffeine consumption. For example, the prevalence of stomachache once a week or more increased from one third among nondrinkers to almost two thirds among binge drinkers. Caffeine intake demonstrated a similar pattern with the proportion of frequent headache sufferers increasing 30 percentage points between low- and high-caffeine consumers (data available on request). Bullying experienced at least once a week increased the risk for all symptoms. Bullied girls were 1.4 times (95% CI, 1.1-1.8) more likely to report headaches and 1.3 times (95% CI, 1.1-1.6) more likely to experience backaches once a week or more compared with girls who were not bullied. Prevalence estimates supported these findings with three quarters of the girls who experienced bullying at this rate also reporting morning fatigue (data available on request). No evidence of either multicollinearity or interactions were found in the multivariate models.
This first national study to report on multiple somatic complaints indicates that US adolescent females are at particularly high risk for headache, stomachache, backache, and morning fatigue. While other studies have documented the high prevalence of these symptoms among adolescents and girls in particular, few have pursued a gender-specific analysis of multiple co-occurring symptoms.1- 7,31 Adolescent females who report experiencing one symptom more than once a week often experience at least one other symptom at the same frequency rate. Among girls who experienced headaches more than once a week, 3.2 million (53.3%) also reported stomachaches more than once a week and 4.1 million (74.3%) reported morning fatigue more than once a week. The high prevalence of these somatic symptoms experienced alone and concurrently could have important implications for this population's daily functional status and clinical treatment.
This investigation found a generally positive association between grade level and symptom frequency or intensity. Age effects reported for the prevalence of somatic symptoms vary by complaint and symptom frequency in previous studies. Other studies found a strong association between increased age and the prevalence of backache,19,32 while Egger et al4 found no age effects for headaches, stomachaches, or musculoskeletal pains. In this study, significant grade-level effects were observed primarily for symptoms occurring at least weekly and in some cases monthly, possibly capturing the onset of menstruation and associated symptoms.33- 36 The variance in age effects observed in previous studies may be attributed to different sample sizes, age groupings, and study designs. Further research is needed to determine the extent to which somatic complaints can be explained by pubertal developments such as menstruation.
Future research on the relationship between age and somatic complaints may need to consider the degree to which observed age effects are potentially a reflection of girls' increasing participation in risky health behaviors and reduced involvement in protective activities. A study reported significant increases in the proportion of girls using alcohol and tobacco products between the 8th and 10th grades.37 Increased participation in these risky behaviors may be occurring in conjunction with decreased involvement in potentially protective behaviors like physical activity and activities that foster family connectedness. A recent US study found that levels of physical activity decline sharply among adolescent girls, with the median activity score decreasing by 83% between the ages of 9 and 10 years and 18 and 19 years.38 Physical activity among young people has been linked to higher-quality relationships with parents, improved academic performance, and a lower prevalence of depression.39 This study did not find physical activity to be a protective factor for somatic symptoms except backache. This may be owing to the inability to measure total physical activity in the survey. In addition to reduced physical activity, as adolescents age, less leisure time may be spent engaged in family-related activities. Several studies have found parent-child connectedness to be protective against a range of risk behaviors and conditions associated with somatic symptoms including emotional distress, suicidal thoughts, violence, and substance abuse, even among high-risk groups.22,40- 42
Social and institutional support from peers and instructors may also have a significant effect on the health of adolescent girls.19,20,22 In this study, both parent and teacher support measures had a more notable association with somatic symptoms than peer support. For both factors, a protective association was observed for most symptoms even at the low and moderate levels of perceived support (compared with very low). Previous studies have documented the protective effects of parental support in particular.22,40- 42 Resnick et al22 reported that parent-family connectedness was protective against several risk factors for somatic symptoms including substance abuse, violence, and emotional distress. These findings were echoed in studies of 2 high-risk populations: overweight adolescents and African American girls. Among overweight adolescents, Mellin et al41 reported that high levels of family connectedness were positively associated with better school performance and lower levels of emotional distress. Among African American girls, Aronowitz and Morrison-Beedy40 found that connectedness with mothers was associated with fewer risky health behaviors and more positive future aspirations. Similar protective effects have been observed for teacher support.20 These findings suggest that increased parent and teacher support may act to help reduce risk behaviors shown to be related to health symptoms among adolescent girls. Additional research is needed to identify effective methods for fostering parent- and school-connectedness within familial and community structures that may already be overburdened.
Although peer support appeared to play a less protective role than parental or teacher support in this study, the proportion of girls reporting somatic complaints once a week or more in this study decreased by almost 10% for all 4 symptoms between those reporting very low levels and those reporting high levels of peer support. These findings are consistent with those of Natvig et al20 who reported that increasing support from fellow students was associated with decreased risk for somatic symptoms. In particular, Natvig et al observed a protective effect between student support and depression, a symptom commonly linked to a range of somatic symptoms in other studies.4,7,18 Additional research is needed to determine whether some risk factors for somatic symptoms can be effectively addressed through positive peer networks, and if so, what mechanisms can be used to mitigate the negative effects of poor peer relations associated with bullying.
In this study, girls who experienced bullying at least once a week were more likely to experience headaches, backaches, and morning fatigue compared with girls who had not been bullied during that term. Previous studies have documented the association between bullying and both health behaviors and symptoms.21,43,44 These studies indicate that while being bullied is associated with difficulty making friends and lower use of alcohol,21 perpetration of bullying is associated with more time spent with friends44 and increased use of alcohol and cigarettes.21 This suggests that although adolescents who are bullied may be at less risk for adverse health outcomes associated with substance use, they may be at increased risk for somatic complaints associated with poor peer relationships.3,18 New initiatives such as the US Department of Health and Human Services, Health Resources and Services Administration's Stop Bullying Now campaign45 should be evaluated for their efficacy in raising public awareness and reducing the prevalence of bullying.
Binge drinking and frequent smoking were significantly associated with all symptoms. Of note was the similarity in prevalence of frequent headaches, stomachaches, and morning fatigue between girls reporting at least weekly smoking and those reporting less than weekly smoking. This suggests that even sporadic smoking may be associated with somatic symptoms. Smoking among adolescent girls has been associated with somatic symptoms,3 lower parent-family connectedness, and weight control efforts.46 The reported relationship between weight control and smoking is particularly salient to the health status of adolescent girls. As they age, girls report increased dissatisfaction with their bodies,47 participation in dieting behaviors,48 and cigarette smoking.49 It is possible that some of the symptoms associated with adolescent female smoking may also reflect negative health effects associated with the psychological pressure to fit social body norms and the physical effects of food restriction and disordered eating.
Alcohol consumption was modestly associated with increased risk for headaches, stomachaches, and morning fatigue in this study. The association between alcohol abuse and lifetime negative health effects has been documented.50,51 Adolescent alcohol use has been associated with frequent physical symptoms ranging from headaches to abdominal pain, risky health habits including poor diet, smoking, low levels of regular physical activity, and delayed puberty.23,52 At least 1 study has suggested that depression and anxiety among this population may be more significantly associated with reported symptoms than the physical effects of alcohol.53 Another study has reported a strong association between somatic symptoms and "relief" substance abuse among adolescent girls, a practice characterized by substance use during stressful events.54 These findings suggest that the physical symptoms associated with substance use may also be reflecting the somatization of concurrent emotional distress. Further research is necessary to determine to what extent students who report risky health behaviors may be otherwise prone to somatic symptoms and, in particular, how stress and depression, may impact health behaviors and outcomes.
Several limitations should be noted. The school-based study population excludes high-risk teens not attending school. Students in special education classes were also less likely to be interviewed. Students with language difficulties or learning disabilities may have been less likely to complete the questionnaires. Students who were absent on the day of the survey may be more likely to have chronic illnesses or have higher risks associated with truancy or other related factors assessed here. Assessment of effects of absenteeism on adolescent reports of health behaviors has shown it may be a problem.55 Students in alternative schools are excluded from the sample and have been shown to have higher levels of health-related risk behaviors than high school students surveyed in the Centers for Disease Control and Prevention's Youth Risk Behavior Survey.56 Thus, the levels of risk reported here may be lower than those of the adolescent population as a whole.
Assessment of the association between food intake and restriction practices and somatic symptoms was omitted because of concerns about the reliability of self-report data for these measures. Self-reported data on risk behaviors among adolescents have been found to be lower for dietary behaviors compared with other health behaviors,57 while comparisons of self-reported energy intake and expenditure have shown adolescents to underestimate energy intake.58 Food restriction practices were also omitted because data were unavailable on the intensity or frequency of dieting behaviors.
The association between stressful life events, such as familial illness and violence, on somatic symptoms experienced by adolescent girls was not addressed in this study.54 Given the prevalence of sexual assault59 and partner violence60 experienced by adolescent girls, subsequent studies of multiple co-occurring symptoms among this population would benefit from inclusion of these measures.
Finally, a similar study of somatic symptoms among adolescent boys would be valuable. Male adolescents do not report suffering from somatic symptoms as frequently as female adolescents, with the exception of morning fatigue26; however, further research is needed to determine how social and environmental factors influence the frequency and prevalence of this condition and others among boys.
Somatic complaints of headache, stomachache, backache, and morning fatigue are common among US adolescent girls. These health conditions are reported frequently and often in tandem with other conditions affecting daily functional levels. These findings suggest that effective clinical treatment may require comprehensive assessment of all female adolescents presenting with seemingly isolated somatic complaints to accurately identify and treat both the presenting symptom and any related conditions. While linkages may be drawn between selected complaints and other biological functions such as menstruation, most of these complaints seem to be associated more strongly with social, environmental, and behavioral risk factors such as perceived social support and alcohol and caffeine consumption.
Headache, stomachache, backache, and morning fatigue are frequently reported somatic complaints among children and adolescents. Studies have shown that girls report a higher prevalence of these symptoms and are at greater risk for experiencing these symptoms concurrently. The disproportionate rates of symptoms experienced by adolescent girls, in conjunction with the detrimental effects of these symptoms on social development, suggest the need for a better understanding of the nature, extent, and possible causes for frequent somatic symptoms among adolescent girls in the United States.
This first national study to report on multiple somatic complaints among US adolescent females indicates that this population is at particularly high risk for headaches, stomachaches, backaches, and morning fatigue and frequently experiences symptoms in tandem with other conditions. Among girls who experienced headaches more than once a week, 3.2 million (53.3%) also reported stomachaches more than once a week, and 4.1 million (74.3%) reported morning fatigue more than once a week. Analysis of selected social, environmental, and behavioral factors found protective associations for parent and teacher support, while cigarette smoking and heavy caffeine and alcohol consumption were strongly associated with symptom prevalence.
Correspondence: Reem M. Ghandour, MPA, Office of Women's Health, Health Resources and Services Administration, Parklawn Bldg, Room 10C-09, 5600 Fishers Lane, Rockville, MD 20857 (Rghandour@hrsa.gov).
Accepted for publication March 29, 2004.
Ghandour RM, Overpeck MD, Huang ZJ, Kogan MD, Scheidt PC. Headache, Stomachache, Backache, and Morning Fatigue Among Adolescent Girls in the United StatesAssociations With Behavioral, Sociodemographic, and Environmental Factors. Arch Pediatr Adolesc Med. 2004;158(8):797-803. doi:10.1001/archpedi.158.8.797