Sense of Coherence and Medicine Use for Headache Among Adolescents
Article Outline
Abstract
Objective
To examine the association between headache, sense of coherence (SOC), and medicine use for headaches in a community-based sample of adolescents.
Methods
Epidemiological cross-sectional study, encompassing 20 out of 23 schools in the network of health-promoting schools in the county of South Jutland, Denmark. The study population consisted of students from seventh and ninth grade, participation rate 93%, n
=1393. The students answered questions on demographic variables, health behavior including medicine use, psychosocial health aspects, and sense of coherence, in an anonymous standardized questionnaire. The outcome measure was self-reported medicine use for headaches. The determinants were headache frequency and SOC measured by Wold and Torsheim's version for children of Antonovsky's 13-item SOC scale.
Results
Analyses adjusted for age group, family social class, exposure to bullying, and headache frequency showed increasing odds for medicine use for headaches (hereafter: medicine use) by decreasing SOC. There was no association between SOC and medicine use among students with a rare experience of headaches but a significant and graded association among students with at least weekly experience of headaches, that is, frequency of headaches modified the association between SOC and medicine use.
Conclusions
We found that adolescents with low SOC used medicine to cope with headaches to a greater extent than adolescents with high SOC.
Keywords: Adolescents, Medicine use, Pharmacoepidemiology, Sense of coherence
Adolescents frequently use medicine to alleviate common ailments [1], [2], [3], [4]. Furthermore, during the last 2 decades there has been an increase in the consumption of pain-relieving medicine among adolescents that is not because of an increase in pain symptoms [5], [6], [7]. Inappropriate use of pain-relieving medicine is a concern because use of pain-relieving medicine is widespread and may have adverse health effects, for example, medicine-induced headaches [8], [9], [10].
Several studies have demonstrated that headaches is not the only reason for use of medicine for headaches. Both Abbott and Fraser [8] and Hansen et al [11] concluded that to a large extent over-the-counter medicine (primarily pain-relieving medicine) is used to relieve feelings of stress. In epidemiological studies of over-the-counter medicine use, primarily for headaches, factors such as exposure to bullying [12], alcohol use [13], and psychological symptoms as depression and anxiety [2] have been associated with medicine use. Andersen et al [13] found medicine use to cluster with smoking and drunkenness, and suggested that medicine use is part of a behavioral pattern of misuse.
Due et al's study [12] among adolescents shows a strong association between medicine use and frequency of headache symptoms, but not all who experienced headaches used medicine and not all who used medicine experienced headaches. Studies among adults as well as among children show that headaches are often associated with an overuse of medicine, and that an overuse of pain-relieving medicine can lead to daily headaches and dependence [10], [14]. Barea et al [1] conclude in their study of medicine for headaches among school children in Brazil that the usage of pain-relieving medicine does not follow abuse patterns, and that headaches are a genuine problem in the study population. Several researchers have suggested that medicine use may be a behavior that over and above formal therapeutic indications may reflect a general coping strategy to overcome daily stressors [11], [13], [15], [16].
Antonovsky [17], [18] proposed that sense of coherence (SOC) is an individual attribute that protects the individual against the harmful consequences of stressors. He defined SOC as: “The extent to which one has a pervasive, enduring though dynamic, feeling of confidence that one's environment is predictable, and that things will work out as well as can reasonably be expected.” In other words, SOC is a combination of optimism and control. It has three components: comprehensibility, manageability, and meaningfulness. SOC is not a coping mechanism but rather a measure of an individual's capacity to use an appropriate variety of coping mechanisms and other strategies when faced with a stressor. Adults with high SOC generally handle daily stressors better and remain healthy, whereas adults with low SOC seem to be more vulnerable with more health problems [19], [20], [21], [22], [23]. Among adults, high SOC is associated with good self-rated health [24], low risk of psychological despair [25], absence of disease [21], [24]. Findings also support a relationship between SOC and a range of health behaviors, for example, exercise and food selection [26], [27]. The stronger the SOC, the healthier the behavior in general.
A few studies indicate that SOC among adolescents may affect stress and coping in much the same way as it is seen in adults [28], [29]. These observations suggest that high SOC may also reduce the tendency to use medicine when experiencing a headache.
A few studies suggest that the experience of stress, negative coping expectations, and inappropriate coping responses is associated with physiological processes that can affect the threshold for experiencing symptoms [20], [28]. McSherry and Holm [30] find an association between SOC and physical reactions to stress. Antonovsky [17] suggests that SOC may affect the equilibrium of the physiological system. Torsheim et al [28], in their study of adolescents, find a strong association between SOC and symptoms, also in analyses adjusted for the level stress. They suggest that this may be because of individuals with high SOC who show resistance to such a sensitization.
The role of medicine use as a coping mechanism among adolescents is an understudied issue. Most health behaviors are formed in childhood and adolescence, which makes children and adolescents an important target group from a public health point of view. Knowledge of the underlying mechanisms behind health behaviors is important. We have not been able to identify studies on the relationship between SOC and medicine use among either adults or adolescents, or studies on how SOC influences the association between stressors and medicine use. The above studies suggest that high SOC may be associated with a low likelihood of using medicine among both adults and adolescents. This study focused on adolescents only, and the objective was to examine the association between the stressor headaches, SOC, and medicine use in a community-based sample of adolescents. Many of the above studies suggest that the pattern of associations between SOC and health-related variables may differ by gender, and we decided to present our findings separately for boys and girls.
Methods
Design and study population
We used data from the Danish contribution to the international cross-sectional study Health Behavior in School-Aged Children (HBSC) [31]. The general purpose of the HBSC study is to gain insight in young people's health and health-related behaviors [32]. The study in 1998 included a subpopulation from which we collected data about SOC, frequency of headaches, medicine use for headaches, and a range of potential confounder variables. The study population comprised all pupils in grades 7 and 9 (average age 13.6 and 15.6, respectively) from the network of health-promoting schools in one region of Denmark, the county of South Jutland.
We invited all 23 schools in the network of health-promoting schools by separate invitation letters to the headmaster, the school board, and the board of pupils at each school. Twenty of these schools agreed to participate. In total, 98.8% of the pupils present on the day of the data collection participated in the survey, corresponding to 92.6% of the pupils enrolled in the relevant classes, n=1393. Because the study was conducted anonymously, we were unable to analyze the characteristics of nonparticipants. There is no formal agency for ethical approval of questionnaire-based survey studies in Denmark. Therefore, we asked for ethical approval from each individual school's board of parents, school director, and pupil board. The study complies with the Helsinki 2 declaration on ethics.
Data collection
Data was collected in the classroom by means of the internationally standardized HBSC questionnaire after standardized instructions from the teacher. The HBSC questionnaire has been used in all HBSC studies with only few revisions from the beginning of the 1980s [33]. To ensure anonymity, the pupils returned their completed questionnaire in a sealed envelope. No members of the school staff had access to the completed questionnaire, which was sent unopened to the project leader.
Measurements
The dependent variable medicine use was measured by the item: “During the past month, did you take any tablets or medicine for headache?” The responses were categorized as 1: Yes, once + yes several times, and 2: No. There were very few missing data (<2%) and we coded missing as 2. This coding procedure may have introduced some bias; however, this bias would in the worst case mean an underestimation of the association between SOC and medicine use. The validity of self-reported medicine use for headaches among adolescents has been tested by means of a student–parent agreement study that showed a high agreement [34]. Further, a qualitative study with interviews of students who just completed the questionnaire demonstrated that the students understood the questions and were willing and able to respond [35]. The study focuses on the behavior medicine use, but does not include data about what kind of medicine. Qualitative studies among children and adolescents have demonstrated that medicine use for headache in these age groups are predominantly use of over-the-counter medicines [2], [11], [35].
The independent variable SOC was measured by the use of Wold and Torsheim's version for children of Antonovsky's 13-item SOC scale [33]. For an SOC value to be calculated the pupils had to answer at least 11 of the 13 items. SOC was divided into tertiles, that is, an approximately even distribution into three groups from low to high SOC. The 1.4% with missing data was not included in the analyses (9 girls, 10 boys). Wold and Torsheim's version of the SOC scale has been tested in a few other countries as part of the HBSC surveys. Although this version has not been validated, the version for adults has shown to be a valid instrument [18], [23]. We assessed the reliability of the scale that was satisfying, Cronbach's coefficient alpha=0.83.
The selection of confounder variables included four variables that were closely associated with both the independent variable SOC and with the dependent variable medicine use, and that were not considered part of a causal chain between SOC and medicine use: age, frequency of headaches, family social class, and exposure to bullying at school [12].
Headache frequency was measured by the validated HBSC Symptom Check List (HSC-SCL) [36]: “During the past 6 months, how often have you experienced… “ (hereafter were mentioned eight symptoms). We only included headache symptoms in our analyses. The responses were recoded into three levels: 1=almost daily + more than once a week + about once a week, 2=about once a month, 3=less often or never. Missing (1.4%) was coded as 3 so as not to overestimate the association between frequency of headaches and medicine use for headaches.
Family social class [37] was derived from two items on father's and mother's occupation, respectively: “What are your parents' jobs? Please describe exactly what they do, for example shop assistant, farm worker, lorry driver, dentist, hairdresser, teacher. You can write ‘don't know’ or ‘has no paid job at the moment’ or ‘unemployed.’” We coded the responses in accordance with the social class classification of the National Danish Institute of Social Research, a classification that is almost identical with the Registrar General coding often used in the United Kingdom. Each student was categorized in accordance with the highest ranking parent into four levels: I–II (high), III–IV (middle), V–VI (low), and unknown. Children's reports of their parents’ occupation have been shown to be valid [38].
The variable bullying was included as a potential confounder between SOC and medicine use for headaches. Exposure to bullying is an important stressor for young people and probably an important causal factor behind headaches [12]. It may request a high degree of SOC to overcome the burden of bullying. Further, exposure to bullying in itself is a strong predictor of medicine use [12]. It was measured by the item: “How often have you been bullied in school this term?” The responses were categorized into three levels: 1=not at all, 2=once or twice, and 3=a few times + once a week + several times a week. Missing (0.3%) was coded as 1 so as not to overestimate the association between bullying and medicine use. Previous studies have demonstrated a high degree of validity in the applied measurement of bullying [39].
We also considered other potential confounder variables such as self-rated health. We decided not to control for this variable so as to avoid the risk of overcontrol. If, for instance, headaches influence the individual's general self-rated health, and if self-rated health influences medicine use, self-rated health would be a mediator between headaches and medicine use. In this case, it is inappropriate to control for mediating variables.
Statistical procedures
We used SAS software version 9.1 for all analyses. First, we inspected the distribution of the employed variables by gender and grade. Differences between girls and boys and seventh and ninth grade were analyzed using a chi-square test. The distribution of social class and bullying were not significantly different for boys and girls (psocial class=.6544; pbullying=.6093). The distribution of headaches and SOC was not significantly different between seventh and ninth grade (pheadache 7 vs. 9=.5466, pSOC 7 vs. 9=.8073). However, more girls than boys had weekly headaches (p < .0001) and boys had higher SOC than girls (p=.0205). These findings suggested that the pattern of associations could be different for boys and girls and we decided to carry out all analyses separately for boys and girls.
Second, we used logistic regression analyses to estimate the crude association between SOC and medicine use, stratified by gender and age group. As a reference group we consequently chose the least strained group: boys, seventh grade, high social class, rare or never headaches, not bullied, and pupils with high SOC. Third, we used multiple logistic regression analyses to study the association between SOC and medicine use, controlled by the other variables simultaneously as well as individually. The three covariates grade, social class, and bullying were associated with medicine use but did not confound the association between SOC and medicine use. Therefore, we only present analyses that are not adjusted for these variables.
We also tested an interaction term between SOC and frequency of headache, but this interaction was not significant (p=.2498) and was therefore not included in the statistical models. Despite the insignificant interaction term we decided to study the combined effect of SOC and frequency of headaches on medicine use. The reference group in these analyses was students with the combination of high SOC and never/rarely headaches.
Associations were reported as odds ratios (OR) with 95% confidence interval (CI). The final analyses included 675 girls and 699 boys.
Results
Table 1 shows the gender and grade specific distribution of social class, headache frequency, bullying, SOC, and medicine use.
Table 1. Distribution of variables by gender and grade
| Sample | Boys | Girls | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Total | 7th grade | 9th grade | 7th grade | 9th grade | |||||
| n=1393 | n=340 | n=369 | n=362 | n=322 | |||||
| % | n | % | n | % | n | % | n | ||
| Medicine use for headache | |||||||||
| Yes | 30.3 | 103 | 36.6 | 135 | 43.9 | 159 | 54.4 | 175 | |
| No | 69.7 | 237 | 63.4 | 234 | 56.1 | 203 | 45.7 | 147 | |
| Sense of coherence | |||||||||
| High | 37.9 | 129 | 38.5 | 142 | 31.8 | 115 | 31.4 | 101 | |
| Medium | 31.8 | 108 | 32.8 | 121 | 32.6 | 118 | 34.5 | 111 | |
| Low | 29.1 | 99 | 27.1 | 100 | 33.7 | 122 | 33.5 | 108 | |
| Missing | 1.2 | 4 | 1.6 | 6 | 1.9 | 7 | 0.6 | 2 | |
| Headache frequency | |||||||||
| Weekly | 17.7 | 60 | 17.3 | 64 | 29.0 | 105 | 37.0 | 119 | |
| Monthly | 27.1 | 92 | 31.2 | 115 | 29.8 | 108 | 24.5 | 79 | |
| Rare/never | 55.3 | 188 | 51.5 | 190 | 41.2 | 149 | 38.5 | 124 | |
| Social class | |||||||||
| I–II | 21,5 | 73 | 23,6 | 87 | 20.4 | 74 | 24,2 | 78 | |
| III–IV | 50.0 | 170 | 56.1 | 207 | 54.7 | 198 | 57.8 | 186 | |
| V–VI | 17.1 | 58 | 13.0 | 48 | 13.3 | 48 | 13.4 | 43 | |
| Unknown | 11.5 | 39 | 7.3 | 27 | 11.6 | 42 | 4.7 | 15 | |
| Bullying | |||||||||
| Not at all | 47.4 | 161 | 59.6 | 220 | 49.2 | 178 | 60.6 | 195 | |
| Once or twice | 29.7 | 101 | 21.4 | 79 | 25.7 | 93 | 20.5 | 66 | |
| Several times | 22.9 | 78 | 19.0 | 70 | 25.1 | 91 | 18.9 | 61 | |
Significantly more girls than boys experienced headaches on a weekly basis (p < .0001) and used medicine for headaches (χ2=33.5071, df=1, p < .0001). In the seventh grade, 43.9% of the girls used medicine for headaches compared to 30.3% of the boys, and in the ninth grade, the difference is even more pronounced, with 54.4% and 36.6%, respectively. Significantly more pupils in the ninth grade compared to the seventh grade used medicine for headaches (χ2=8.1816, p=.0042). Boys had significantly higher SOC than girls (p=.0205). The covariates headache, bullying, and social class were significantly correlated with SOC (p < .0001).
Table 2 shows the association between SOC and medicine use and the association between frequency of headache and medicine use. The crude analyses showed a graded and significant association, that is, increasing odds for medicine use by decreasing SOC and increasing odds for medicine use by increasing headache frequency. The association between SOC and medicine use attenuated and became insignificant when controlled for frequency of headache. Control for the other confounder variables (grade, social class, bullying) did not influence the association between SOC and medicine use (data not shown).
Table 2. OR (95% CI) for medicine use for headache: girls and boys separately, unadjusted and adjusted models
| n | Crude OR | Adjusted for headache | |
|---|---|---|---|
| Girls, total | 675 | ||
| Sense of coherence | |||
| High | 216 | 1.0 | 1.0 |
| Medium | 229 | 1.39 (0.96–2.02) | 1.06 (0.70–1.60) |
| Low | 230 | 1.79 (1.23–2.61) | 1.25 (0.82–1.89) |
| Headache frequency | |||
| Weekly | 222 | 8.08 (5.40–12.09) | 7.55 (5.01–11.36) |
| Monthly | 184 | 3.47 (2.34–5.16) | 3.33 (2.23–4.98) |
| Rare/Never | 269 | 1.0 | 1.0 |
| n | |||
| Boys, total | 699 | ||
| Sense of coherence | |||
| High | 271 | 1.0 | 1.0 |
| Medium | 229 | 1.73 (1.18–2.53) | 1.20 (0.78–1.84) |
| Low | 199 | 2.09 (1.41–3.10) | 1.40 (0.90–2.18) |
| Headache frequency | |||
| Weekly | 123 | 14.32 (8.84–23.19) | 13.64 (8.33–22.32) |
| Monthly | 202 | 4.51 (3.04–6.67) | 4.43 (2.98–6.60) |
| Never/rarely | 374 | 1.0 | 1.0 |
Finally, we studied how medicine use was related to combinations of SOC and frequency of headache (Table 3). There was a significant different distribution of headache frequency among pupils with low, medium, and high SOC respectively, among girls (χ2=36.2427, df=4, p<0.0001) and boys (χ2=36.3687, df=4, p < .0001). Among students with infrequent headache there was no association between SOC and medicine use. Among students with monthly headache there was an increased odds ratio of using medicine regardless of SOC and there was no association between SOC and medicine use. Among students with weekly or more often headache there was a graded association between SOC and medicine use, that is, increasing odds for medicine use by decreasing SOC. Girls with weekly headaches and low SOC had almost 11.5 times higher odds of medicine use compared to girls with rare/never headaches and high SOC. The equivalent OR for the boys was 20.73 (CI: 9.43–45.57).
Table 3. OR (95% CI) for medicine use by headache frequency and SOC
| Girls: Headache frequency | |||
|---|---|---|---|
| Sense of coherence | Never/rarely | Monthly | Weekly |
| High | 1.0 | 4.59 (2.29–9.21) | 5.94 (2.86–12.36) |
| Medium | 1.37 (0.72–2.61) | 3.00 (1.60–5.64) | 8.38 (4.32–16.27) |
| Low | 1.07 (0.54–2.12) | 4.30 (2.19–8.45) | 11.41 (6.01–21.65) |
| Boys: Headache frequency | |||
|---|---|---|---|
| Sense of coherence | Never/rarely | Monthly | Weekly |
| High | 1.0 | 4.62 (2.45–8.70) | 6.33 (2.61–15.39) |
| Medium | 0.74 (0.36–1.53) | 5.54 (2.99–10.27) | 19.66 (8.91–43.38) |
| Low | 1.59 (0.82–3.06) | 4.18 (2.16–8.08) | 20.73 (9.43–45.57) |
Discussion
Our study showed a significant and graded association between SOC and medicine use for headaches, that is, the higher the SOC, the lower the medicine use. This association became insignificant when adjusted for frequency of headaches. There was no association between SOC and medicine use in strata with rare or monthly appearance of headaches, but a graded association among students who experienced headaches at least weekly. This means that frequency of headaches modified the association between SOC and medicine use. Although the distribution of SOC, headache frequency, and medicine use differed by gender, the pattern of associations between SOC and medicine use was similar for boys and girls.
We cannot directly compare our findings to similar studies as we have been unable to find other studies of the association between SOC and medicine use among adolescents. The strong association between frequency of headaches and medicine use for headaches has been found in other studies, for example, Due et al [12].
Our study is cross-sectional, and does not reveal how SOC may contribute to medicine use or what role headache plays in this association, but there are several possible explanations. Our analyses show increasing odds for medicine use with decreasing SOC among pupils who experience weekly headache symptoms. This is in concordance with studies that show that low SOC in connection with other strains increase the likelihood for health problems, whereas high SOC reduces these health problems [19], [20], [21], [22], [23]. According to Antonovsky's theory, SOC acts as a buffer when faced with stressors, and the effect of SOC is thus only seen when a person is exposed to the stressor headache [17].
Further, we cannot exclude that headaches are a mediating factor between SOC and medicine use. A few studies suggest that the experience of stress and inappropriate coping responses is associated with physiological processes that can affect the threshold for experiencing symptoms [20], [28]. McSherry and Holm [30] find an association between SOC and physical reactions to stress. Antonovsky [17] suggests that SOC may affect the equilibrium of the physiological system and Torsheim et al [28] find a strong association between SOC and symptoms when they adjust for stress. They point out that this may be because of individuals with high SOC showing resistance to such a sensitization. The girls in this study had significantly lower SOC than the boys, which might explain the higher headache frequency and the higher usage of medicine among the girls.
There is a strong association between subjective symptoms and SOC [24], [28]. Larsson and Kallenberg [20] found a significant association between headaches and SOC among adults, and others have found significant correlations between SOC and depression and anxiety [22], [30]. The 13-item SOC scale has been criticized for being contaminated by emotionality [40]. Geyer [40] suggests that measurements of SOC partly reflect negative affectivity. According to the theory of symptom sensation, negative affectivity is associated with high introspection and a low threshold for symptom sensation [28]. Seen from this perspective, the association between SOC and headaches could reflect a mutual influence of negative affectivity. Schnyder et al [22], however, conclude on the basis of two longitudinal studies of SOC among adults that SOC is not merely a result of psychopathology but rather a measure of an individual's world view. Our study shows that even pupils with high SOC experience weekly headaches. The significant association between headaches and medicine use when adjusting for SOC suggests that SOC and headaches are not merely two different measures of the same latent phenomenon.
The study was appropriate for the planned analyses because it included a large community-based population, had a high response rate, and validation studies have proven the included variables to be appropriate for age equivalent populations. The study was based on the internationally standardized HBSC questionnaire that has been used without substantial changes since the 1990s. It has been tested several times in a number of countries [33].
The study has limitations also: the most important limitation is the measurement of headaches by frequency instead of measurement by severity. We would have liked to analyze the association between SOC and medicine use in strata of adolescents with mild and severe headache, but this was unfortunately not possible. There is a risk of selection bias because pupils who were not present on the day of the survey, or who did not answer the questions may have a higher headache frequency and medicine use and may be characterized by lower SOC. If this is the case, we are likely to have underestimated the associations between SOC and medicine use. Further, we systematically coded (the very few) missing as our reference category that may have introduced a slight bias. We think that a bias in order not to overestimate the associations is an appropriate and cautious way of reporting association patterns.
The data were collected 10 years ago, and the prevalence of the key variables (SOC, headache frequency, medicine use) may have changed since the late 1990s. There is, however, no prior reason to believe that the association between SOC and medicine use is sensitive to time.
In a long-term perspective, early use of medicine to alleviate pain and stress may prevent learning more appropriate ways of coping, and potentially lead to life-long medicine use. To prevent excess medication use in adolescence, it is important to know when adolescents use medicine for its intentional purpose, when they use it as a coping mechanism, and when they take medicine for other purposes such as recreational use. It is important for parents and healthcare practitioners who work with young people to be aware that medicine use for headaches is common, and may be influenced by many other factors than a headache itself.
Both qualitative studies and longitudinal research are needed to further study the causal mechanisms behind adolescents' use of medicine. Research in this area is challenged by the developmental changes that occur over the adolescents' life period, and should address the reasons for use of medicine as well as any change in causes that may occur over time.
Acknowledgments
The study used data from the Danish part of the international research project Health Behaviour in School-Aged Children—a WHO crossnational study (HBSC) from 1998. The Principal Investigator in Denmark, Dr. Pernille Due, from the National Institute of Public Health at the University of Southern Denmark, provided the data. The research was supported by the Health Insurance Foundation, the Ministry of Health and the Interior, the Ministry of Social Affairs, and the Nordea Denmark Foundation.
References
- . Analgesic use for headache treatment by schoolchildren of Southern Brazil. Pharmacoepidemiol Drug Safety. 1997;6:359–366
- . Adolescents' use of prescribed drugs and over-the-counter preparations. J Public Health Med. 1996;18:437–442
- . International survey of self-reported medicine use among adolescents. Ann Pharmacother. 2003;37:361–366
- . Adolescents' drug use and drug knowledge. Eur J Pediatr. 2000;159:608–614
- . Self-reported medicine use among 11- to 15-year-old girls and boys in Denmark 1988–1998. Scand J Public Health. 2003;31:334–341
- Holstein BE, Andersen A, Hansen EH, Due P. Children's and adolescents' medicine use for pains and psychological problems: secular trends from 1988 to 2006 (in Danish: Børn og unges brug af lægemidler for smerter og psykiske problemer 1988–2006). Ugeskrift for laeger, in press.
- . Trends in the use of psychotropic medications among adolescents, 1994 to 2001. Psychiatr Serv. 2006;57:63–69
- . Use and abuse of over-the-counter analgesic agents. J Psychiatr Neurosci. 1998;23:13–34
- . Consumption of medicines among adolescent students: a concern. J Pediatr. 2004;80:326–332
- . Analgesic overuse among adolescents with headache: the Head–HUNT–Youth study. Neurology. 2006;66:198–201
- . Using analgesics as tools: young women's treatment for headache. Qual Health Res. 2008;18:234–243
- . Is victimization of bullying associated with medicine use among adolescents: a nationally representative cross-sectional survey in Denmark. Peadiatrics. 2007;120:110–117
- . Is medicine use in adolescence risk behaviour? Cross-sectional survey of school-aged children from 11 to 15. J Adolesc Health. 2006;39:362–366
- . Successful withdrawal form analgesic abuse in a group of youngsters with chronic daily headache. J Child Neurol. 2001;16:448–449
- . “Social medication” and the control of children: a qualitative study of the over-the-counter medication among Australian children. Paediatrics. 2004;114:378–383
- . Epidemiological evaluation of drug use in children. J Clin Pharmacol. 1994;34:300–305
- . Unravelling the Mystery of Health. San Francisco, CA: Jossey-Bass; 1987;
- . The structure and properties of the sence of coherence scale. Soc Sci Med. 1993;36:725–733
- . Trait anxiety and sense of coherence: a longitudinal study. Psychol Rep. 1989;65:221–232
- . Sense of coherence socioeconomic conditions and health—interrelationships in a nation-wide Swedish sample. Eur J Public Health. 1996;6:175–180
- . Sense of coherence and health: evidence from two crosslagged longitudinal samples. Soc Sci Med. 1991;50:583–597
- . Antonovsky's sence of coherence: trait or state?. Psychother Psychosom. 2000;69:296–302
- . Validity of Antonovsky's sense of coherence scale: a systematic review. J Epidemiol Community Health. 2005;59:460–466
- . Sense of coherence, social class, and health in a population based study in Denmark (in Danish: “Sense of coherence,” socialgruppe og helbred i en dansk befolkningsundersøgelse. Ugeskrift for laeger. 1998;160:724–729
- . Sense of coherence, life stress and psychological distress: a prospective methodological inquiry. J Clin Psychol. 1990;46:415–421
- . Physical exercise and psychological well-being: a population study in Finland. Prev Med. 2000;30:17–25
- . Food selection associated with sense of coherence in adults. Nutr J. 2005;4:9
- . Sense of coherence and school-related stress as predictors of subjective health complaints in early adolescence: interactive, indirect or direct relationships. Soc Sci Med. 2001;53:603–614
- . School-related strain, coping, and mental health complaints among Norwegian 15-year-olds (in Norwegian: Skolerelaterte belastninger, mestring og subjektive helseplager blant norske 15-åringer). Tidsskrift Norsk Psykologforening. 1999;36:315–321
- . Sense of coherence: its effects on psychological and physiological processes prior to, during, and after a stressful situation. J Clin Psychol. 1994;50:476–487
- . Socioeconomic health inequalities among a nationally representative sample of Danish adolescents: the role of social relations. J Epidemiol Community Health. 2003;57:692–698
- In: Currie C, Hurrelmann K, Settertobulte W, Smith R, Todd J editor. Health and Health Behaviour among Young People. Copenhagen: WHO; 2000;
- . Measuring the health and health behaviours among adolescents through cross-national survey research: recent developments in the Health Behaviour in School-Aged Children (HBSC) study. J Public Health. 2007;15:179–186
- . Medicine use among 11- and 13-year-olds: agreement between parents' reports and children's self-reports. Ann Pharmacother. 2007;41:581–586
- . School-children's symptom related medicine use (M.Sc.-thesis) (in Danish: Skolebørns symptomrelaterede lægemiddelanvendelse). Copenhagen: The Danish University of Pharmaceutical Sciences; 2004;
- . Subjective health complaints in adolescence: reliability and validity of survey methods. J Adolesc. 2001;24:611–624
- . Social class variation in medicine use among adolescents. Eur J Public Health. 2004;14:49–52
- . Adolescents' proxy reports of parents' socioeconomic status: how valid are they?. J Epidemiol Community Health. 2001;55:731–737
- . Prevalence estimation of school bullying with the Olweus bully/victim questionnaire. Aggress Behav. 2003;29:239–268
- . Some conceptual considerations on the sense of coherence. Soc Sci Med. 1997;44:1771–1779
PII: S1054-139X(08)00677-0
doi:10.1016/j.jadohealth.2008.12.009
© 2009 Society for Adolescent Medicine. Published by Elsevier Inc. All rights reserved.
