Recurrent pain in school-aged children: a longitudinal study focusing on the relation to academic achievement

1. Introduction

Recurrent pain has, during the last decades, been recognized as an increasing health problem among school-aged children, particularly adolescent girls.5,14 Up to 90% of school-aged children experience recurrent pain,16 and about 5% suffer from moderate-to-severe chronic pain.11 Children with recurrent pain may experience impairment of daily life related to pain, such as impaired school functioning.12,20,25 A higher pain frequency along with a higher number of pain sites have been associated with a higher degree of impaired functioning.12,22

School is an essential part of children's everyday life, and success in elementary school also has implications for future opportunities, such as access to higher education and later employment, and may also be a predictor for later well-being and health.3,35 However, not all children are successful in school. In Sweden, for instance, about 10% to 15% of pupils leaving elementary school do not meet the requirements to enter upper secondary school.33 During the last decades, there has been a decline in children's academic achievement in Sweden.24

The period of decreasing results in Swedish schools coincides with an increase in the prevalence of recurrent pain among school-aged children.9 The question is whether these trends are interrelated. In a recent systematic review, we show that recurrent pain may be a predictor of lower academic achievement, including failure to achieve school qualifications.29 Previous studies indicate that recurrent pain may intrude on children's sleep quality and their capacity to concentrate,34,36 which may in turn lower their ability to achieve in school.2,6,8 In addition, we have shown that children with recurrent pain perceive more problems regarding their academic achievement than other children.28 Perceived problems with academic achievement (PPAA) reflect the child's academic self-esteem, a concept that is related to the child's self-construct but also to lower actual achievement.27 Thus, the path between recurrent pain and academic achievement may potentially go through sleep and concentration problems, and the child's perception of his or her ability to achieve academically may contribute to this relationship.

In our abovementioned review, we conclude that the overall evidence for an association between recurrent pain and academic achievement is limited.29 There is particularly a lack of longitudinal studies that capture the most common pain conditions in childhood simultaneously and the significance of multiple pain in relation to objectively measured academic achievement.

The aim of this study was to investigate the longitudinal relationship between recurrent pain and academic achievement among school-aged boys and girls.

More specifically, we aimed to find out (1) whether recurrent pain (headache, stomachache, or backache) in school year 6 predicts the final overall grade points at the end of elementary school in school year 9 or the eligibility for entering upper secondary school; (2) whether these potential associations differ depending on pain frequency, location, and co-occurrence; and (3) whether they are mediated by sleep problems, concentration problems, and PPAA. We hypothesized that (1), (2), and (3) would all be confirmed.

2. Methods

The study was based on data from the “Study of Health In School-Aged Children from Umeå” (SISU), which is a longitudinal total population-based study performed in northern Sweden. Further information about SISU is available elsewhere.25 In Sweden, elementary school includes years 1 to 9, further divided into 3 stages with school years 1 to 3 being the first stage, years 4 to 6 the second, and years 7 to 9 the last. Elementary school is compulsory, and children generally start school the calendar year they turn 7 years old and finish elementary school the calendar year they turn 16.

2.1. Study population and procedure

This study included all children who attended school year 6 (ie, aged approximately 12-13 years) in any school in the city of Umeå in 2003 or 2006 (Fig. 1). These 2 cohorts were followed up 3 years later, in school year 9 (aged approximately 16 years). All invited schools and 1644 (97%) of the 1691 invited children accepted the invitation to take part in the study. In school year 6 (T1), the children were asked to fill out a questionnaire at school and their parents were asked to fill out a background form at home. The children's final grades in school year 9 (T2) were collected from school registers. The participants included in this study were all children who provided complete answers on questions regarding recurrent pain (independent variable) in school year 6 and for whom information about final grades or secondary school eligibility (dependent variable) in school year 9 were available from the school registry (n = 1567, participation rate = 93%). The reasons for the internal missing were that the child did not provide complete answers on the pain variables at T1 (n = 50) and problems identifying the grades at T2 (n = 74), for example, because of children moving abroad.

F1Figure 1.:

Flow chart for the data collection of the cohorts. T1, time 1, T2, time 2.

2.2. Measures 2.2.1. Recurrent pain

Recurrent pain was assessed by 3 questions asking how often the child had experienced headache, stomachache, and backache, respectively, in the past 6 months. The questions had 5 response alternatives: (1) “seldom or never,” (2) “about every month,” (3) “about every week,” (4) “more than once per week,” and (5) “about every day”. Pain frequency was labelled as pain never or seldom (response alternative (1)), monthly pain (alternative (2)), and weekly pain or frequent pain (alternatives (3)–(5) merged). Recurrent pain was defined as pain occurring at least once a month during the past 6 months (response alternatives (2)–(5) merged). Recurrent pain from only one site was named single pain, and recurrent pain from more than one site was named multiple pains. The questions were derived from the international survey Health Behavior in School-Aged Children and had been previously shown to have adequate validity and reliability.10

2.2.2. Academic achievement

Academic achievement was captured by 2 measures: children's final overall grade points in school year 9 and eligibility to enter upper secondary school, hereafter named overall grade points and secondary school eligibility. The overall grade points were estimated as the sum of the 16 subjects in which the child had obtained the highest grades. The performance in each subject was graded at 4 levels: “not passed” (0 points), “passed” (10 points), “passed with distinction” (15 points), and “passed with special distinction” (20 points). Thus, the overall grade points ranged from 0, which means no subjects were passed, to 320, which means all 16 subjects were passed with special distinction. Secondary school eligibility was granted to children who at least “passed” in the 3 core subjects of mathematics, English, and Swedish when leaving elementary school in school year 9. An exception was 20 participants who attended a school that graded the overall qualification for high-school education but gave no grades for individual subjects. These 20 participants were only included in the analyses regarding secondary school eligibility.

2.2.3. Potential mediators

In school year 6 (T1), the children also answered 3 questions capturing how often they experienced problems with sleep and concentration and how often they perceived having problems with academic achievement. These questions each had 5 response alternatives ranging from “never” to “almost always” during the past month. When problems occurred almost always, often, or sometimes, we coded this as impaired sleep, impaired concentration, and PPAA. The questions derived from the Pediatric Quality of Life Inventory (PedsQL), and reliability has been verified for each question.26

2.2.4. Background variables

Backgrounds variables consisted of sex, parental country of birth, and family structure, as reported by the children. The parent reported their level of education and whether the child had any long-term illness. For this study, we labelled illnesses that were not specific pain conditions as “nonpain long-term illnesses”. Finally, we registered whether the child belonged to the 2003 or the 2006 cohort.

2.3. Statistical analyses

All statistical analyses were performed in the Statistical Package for the Social Sciences versions 24.0 and 26.0.

2.3.1. Missing data

Missing data were found in 3 of the background variables and in 2 of the potential mediators, namely, family structure (n = 7), parental education (n = 131), long-term illnesses other than pain (n = 158), sleep (n = 8), and concentration (n = 3). Missing data were found to occur at random and were therefore replaced by use of the function “multiple imputation” in Statistical Package for the Social Sciences.

2.3.2. Associations

Correlations between the variables recurrent pain, country of birth, parental education, family structure, and nonpain long-term illnesses were tested with Spearman rho, and multicollinearity was rejected by the use of the variance inflation factor, which was not > 2 for any correlation. Owing to the rather large sample size, overall grade points in school year 9 were treated as approximatively normally distributed.15 General linear models were used to analyze the longitudinal association between recurrent pain in school year 6 and later academic achievement expressed as overall grade points in school year 9. Likewise, univariate and multivariate logistic regression analyses were performed to test the longitudinal association between recurrent pain in school year 6 and later academic achievement expressed as secondary school eligibility in school year 9.

First, a crude model (model 1) without any covariates was used to test associations between pain and academic achievement. Pain was operationalized as frequencies of headache, stomachache, and backache, separately, and merged into frequency of pain at any of these 3 sites. In addition, the number of recurrent pain sites, “pain co-occurrence,” and a mixture of frequencies and number of pain sites, “pain co-occurrence, and frequency” was applied. In a second model, 5 covariates were added, cohort (2003 vs 2006), nonpain long-term illnesses (yes vs no), parental country of birth (both parents born in Europe or North America: yes vs no), family structure (living with both parents: yes vs no), and parental education (at least one of the parents had a university or college degree: yes vs no).

2.3.3. Potential mediators

In a third model, the potential mediators (problems with concentration, sleep, and PPAA) were added one by one into the adjusted model described in the previous section (model 2), using the independent variable recurrent pain at any site. Before adding the potential mediators into the model, associations between the independent variable and the potential mediators were tested using Spearman correlation. Likewise, associations were tested between the potential mediators and the dependent variables, that is, overall grade points (Spearman correlation) and secondary school eligibility (chi-square analysis). Only potential mediators with verified association to both the independent and the dependent variables were added to the third model.

2.3.4. Sensitivity analyses

At the time of the study, no grades were given in school year 6, but self-rated and parent-rated academic achievements were available from school years 6 and 3, respectively. These data were included in every model to test the impact of outcome at baseline. More specifically, as part of the SISU study, the children in one cohort had rated their academic achievement in 6 specific subjects when the child attended school year 6 (2003 cohort), and in the other cohort, parents had rated the child's academic achievement when the child attended school year 3 (2006 cohort). This information was sampled using the standardized questionnaires Youth Self-Report and Child Behavior Check List (parent report).

In addition, the potential mediators were added one by one, but with all 5 response alternatives (“never” to “almost always”) instead of a dichotomized variable “problem” vs “no problem.”

Furthermore, multilevel analyses were performed to test the impact of clustering of children within schools and classes. Two-level and 3-level analyses were performed: individuals nested within classes at year 6, individuals nested within schools at year 6, and 3-level models with individuals nested within classes, which were in turn nested within schools (also year 6). The log likelihood-ratio test was used to determine whether multilevel models were a better fit to the data than single-level regression models.

All analyses were stratified by sex. The significant level was set to 95%, corresponding to P < 0.05. All results presented as differences and associations in the results section are statistically significant at this level.

2.4. Ethics

Children and their parents received an information letter about the study. Moreover, informed consent was obtained from the child as well as from the parents. Ethics approval was attained from the Research Ethics Committee of the Medical Faculty, Umeå University, Sweden UmDnr 03-352 (05-152).

3. Results 3.1. Study characteristics

Descriptive statistics of the background factors and potential mediators are presented in Table 1. About 2 thirds of the included 1567 children lived with both parents, almost all parents were born in Sweden, and about 2 thirds of the parents had a university or college degree. A minority of the children had a long-term illness that was not a specific pain condition, such as asthma, allergy, or diabetes (∼15%). About 15% to 30% of the children had problems with concentration, sleep, or PPAA. Close to 80% of the girls and 70% of the boys experienced recurrent pain in school year 6 (Table 2). In both girls and boys, the most commonly reported pain was headache, followed by stomachache and backache. The prevalence of recurrent headache was 61% in girls and 54% in boys. Recurrent stomachache was reported by about 50% of both girls and boys and backache by about 30% of the girls and boys. Concerning academic achievement in school year 9, the mean overall grade points were 229 points in girls and 205 points in boys, of 320 possible points, and 8% of the girls and boys left elementary school without secondary school eligibility.

Table 1 - Background characteristics and potential mediators among children in school year 6 presented by sex (girls N = 758 and boys N = 809). Girls, N (%) Boys, N (%) Participants 758 (48) 809 (52) Cohort 2003 362 (49) 441 (55) Both parents born in Europe or North America 715 (94) 740 (91) At least one parent with a college or university degree 477 (63) 537 (69) Live with both parents 501 (66) 559 (70) Long-term illness (child)* 110 (15) 140 (17) PPAA 72 (10) 148 (18) Sleep problem 170 (22) 141 (17) Concentration problem 115 (15) 218 (27)

*Specific pain conditions excluded. PPAA, perceived problem with academic achievement.


Table 2 - Prevalence of recurrent pain and secondary school eligibility among the participating girls (N = 758) and boys (N = 809), along with their overall grade points. Girls, N (%) Boys, N (%) Any pain site  Seldom or never 170 (22) 247 (31)  Monthly 253 (33) 274 (34)  Weekly 336 (44) 288 (35) Headache  Seldom or never 297 (39) 369 (46)  Monthly 243 (32) 250 (31)  Weekly 219 (29) 190 (23) Stomachache  Seldom or never 353 (47) 449 (56)  Monthly 240 (32) 247 (31)  Weekly 166 (22) 113 (14) Backache  Seldom or never 548 (72) 599 (74)  Monthly 117 (15) 119 (15)  Weekly 94 (12) 91 (11) Pain co-occurrence  Seldom or never 170 (22) 247 (31)  Singe pain 219 (29) 230 (28)  Multiple pain 370 (49) 332 (41) Pain co-occurrence and frequency  Seldom or never 170 (23) 247 (30)  Monthly SP 147 (19) 159 (20)  Weekly SP 72 (9) 71 (9)  Monthly MP 106 (14) 115 (14)  Weekly MP 264 (35) 217 (26) Secondary school eligibility  Yes 698 (92) 743 (92)  No 60 (8) 66 (8)  Overall grading points M (SD) M (SD  M (SD) 229 (61) 205 (55)  Range 0-320 0-320

M, mean; SD, standard deviation; SP, single-site pain; MP, multisite pain.


3.2. Association between recurrent pain in school year 6 and overall grade points in school year 9

The identified associations between recurrent pain in school year 6 and overall grade points in school year 9 generally remained statistically significant after adjusting for the background factors: parental country of birth, family structure, parental education, nonpain long-term illnesses, and cohort (Table 3). Therefore, the text below only summarizes the results from the adjusted models.

Table 3 - Differences in overall grade points in school year 9 between girls and boys with recurrent pain at different frequencies and locations in school year 6. Girls (N = 746) Boys (N = 802) N B (SE) crude B (SE) adj† N B (SE) crude B (SE) adj† Any pain  Seldom or never 169 Ref Ref 244 Ref Ref  Monthly 247 −10.89 (5.72) −8.18 (5.91) 273 −1.96 (4.71) –0.406 (4.65)  Weekly 330 −20.12 (5.77)* −15.35 (5.64)* 285 −12.47 (5.17)* −11.45 (4.77)* Headache  Seldom or never 292 Ref Ref 366 Ref Ref  Monthly 239 −7.04 (5.32) −5.29 (5.20) 249 −3.31 (4.52) −2.41 (4.48)  Weekly 215 −15.73 (5.49)* −11.49 (5.35)* 187 −13.65 (4.94)* –12.74 (4.90)* Stomachache  Seldom or never 350 Ref Ref 444 Ref Ref  Monthly 233 −8.84 (5.08) −6.55 (4.98) 246 −7.26 (4.37) −5.54 (4.33)  Weekly 163 −28.29 (5.74)* −27.67 (6.34)* 112 −18.12 (5.81)* −16.08 (5.79)* Backache  Seldom or never 538 Ref Ref 594 Ref Ref  Monthly 115 −15.97 (6.26)* −11.85 (6.10) 118 0.525 (5.56) 1.84 (5.52)  Weekly 93 −20.82 (6.84)* −17.58 (6.65)* 90 −13.77 (6.23)* −12.57 (6.15)* Pain co-occurrence  Seldom or never 169 Ref Ref 244 Ref Ref  SP 214 −7.01 (6.26) −5.53 (6.08) 229 –0.745 (5.06) –0.608 (4.99)  MP 363 −21.57 (5.66)* −16.98 (5.55)* 329 −12.34 (4.64)* −10.54 (4.65)* Pain co-occurrence + frequency  Seldom or never 169 Ref Ref 244 Ref Ref  Monthly SP 143 −9.90 (6.88) −9.09 (6.68) 158 0.221 (5.58) 0.509 (5.54)  Weekly SP 71 −1.19 (8.56) −1.78 (8.38) 71 −2.89 (7.37) −3.00 (7.32)  Monthly MP 104 −12.25 (7.54) −7.06 (5.88) 115 −4.97 (6.18) –0.3.27 (6.21)  Weekly MP 259 −25.31 (6.00)* −20.09 (5.88)* 214 −16.30 (5.14)* −14.38 (5.13)*

*P < 0,05.

†Adjusted for parental country of birth, parental education, family structure, chronic diseases, and cohort.

SP, single-site pain; MP, multisite pain.


3.2.1. Pain frequency

Regarding pain frequency, weekly (but not monthly) pain in school year 6 was associated with overall grade points in school year 9 (Table 3). In girls with weekly pain, the overall grade points were 15 points lower compared with girls who never or seldom experienced pain. In boys, the corresponding number was 11 points lower.

3.2.2. Pain locations

Similar results were obtained for each of the 3 studied pain locations, that is, associations between recurrent pain in school year 6 and overall grade points in school year 9 were found for weekly pain but not for monthly pain (Table 3). Comparing those with weekly headache and those who seldom or never had pain (year 6), the differences in overall grade points in school year 9 were 11 points in girls and 12 points in boys. For weekly stomachache, these differences were 27 points in girls and 16 points in boys, and for weekly backache, they were 17 points in girls and 12 points in boys.

3.2.3. Pain co-occurrence

Girls and boys who experienced recurrent pain at only one location in school year 6 (single pain) had similar overall grade points in school year 9 to the reference category (who seldom or never experienced pain) (Table 3). However, there was a negative association between multiple recurrent pain in school year 6 and overall grade points in school year 9; that is, overall grade points were on average 11 (girls) and 17 (boys) points lower in those experiencing recurrent pain at multiple locations than in those who seldom or never experienced pain.

In children who experienced pain at multiple locations on a weekly basis, the differences were even greater, that is, about 25 points in girls and 14 points in boys (Table 3).

3.2.4. Mediators

Mediation was tested in the models and showed that a higher frequency of recurrent pain (weekly pain) in school year 6 was associated with a lower grade point average in school year 9 in both boys and girls. The mediation tests showed that 2 of the 3 potential mediators, namely, concentration problems and PPAA, partly explained this association (Fig. 2).

F2Figure 2.:

Mediation of the association between recurrent pain in school year 6 and overall grade points in school year 9. R, correlation coefficient; B, beta coefficient; SE, standard error; A, difference in merit ratings between children with weekly pain and seldom or no pain, adjusted for background factors; A1, A further adjusted for concentration problems; A2, A further adjusted for perceived problems with academic achievement; *P < 0.05.

First, pain was positively correlated with each of the 3 potential mediators, that is, as the frequency of pain increased, the frequency of concentration problems, sleep problems, and PPAA increased as well (Fig. 2). Second, there was a negative correlation between 2 of the potential mediators and overall grade points, that is, as the frequency of problems with concentration and PPAA increased, the overall grade points decreased. Sleep problems were not correlated with the overall grade points. Thus, sleep problems were not tested as a potential mediator in model 3.

Finally, among girls, the association seen in model 2 between weekly pain in school year 6 and overall grade points in school year 9 remained but decreased by about 30% when problems with concentration were added in model 3 and by about 20% when PPAA was added (Fig. 2). In boys, the association between weekly pain in school year 6 and overall grade points in school year 9 disappeared when the 2 remaining potential mediators were added one by one into model 3.

3.3. Association between recurrent pain in school year 6 and secondary school eligibility in school year 9 3.3.1. Pain frequency, location, and co-occurrence

Independent of the frequency, location, and co-occurrence of pain, there was generally no association between the occurrence of pain in school year 6 and secondary school eligibility in school year 9. This was true for girls as well as boys (Table 4). The only exception was that girls who suffered from weekly stomachache in school year 6 had about 2 times higher odds of lacking secondary school eligibility when leaving elementary school in school year 9 than girls who never or seldom experienced stomachache.

Table 4 - Differences in secondary school eligibility (school year 9) between girls and boys with recurrent pain at different frequencies and locations in school year 6. Grade 6 Girls(N=758) Boys(N=809) N Crude, OR (95% CI) Adjusted*, OR (95% CI) N Crude, OR (95% CI) Adjusted*, OR (95% CI) Pain at any site  Seldom or never 171 1 1 247 1 1  Monthly 253 1.14 (0.54-2.39) 1.02 (0.48-2.18) 274 1.17 (0.61-2.22) 1.02 (0.52-2.00)  Weekly 253 1.24 (0.62-2.50) 1.06 (0.52-2.19) 274 1.21 (0.64-2.27) 1.08 (0.56-2.08) Headache  Seldom or never 297 1 1 369 1 1  Monthly 243 0.69 (0.36-1.33) 0.63 (0.32-1.25) 250 1.10 (0.60-2.01) 0.99 (0.53-1.81)  Weekly 297 1.05 (0.34-1.65) 0.95 (0.50-1.78) 190 1.12 (0.53-2.39) 1.34 (0.71-2.53) Stomachache  Seldom or never 353 1 1 449 1 1  Monthly 240 1.60 (0.85-3.02) 1.55 (0.81-2.97) 247 1.52 (0.88-2.64) 1.37 (0.77-2.43)  Weekly 166 2.28 (1.19-4.37)† 2.09 (1.06-4.12)† 113 1.14 (0.48-2.67) 1.20 (0.55-2.60) Backache  Seldom or never 549 1 1 599 1 1  Monthly 116 0.99 (0.47-2.09) 0.80 (0.36--1.78) 119 0.55 (0.23-1.30) 0.47 (0.19-1.15)  Weekly 94

留言 (0)

沒有登入
gif