- Email: [email protected]
Prognosis of non-specific musculoskeletal pain in preadolescents: A prospective 4-year follow-up study till adolescence
Pain 110 (2004) 550–559 www.elsevier.com/locate/pain
Prognosis of non-specific musculoskeletal pain in preadolescents: A prospective 4-year follow-up study till adolescence Ashraf El-Metwallya,b,*, Jouko J. Salminenc, Anssi Auvinenb, Hannu Kautiainena, Marja Mikkelssona a
Department of Physical and Rehabilitation Medicine, The Rheumatism Foundation Hospital, Pikija¨rventie 1, 18120 Heinola, Finland b Tampere School of Public Health, FIN-33014 University of Tampere, Tampere, Finland c Department of Physical and Rehabilitation Medicine, University Hospital of Turku, P.O box 52, 20520 Turku, Finland Received 9 September 2003; received in revised form 11 March 2004; accepted 11 March 2004
Abstract Musculoskeletal pain is common in children but studies on the outcome and predictive factors for persistence/recurrence of these symptoms are scarce. A baseline cross-sectional survey of 1756 schoolchildren (mean age 10.8) identified 564 (32.1%) children with musculoskeletal pain. At baseline, these children were evaluated using a structured questionnaire and examined for hypermobility and physical fitness. The children were re-evaluated after one year, and four years (at adolescence) using the same pain questionnaire. At 1-year follow-up, 53.8% (95% CI 48.8 – 58.8) of the children reported pain persistence (persistent preadolescent musculoskeletal pain). At 4-year follow-up, 63.5% (95% CI 58.7– 68.1) of them had musculoskeletal pain. Neck was the site with most persistent/recurrent musculoskeletal pain. Those with persistent preadolescent musculoskeletal pain had approximately three times higher risk of pain recurrence (OR ¼ 2.90 [95% CI 1.9– 4.4]). In the univariate analysis, female gender, older age group (11 þ ), hypermobility, co-existence of psychosomatic symptoms, having high disability index, and reporting combined musculoskeletal pain at baseline predicted pain recurrence at adolescence. In the multivariate analysis, age, headache, hypermobility and having combined musculoskeletal pain were found as independent predictors. Statistically significant sex interactions were found for age, depressive feelings, waking up during nights and hypermobility. More psychosomatic symptoms predicted pain recurrence in girls than in boys, and hypermobility was a strong predictor in females only. Musculoskeletal pain in preadolescents is not a self-limiting phenomenon and more studies are still warranted to explore its determinants aiming to improve the long-term outcome of these symptoms. q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Musculoskeletal Pain; Preadolescents; Children; Hypermobility; Prognosis
1. Introduction Several recent studies have suggested a possible association between musculoskeletal pain in childhood and development of musculoskeletal disorders in adults (Harreby et al., 1995; Leboeuf-Yde and Kyvik, 1998; Salminen et al., 1999). Hence, a better understanding of determinants of these pain symptoms in children is important, not only for developing effective preventive strategies to the youth, but also for providing a better understanding of the origin of chronic pain in adults (Goodman and McGrath, 1991). * Corresponding author. Address: Tampere School of Public Health, FIN33014 University of Tampere, Tampere, Finland. Tel.: þ358-0504630701; fax: þ 358-38491516. E-mail address: [email protected] (A. El-Metwally).
Previous studies have shown that musculoskeletal pain is common in pre-adult populations. Prevalence varies by age and case definition of musculoskeletal pain. In a study conducted in Denmark, 38% of 16 – 19 year olds reported musculoskeletal pain during the past two weeks (Kamper-Jorgensen et al., 1992). In our previous study in ages 10– 12, 32% reported musculoskeletal pain during the preceding 3 months (Mikkelsson et al., 1997). Furthermore, biennial nationwide postal surveys (1985 – 2001), and annual classroom surveys (1996 – 2001) in Finland have shown that musculoskeletal pain symptoms are becoming more common in adolescents (Hakala et al., 2002). Current evidence indicates that some factors might be causally related to the onset of these complaints in schoolchildren. These include the child’s age and gender, anthropometric parameters, physical activity, after-school
0304-3959/$20.00 q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2004.03.021
A. El-Metwally et al. / Pain 110 (2004) 550–559
work, heavy backpack, smoking, psychological factors and family history of musculoskeletal pain (Balague´ et al., 1999). However, a prognostic factor, that can potentially predict the disease course subsequent to the onset, need not necessarily be a risk factor for the disease incidence (Laupacis et al., 1994). Studies on the natural course of musculoskeletal pain in childhood are scarce and most of the previous studies had the limitations of small, hospital-based study populations. However, in a prospective 3- and 9-year follow-up study of 14-year old children initially complaining of low back pain, 35% of them still reported recurrent LBP at 18 years of age and up to early adulthood. Early degeneration of lower lumbar discs was the most significant predictive factor for LBP persistence in this controlled study (Salminen et al., 1999). In another 9– 12 year follow-up study of 16-year-old Norwegian adolescents, 59% of baseline subjects had neck pain and 56% had lumbar pain. Muscular tenderness in the neck/shoulder area was the most significant risk factor for later cervical pain (Hertzberg, 1985). However, both studies were limited to a specific local pain in adolescence. Cooccurrence of different localized pain states in the same child has been shown in earlier studies (Øster, 1972; Salminen, 1984; Mikkelsson et al., 1997). Hence, further research is still needed using a younger age group, with a sufficient sample size and a wider case definition of musculoskeletal pain. This study is a continuation of a previous survey of a representative preadolescent population. The results on musculoskeletal pain and risk factors at baseline and at 1-year follow-up have been reported earlier (Mikkelsson et al., 1997,1998). We evaluated the same cohort 3 years later to determine to what extent musculoskeletal pain recurred at adolescence. We also investigated which baseline factors could contribute to recurrence of musculoskeletal pain at 4-year follow-up.
2. Materials and methods 2.1. The study population The initial study took place in Lahti, a town of 97 689 (2002) inhabitants in southern Finland. Nineteen of the 21 primary schools in the town took part in the study. The Steiner school, the hospital school, and the schools for hearing and physically disabled and the mentally handicapped were excluded because some of the methods used in this study were not suitable for their pupils. All pupils from the third and fifth grades and present at school on the day of the survey participated in the study. The initial sample consisted of 867 third-grade (mean age 9.8 [SD ¼ 0.4] years) and 889 fifth-grade schoolchildren (mean age 11.8 [SD ¼ 0.4] years), representing 82.9% of all schoolchildren in these grades in Lahti (Fig. 1). These children completed a pain questionnaire and were tested for hypermobility and
551
Fig. 1. Flow chart of the study.
cardio-respiratory fitness in March 1995. Of them 564 children (32%) reported musculoskeletal pain symptoms. These children constituted our study population and were re-evaluated after one year and four years to determine persistence/recurrence of preadolescent musculoskeletal pain and investigate factors contributing to it. 2.2. Instruments 2.2.1. Pain questionnaire A structured pain questionnaire was designed to asssess musculoskeletal pain during the previous three months. Musculoskeletal pain symptoms were classified according to body area (neck, upper limb, chest, lower limb, upper back, lower back, buttock) and frequency (pain seldom or never, once a month, once a week, more than once a week, almost daily). The 5-level frequency classification was adopted from the questionnaire used in the nationwide survey on health and health-related behaviors in schoolchildren by the WHO (King et al., 1996). Psychosomatic symptoms (headache, abdominal pain, depressive mood, day tiredness, difficulties in falling asleep, waking up during nights) were asked about with the same frequency categorization as for musculoskeletal pain. Each symptom was considered positive if it was present at least once a week. Disability due to pain was assessed by the following questions (1) do you have difficulties in falling asleep
552
A. El-Metwally et al. / Pain 110 (2004) 550–559
because of your pain or does your pain disturb your sleep (2) do you have difficulties while sitting during lessons (3) do you feel pain if you walk more than one kilometer (4) do you feel pain during physical exercise class (5) does your pain disturb your hobbies. Preadolescents were categorized into 3 groups according to their pain disability symptoms (no symptoms, 1 –2 symptoms and 3 –5 disability symptoms). Absence from school due to pain or aches was assessed to ascertain functional limitations. However, this question did not distinguish musculoskeletal pain from other types of pain (e.g. headache, abdominal pain). The amount of physical exercise was evaluated and children were categorized into 2 groups according to its frequency (less than 3 times a week and 3 or more times a week). The pain questionnaire included also questions about injury or trauma to the reported pain area, other diseases, and regular use of medications. During the design procedure, two versions of the questionnaire were tested in two different schools in Nastola, a neighboring community in Lahti, Finland and the test-retest reliability of the final version evaluated by repeating the survey at one-week interval (Mikkelsson et al., 1996). The test-retest reliability of the questionnaire in pain at least once a week was good (k 0.9). The concurrent validity of the pain questionnaire was examined by comparing it with interviews of 31 third- and 25 fifthgrade children. The observed agreement between pain questionnaire and interview technique was 86% (95% confidence interval [95% CI] 74 –94%) and k was 0.67 (Mikkelsson et al., 1997). 2.2.2. Hypermobility test Hypermobility test was conducted only at baseline, using Beighton’s method (Beighton et al , 1973). No stretching was allowed before the test. Intra and inter-observer reliability were measured earlier with a k coefficient of 0.75 and 0.78, respectively (Mikkelsson et al., 1996). Out of 403 children included in the analysis, 381 (94.5%) were examined for hypermobility. School absence on the examination date was the reason for non-participation. Beighton score of six was chosen as the cutoff point for hypermobility on the basis of the distribution of the results. 2.2.3. Shuttle run test As the hypermobility test, the shuttle run test (Leger and Lambert, 1982) was conducted only at baseline It measures maximal performance indoors and provides a valid and reliable index of cardio-respiratory endurance or maximal oxygen uptake (VO2max) (Committee of experts on sport research 1988). It was performed to all the study subjects except to children with acute musculoskeletal injury, acute respiratory infection, or other diseases inhibiting maximal physical strain. Children from schools that did not have room for the 20-m distance, needed to perform this test, were also excluded. Out of 403 children, who were included in the analysis, 368 (91.3%) underwent the shuttle run test.
2.3. Case definition of musculoskeletal pain Those who reported musculoskeletal pain in one or more area (s) at least once a week, and did not report injuring the pain site, fulfilled the case definition of musculoskeletal pain at baseline and constituted the study population for the follow-up study. The same case definition of musculoskeletal pain was used at the follow-up assessments. We have excluded children who reported injuring the pain area as, in this report, we are specifically investigating the outcome and contributing factors of recurrence of non-traumatic preadolescent musculoskeletal pain. It must be noted that we have only excluded children who reported directly injuring the musculoskeletal area with pain while overuse injuries were included in our case definition of musculoskeletal pain at both preadolescence and adolescence. 2.4. Follow-up The one-year follow-up was conducted in March 1996 with the same pain questionnaire as at baseline. At this follow-up, 515 (91.3%) of the children who had reported musculoskeletal pain at baseline filled in the pain questionnaire. At 4-year follow-up, 430 (76.2%) of the children in our original study population completed the pain questionnaire again. However, 27 children had incomplete data and were excluded. The final follow-up study population of 403 children with complete data was used in the analysis. These 403 children represented 71.5% of the original 564 subjects with musculoskeletal pain at baseline. Reasons for non-participation in the follow-ups included absence from school on the day of testing, changing schools, moving away and refusing to participate. Children who constituted our final study population were compared with those who were lost to follow-up or had incomplete data (Table 1). Both groups were similar with respect to almost all the baseline variables. The mean age (10.8 versus 10.9 years) and the mean VO2 max (34.4 versus 33.5 ml/kg/min) were comparable in the two groups. Furthermore, a multivariate logistic regression analysis was conducted to assess the impact of various baseline factors on the odds that a child would be lost to follow-up, and none of the baseline variables significantly influenced the rate at which children dropout. 2.5. Statistical methods The statistical analysis consisted of descriptive statistics with the most important values expressed with 95% CI. General linear models for the binomial family with log link function (risk ratio [RR]) were used to identify predictive factors for recurrence of musculoskeletal pain in adolescence. Musculoskeletal pain at 4-year follow-up was used as the dependent variable in all the models i.e. case was defined as recurrence of musculoskeletal pain at four years from baseline. The analysis included the following
A. El-Metwally et al. / Pain 110 (2004) 550–559 Table 1 Comparison of baseline data of included subjects and those not included in the analysis Variables
Included in the study
Lost at follow-up
Total, N Males, N (%)
403 194 (48.1)
161 79 (49.1)
Musculoskeletal pain, N (%) Neck pain Upper limb pain Chest pain Lower limb pain Upper back pain Lower back pain Buttock pain
187 (46.4) 64 (15.9) 68 (16.8) 219 (54.3) 70 (17.3) 47 (11.7) 25 (6.2)
73 43 24 85 37 23 13
Psychosomatic symptoms, N (%) Headache Abdominal pain Depressive mood
214 (53.0) 148 (36.5) 161 (40.0)
34 (56.5) 55 (34.2) 68 (42.2)
Difficulty in falling asleep Day tiredness Waking up during nights
147 (36.5) 228 (56.6) 110 (27.3)
72 (44.7) 98 (60.9) 53 (32.9)
Pain Disability symptoms, N (%) Pain disturbs sleeping Pain disturbs sitting during lessons Pain disturbs walking more than 1 km Pain disturbs physical exercise class Pain disturbs hobbies
119 (29.5) 62 (15.4) 129 (32.0) 105 (26.1) 90 (22.3)
51 26 47 44 52
Frequency of exercisea, N (%) Hypermobilityb, N (%) School absence due to pain, N (%)
202 (50.1) 30 (7.4) 101 (25.1)
78 (48.4) 7 (4.0) 43 (26.7)
(45.3) (27.0) (14.9) (52.8) (23.0) (14.3) (8.1)
(31.7) (16.1) (29.2) (27.3) (32.3)
553
3. Results At 1-year follow-up, 53.8% (95% CI 48.8 – 58.8) reported musculoskeletal pain at least once a week in at least one part of the body. At 4-year follow-up (at adolescence), 63.5% (95% CI 58.7– 68.1) had pain at least once a week in at least one musculoskeletal site. Musculoskeletal pain was reported by 74.7% (95% CI 68.5 – 80.0) of those who reported musculoskeletal pain at 1year follow-up and by 50.5% (95% CI 43.3 –57.6) of those who reported musculoskeletal pain at baseline only. Those who reported musculoskeletal pain at both baseline and 1year follow-up (persistent preadolescent musculoskeletal pain) had approximately three times (OR ¼ 2.9 [95% CI 1.9 – 4.4]) higher risk of musculoskeletal pain recurrence at adolescence compared to those who reported musculoskeletal pain at baseline only (Fig. 2). The most persistent/recurrent musculoskeletal pain site was the neck. Of 187 children reporting neck pain (as a single musculoskeletal pain site or combined with other musculoskeletal pain symptoms) at baseline, 90 (48.1% [95% CI 40.7 – 55.5]) had neck pain at 1-year follow-up and 98 (52.4% [95% CI 45.2– 59.5]) at 4- year follow-up (Fig. 3). At both 1-and 4-year follow-ups, children with regional musculoskeletal pain at baseline (with a single affected musculoskeletal site) had better prognosis than those with combined musculoskeletal pain (with several affected
a
Exercised more than 3 times a week. A child was categorized as hypermobile if his/her hypermobility score was six or higher. b
independent variables, age (below/above 11 years), psychosomatic symptoms (present/absent), disability index (0/1 – 2/3 –5), frequency of exercise (less/more than three times a week), Beighton score of hypermobility (below/above six), musculoskeletal pain location (regional/combined), school absence due to pain (never/once or more), and VO2 max (continuous). Regression models were first fitted with each of these variables separately, together with age, to estimate the age-adjusted predictive effect for each of them then a multivariate analysis was conducted using a backward stepwise method. All independent variables were initially included in the regression equation and elimination of variables, at each step, was based on the likelihood ratio test at 10% level of significance. Modification by sex was assessed by evaluating interactions between sex and other potential predictive factors. In all tests, a P value of less than 0.05 (two tailed) was considered statistically significant. All statistical analyses were performed using the SPSS (for Windows), version 10.0.
Fig. 2. Proportions of persistent/recurrent preadolescent musculoskeletal (MS) pain. 1-and 4-year follow-up assessments of schoolchildren in Lahti, Finland.
554
A. El-Metwally et al. / Pain 110 (2004) 550–559
Fig. 3. Persistence/recurrence proportions of preadolescent musculoskeletal (MS) pain symptoms at 1-year follow-up (1996) and 4-year follow-up (1999) of schoolchildren in Lahti, Finland.
musculoskeletal pain sites) (Fig. 4). Within those who had combined musculoskeletal pain at baseline and reported persistent/recurrent musculoskeletal pain, pain remained combined in the majority at both 1-year follow-up (66.7%) and 4-year follow-up (62.9%). On the other hand, within those who had regional musculoskeletal pain at baseline and reported persistent/recurrent musculoskeletal pain, pain was reported in several musculoskeletal sites by 35% of them at 1-year follow-up then by 51% of them at 4-year follow-up. Children who reported recurrent musculoskeletal pain at 4-year follow-up were older, had higher proportion of combined preadolescent musculoskeletal pain and hypermobility compared to children free of musculoskeletal pain
at follow-up. In addition, children with recurrent musculoskeletal pain had higher proportion of persistent preadolescent musculoskeletal pain and had more psychosomatic symptoms at baseline (Table 2). In the univariate analysis, children with combined musculoskeletal pain at baseline had a 24% higher ageadjusted risk of pain recurrence compared to those with regional musculoskeletal pain. Similarly, girls had a 24% higher risk of pain recurrence than boys and hypermobile children had a 29% higher recurrence risk compared to children with normal joint laxity. In addition to that, children aged 11– 13 years at baseline were more likely to develop musculoskeletal pain recurrence than those aged 9 – 10 and children with high subjective disability index were also at a higher risk. From the psychosomatic symptoms, headache, abdominal pain, and depressive feelings were predictive of musculoskeletal pain recurrence. In the multivariate analysis, only age (RR ¼ 1.28 [95% CI 1.10 – 1.49]), headache (RR ¼ 1.28 [95% CI 1.08– 1.51]), hypermobility (RR ¼ 1.35 [95% CI 1.08– 1.68]), and having combined musculoskeletal pain at baseline (RR ¼ 1.18 [95% CI 1.02 – 1.36]) remained statistically significant. Frequency of exercise, physical fitness and school absence due to pain did not predict the recurrence of musculoskeletal pain (Table 3). When analysed by sex, headache, abdominal pain and having combined musculoskeletal pain predicted pain recurrence among boys, and headache was the only statistically significant predictor in the multivariate analysis. The predictive factors in girls were age, abdominal pain, depressive feelings, waking up during nights and hypermobility. However, only age and depressive feelings and hypermobility were statistically significant in the backward step-wise regression. Statistically significant sex
Fig. 4. Proportions of persistent/recurrent musculoskeletal (MS) pain by number of involved sites at baseline (regional versus combined). 1-and 4-year followups of preadolescents in Lahti, Finland.
A. El-Metwally et al. / Pain 110 (2004) 550–559 Table 2 Baseline variables for adolescents with and without musculoskeletal pain at 4-year follow-up Variables
Musculoskeletal pain at adolescence
Age (mean)
10.9
Sex, N (%) Males Females
112 (43.8) 144 (56.2)
Musculoskeletal pain, N (%)a One area Two or more areas
No musculoskeletal pain at adolescence
Table 3 Relative risk (RR) and 95% confidence intervals (95% CI)) for recurrence of musculoskeletal pain from preadolescence till adolescence (403 children) Variables
Univariate analysis a RR (95% CI)
82 (55.8) 65 (44.2)
Sex Boys Girls
Reference 1.24 (1.07–1.44)
Age (years) 9– 10 11 –13
Reference 1.29 (1.11–1.51)
Reference 1.28 (1.10–1.49)
1.24 (1.06–1.43) 1.25 (1.09–1.43) 1.15 (1.00–1.32) 1.05 (0.90–1.21) 1.01 (0.87–1.17) 1.12 (0.96–1.30)
1.28 (1.08–1.51)
99 (67.3) 48 (32.7)
Pain at 1-year follow-up
162 (63.3)
55 (37.4)
Psychosomatic symptoms, N (%)a Headache Abdominal pain Depressive feelings Difficulty in falling asleep Day tiredness Waking up during nights
Psychosomatic symptoms Headache Abdominal pain Depressive feelings Difficulty falling asleep Day tiredness Waking up during nights
150 (59.1) 105 (41.8) 113 (44.7) 94 (37.3) 146 (57.0) 76 (30.0)
64 (44.1) 43 (29.5) 48 (33.1) 53 (36.1) 82 (56.2) 34 (23.4)
Disability index 0 1– 2 3– 5
Pain Disability symptoms, N (%) Pain disturbing sleeping Pain disturbing sitting Pain disturbing activity Pain disturbing walking Pain disturbing hobbies
Multivariate analysis RR (95% CI)
b
10.5
131 (51.2) 125 (48.9) b
555
c
Amount of exercise Less than 3 times a week More than 3 times a week
Reference 1.02 (0.86–1.21) 1.23 (1.02–1.49) Reference 1.00 (0.86–1.15)
(30.0) (17.2) (31.6) (25.8) (25.4)
42 (28.6) 18 (12.2) 48 (31.3) 39 (26.5) 25 (17.0)
Beighton score for hypermobility ,6 Reference 6– 9 1.23 (1.02–1.49)
Reference 1.35 (1.08–1.68)
Pain location at baselined Regional Combined
Reference 1.24 (1.08–1.42)
Reference 1.18 (1.02–1.36)
77 (30.0) 133 (52.0) 46 (18.0)
50 (34.0) 81 (31.3) 16 (10.9)
School absence due to pain Never absent Absence one day or more
Reference 1.08 (0.92–1.26)
VO2 max (per unit increase)
1.00 (0.98–1.01)
77 44 81 66 65
c
Disability index 0 1–2 3–5
Beighton score for hypermobility ,6 220 (90.5) 6–9 23 (9.5)
131 (94.9) 7 (5.1)
Frequency of exercise, N (%) Less than 3 times a week 127 (50.1) 3 or more times a week 123 (49.2)
74 (51.7) 69 (48.3)
School absence due to pain, N (%) VO2 max (mean)
67 (26.1)
34 (23.1)
34.6
34.1
a
Except for the variable “age”, all other figures were age adjusted. Using backward stepwise procedure. Only entered variables shown. c Number of pain disability symptoms. d Pain in one (regional) versus more than one (combined) musculoskeletal site at baseline. b
higher age-adjusted risk of recurrence within girls in the multivariate analysis. Out of 18 hypermobile girls who had preadolescent musculoskeletal pain at baseline, 17 reported musculoskeletal pain at 4-year follow-up.
a
Symptoms reported in a frequency of at least once a week. Those who had musculoskeletal pain at both baseline and 1-year follow-up (persistent preadolescent musculoskeletal pain). c Number of pain disability symptoms reported. b
interactions were found for age ðP , 0:001Þ; depressive feelings ðP ¼ 0:002Þ; waking up during nights ðP ¼ 0:049Þ and hypermobility ðP ¼ 0:037Þ: These factors predicted pain recurrence in girls but not in boys (Table 4). Hypermobility was associated with approximately a 30%
4. Discussion This population-based follow-up study of preadolescents with non-specific musculoskeletal pain shows that these symptoms are not as transient and self-limiting as previously assumed. Almost two thirds complained of musculoskeletal pain at 4-year follow-up. Children with regional musculoskeletal pain had better prognosis than
556
A. El-Metwally et al. / Pain 110 (2004) 550–559
Table 4 Relative risk (RR) and 95% confidence intervals (95% CI)) for recurrence of musculoskeletal pain from preadolescence till adolescence. Separate analysis for boys and girls Univariate analysis a RR (95% CI)
Multivariate analysis b RR (95% CI)
Boys
Girls
Boys
Age (years) 9– 10 11– 13
Reference 1.11 (0.87–1.42)
Reference 1.47 (1.21–1.78)
Psychosomatic symptoms Headache Abdominal pain Depressive feelings Difficulty falling asleep Day tiredness Waking up during nights
1.43 (1.12–1.81) 1.27 (1.00–1.61) 1.00 (0.78–1.29) 0.98 (0.75–1.26) 1.10 (0.86–1.41) 1.03 (0.79–1.35)
1.05 (0.89–1.24) 1.20 (1.03–1.40) 1.21 (1.03–1.42) 1.08 (0.92–1.27) 0.91 (0.78–1.07) 1.18 (1.01–1.37)
Disability indexc 0 1– 2 3– 5
Reference 1.25 (0.93–1.69) 1.37 (0.94–2.01)
Reference 0.91 (0.75–1.09) 1.13 (0.95–1.34)
More than 3 times a week
Reference 0.99 (0.78–1.27)
Reference 1.04 (0.89–1.26)
Beighton score for hypermobility ,6 6– 9
Reference 0.88 (0.50–1.56)
Reference 1.19 (1.00–1.41)
Pain location at baseline d Regional Combined
Reference 1.32 (1.04–1.66)
Reference 1.11 (0.95–1.31)
School absence due to pain Never absent Absence one day or more VO2 max (per unit increase)
Reference 1.05 (0.80–1.36) 1.01 (0.99–1.03)
Reference 1.13 (0.97–1.33) 0.99 (0.97–1.02)
Variables
Girls
Reference 1.40 (1.17–1.67) 1.43 (1.12–1.83) 1.24 (1.07 –1.43)
Amount of exercise Less than 3 times a week
a b c d
Reference 1.31 (1.18–1.46)
Except for the variable “age”, all other figures were age adjusted. Using backward stepwise procedure. Only entered variables shown. Number of pain disability symptoms. Pain in one (regional) versus more than one (combined) musculoskeletal site at baseline.
those with combined musculoskeletal pain. In addition to that, preadolescents who had regional musculoskeletal pain at baseline reported combined more than regional musculoskeletal pain at adolescence. This might indicate that musculoskeletal pain tends to be more generalized as the child grows up. It is difficult to compare these results with the descriptive results of the few previous follow-up studies. This is due to the differences in the children’s age at baseline, the length of follow-up and the case definitions of musculoskeletal pain. Moreover, the study subjects, in most of the previous prognostic studies, were drawn from hospitals or rheumatology clinics. Clinical study subjects may not represent the population at large due to referral biases, which is likely to overestimate the risk of recurrence, as documented in adult patients with persistent pain (Crook et al., 1989). Yet, our 4year recurrence proportion of musculoskeletal pain of roughly 65% is in agreement with the recurrence proportion
of 59% found in a clinic-based 9-year follow-up study (Flatø, et al., 1997). The high recurrence proportion in our study should be interpreted with caution due to the widecase definition and the short time recall of musculoskeletal pain. However, musculoskeletal pain is an intermittent phenomena and the pattern of recurrence rather than continuity may be a more realistic description of an individual’s experience of a chronic musculoskeletal problem (Deyo, 1993). Another possibility for our high recurrence figures would be that we evaluated musculoskeletal pain symptoms during winter when symptoms can be more frequent (Takala et al., 1992; Ehrmann Feldman et al., 2002). The 30% recurrence proportion of low-back pain in our study is somewhat lower than that of a follow-up study of 14-year-old children with low back pain (LBP) at baseline (Salminen et al., 1995). In their sample, 29% of the boys and 60% of the girls had recurrent LBP at 3-year follow-up, and 35% after 9 years from baseline (Salminen
A. El-Metwally et al. / Pain 110 (2004) 550–559
et al., 1999). In another population-based follow-up study, 27% of children with fibromyalgia at baseline still had similar symptoms at 30 months follow-up (Buskila et al., 1995). As mentioned above, the comparison of the studies is difficult. However, all these studies indicate that there are subpopulations among children with a tendency to develop chronic pain. Our univariate results show a high risk of musculoskeletal pain recurrence among girls, children who were older and children with high disability index at baseline. Furthermore, hypermobile children and children with psychosomatic symptoms and generalized musculoskeletal symptoms were at increased risk of recurrent pain. A strong association between psychosomatic symptoms and musculoskeletal pain in schoolchildren was also found in a recent cross-sectional study in a nationally representative sample of 11276 12 – 18-year-old Finnish adolescents (Vikat et al., 2000). In the multivariate analysis, age, headache, hypermobility and multiple painful musculoskeletal areas were the strongest predictive factors for pain recurrence. These results are in accordance with the previous cross-sectional studies, which have shown that prevalence of musculoskeletal pain in older children is higher than in younger (Salminen, 1984; Balague et al., 1988, 1994; Troussier et al., 1994; Burton et al., 1996) and that headache and musculoskeletal pain in children frequently co-exist (Sherry et al., 1991; Anttila et al., 2002). However, the crosssectional design strongly limits the conclusions of these earlier studies. The contributing factors of musculoskeletal pain recurrence are to a great extent similar to our previous results based on one-year follow-up. However, in this study we have found a statistically significant association between hypermobility at preadolescence and musculoskeletal pain recurrence at adolescence. This is in contrast to our previous findings, where hypermobility was not associated with preadolescent musculoskeletal at baseline nor was it predictive of musculoskeletal pain persistence at 1-year follow-up (Mikkelsson et al., 1996; Mikkelsson et al., 1997; Mikkelsson et al., 1998). An association between hypermobility and different musculoskeletal symptoms has been previously reported in a number of studies on schoolchildren (Beighton et al., 1973; Gedalia et al., 1991, 1993). However, we are not aware of any previous reports on the role of hypermobility in predicting recurrence of childhood musculoskeletal pain in adolescence. Our study shows differences in the factors predicting recurrent musculoskeletal pain between boys and girls. Age, depressive feelings, waking up during nights and hypermobility were predictive factors only in girls. Gender differences in the determinants of paediatric and adolescent chronic pain have been previously reported (Passchier and Orlebeke, 1985; Mirlijnet al., 2003). More psychosomatic symptoms predicted musculoskeletal pain recurrence in girls compared to boys. This is in line with earlier studies
557
where a relatively stronger association between psychosomatic-psychosocial factors and musculoskeletal pain in girls was found (Sherry et al., 1991; Merlijn et al., 2003). Joint hypermobility strongly predicted pain recurrence in girls but not in boys. Females show a greater joint mobility at all ages (Giannini and Brewer, 1982), and hypermobility syndrome is more prevalent among females (Beighton et al., 1973; Jessee el al., 1980). However, confirming the prognostic relevance of hypermobility in children needs a larger sample size of hypermobile children complaining of musculoskeletal pain at baseline in order to have enough statistical power for evaluating the difference between genders as regard to musculoskeletal pain recurrence. We are using the term persistence to describe those who reported musculoskeletal pain at 1-year follow-up and the term recurrence for those who reported musculoskeletal pain at 4-year follow-up. These might not be the correct terms for both. Occurrence of musculoskeletal pain in a frequency of at least once a week during the 3 months preceding the 1-year follow-up assessment does not necessarily indicate that the child had musculoskeletal pain with the same frequency during the whole preceding year, which might be assumed by using the term persistence. On the other hand, occurrence of musculoskeletal pain at 4-year follow-up might have been persistent musculoskeletal pain if the child had had pain during the whole 4-year follow-up period. Unfortunately, we did not have this information. Our questionnaire evaluated the occurrence of musculoskeletal pain only during the preceding 3 months. We have used this rather short time scale to avoid difficulties with recall of pain. Recall problems have been reported in several previous studies which have used a protracted recall time scale in collecting data about pediatric and adolescent pain (Goodman and McGrath, 1991; Savedra et al., 1988; Ehrmann Feldman et al., 2002). By using the term ‘psychosomatic symptoms’ we mean symptoms that are chiefly considered to be expressions of psychological stress and only rarely associated with any organic disease in schoolchildren. Depressive mood is certainly a psychological symptom. Childhood abdominal pain and headache are believed to be having a psychosomatic origin in the great majority of cases (Bury RG. 1987; Alfven G. 1986). However, it might be questionable whether difficulty in falling asleep, waking up during nights and day tiredness were manifestations of psychological stress or consequences of chronic pain. The rationale of excluding pain preceded by injury from our case definition of musculoskeletal pain is our belief that determinants and prognosis of musculoskeletal pain complaints related to trauma (falling down, sports injuries) might be different from non-traumatic pain complaints. Previous studies on young athletes have shown that most injuries in children’s sports are minor and self-limiting (Maffulli and Baxter-Jones, 1995; Maffulli and Bruns, 2000). Including children who
558
A. El-Metwally et al. / Pain 110 (2004) 550–559
reported pain symptoms precipitated by trauma at adolescence might have spuriously over-estimated our recurrence proportion at 4-year follow-up, as it is most probably un-related to the past history of non-traumatic musculoskeletal pain at preadolescence. To our knowledge, the present study is the first prospective follow-up study of preadolescents with musculoskeletal pain till adolescence. Strengths of the study are it’s population based design; fairly long prospective followup; children’s self-report of pain, as was previously recommended in pediatric pain research (Goodman and McGrath, 1991); and the use of a valid pain questionnaire which classifies musculoskeletal pain according to frequency and location aided by a diagram. In addition, pain interference with daily activities was assessed. However, pain intensity was not evaluated and we did not gather information about anatomic/structural factors (except for hypermobility) or socio-demographic factors. These factors might be related both to the onset and prognosis of musculoskeletal pain in children (Salminen et al., 1999; Aasland et al., 1997; Balague´, 1999). In conclusion, the current findings indicate that nonspecific musculoskeletal pain in preadolescents is not a selflimiting phenomenon. The persistence of musculoskeletal pain in preadolescence was strongly associated with pain recurrence in adolescence. As sex was a significant effect modifier for many baseline variables, future studies identifying determinants of musculoskeletal pain in children and adolescents should test for sex interactions. Although this study did not test for psychological variables (except for depressive feelings), the strong association between psychosomatic symptoms and musculoskeletal pain recurrence indicates that psychological issues should be evaluated with pediatric musculoskeletal pain to determine if psychological intervention should be incorporated.
Acknowledgements This study was funded by PATU Development Project and the Medical Research Fund of the Rheumatism Foundation Hospital. We thank Tuija Sulonen, RN, for assistance in collecting the data at follow-up. We also thank Riina Haataja, BA and Rahman Shiri, MD for commenting on the statistical analyses.
References Aasland A, Flatø B, Vandvik IH. Psychosocial factors in children with idiopathic musculoskeletal pain: a prospective, longitudinal study. Acta Paediatr 1997;86:740– 6. Alfven G., Psychological tension headaches among children_a survey. La¨kartidningen, 1986;45:3827–8 (in Swedish). Anttila P, Metsahonkala L, Aromaa M, Sourander A, Salminen J, Helenius H, Alanen P, Sillanpaa M. Determinants of tension-type headache in children. Cephalalgia 2002;22:401–8.
Balague´ F, Dutoit G, Waldburger M. Low back pain in school children. An epidemiological study. Scand J Rehabil Med 1988;20: 175 –9. Balague´ F, Nordin M, Skovron ML, Dutoit G, Yee A, Waldburger M. Nonspecific low-back pain among schoolchildren: a field survey with analysis of some associated factors. J Spinal Disord 1994;7:74– 9. Balague´ F, Troussier B, Salminen JJ. Non-specific low back pain in children and adolescents: risk factors. Review Eur Spine J 1999;8: 429 –38. Beighton P, Solomon L, Soskolne CL. Articular mobility in an African population. Ann Rheum Dis 1973;32:413 –8. Burton AK, Clarke RD, McClune TD, Tillotson KM. The natural history of low back pain in adolescents. Spine 1996;20:2323–8. Bury RG. A study of 111 children with recurrent abdominal pains. Aust Paediatr J 1987;23:117–9. Buskila D, Neumann L, Hershman E, Gedalia A, Press J, Sukenik S. Fibromyalgia syndrome in children-an outcome study. J Rheumatol 1995;22:525–8. Committee of experts on sports research. Eurofit. Handbook for the Eurofit tests of physical exercise. Rome: Council of Europe, Committee for the Development of Sport, 1988 p. 24. Crook J, Weir R, Tunks E. An epidemiological follow-up survey of persistent pain sufferers in a group family practice and specialty pain clinic. Pain 1989;36:49–61. Deyo RA. Practice variations, treatment fads, rising disability. Do we need a new clinical research paradigm? Spine 1993;18:2153 –62. Ehrmann Feldman D, Shrier I, Rossignol M, Abenhaim L. Risk factors for the development of neck and upper limb pain in adolescents. Spine 2002;27:523–8. Flatø B, Aasland A, Vandvik IH, Forre O. Outcome and predictive factors in children with chronic idiopathic musculoskeletal pain. Clin Exp Rheumatol 1997;15:569–77. Gedalia A, Press J:. articular symptoms in hypermobile schoolchildren; a prospective study. J Pediatr 1991;119:944 –6. Gedalia A, Press J, Klein M, Buskila D. Joint hypermobility and fibromyalgia in schoolchildren. Ann Rheum Dis 1993;52:494–6. Giannini EH, Brewer Jr EJ. Standard methodology for Segment I, II, and III Pediatric Rheumatology Collaborative Study Group studies. II. Analysis and presentation of data. J Rheumatol 1982;9: 114 –22. Goodman JE, McGrath PJ. The epidemiology of pain in children and adolescents: a review. Pain 1991;46:247 –64. Hakala P, Rimpela A, Salminen JJ, Virtanen SM, Rimpela M. Back, neck, and shoulder pain in Finnish adolescents: national cross sectional surveys. Br Med J 2002;325:743–6. Harreby M, Neergaard K, Hesselsoe G, Kjer J. Are radiologic changes in the thoracic and lumbar spine of adolescents risk factors for low back pain in adults? A 25-year prospective cohort study of 640 schoolchildren. Spine 1995;20:2298 –302. Hertzberg A. Prediction of cervical and low-back pain based on routine school health examinations. A nine- to twelve-year follow-up study. Scand J Prim Health Care 1985;3:247–53. Jessee EF, Owen Jr DS, Sagar KB. The benign hypermobile joint syndrome. Arthritis Rheum 1980;23:1053– 6. Kamper-Jorgensen L, Mollenvang Nielsen M., Born og unge. Kropsbevidsthet. Forebyggelsesinnsats. Rapport fra et seminar om forebyggelse af og muskel- og skeletsygdomme hos born og unge (in Danish). Dansk institutt for Klinisk Epidemiologi (DIKE). Kobenhavn: 1992. pp. 13 –17. King A, Wold B, Tudor-Smith C, Harel Y. The Health of Youth: A Cross-sectional Survey. Geneva: WHO; 1996. p. 68– 9. European Series No. 69. Laupacis A, Wells G, Richardson WS, Tugwell P. Users’ guides to the medical literature V. How to use an article about prognosis. J Am Med Assoc 1994;272:234–7. Leboeuf-Yde C, Kyvik KO. At what age does low back pain become a common problem? A study of 29,424 individuals aged 12-41 years. Spine 1998;23:228–34.
A. El-Metwally et al. / Pain 110 (2004) 550–559 Leger LA, Lambert J. A maximal multistage 20-m shuttle run test to predict VO2 max. Eur J Appl Physiol Occup Physiol 1982;49:1–12. Maffulli N, Baxter-Jones ADG. Vommon skeletal injuries in young athletes. Sports Med 1995;19:137 –49. Maffulli N, Bruns W. Injuries in young athletes. Eur J Pediatr 2000;159: 59–63. Merlijn VP, Hunfeld JA, van der Wouden JC, Hazebroek-Kampschreur AA, Koes BW, Passchier J. Psychosocial factors associated with chronic pain in adolescents. Pain 2003;101:33– 43. Mikkelsson M, Salminen JJ, Kautiainen H. Joint hypermobility is not a contributing factor to musculoskeletal pain in pre-adolescents. J Rheumatol 1996;23:1963 –7. Mikkelsson M, Salminen JJ, Kautiainen H. Non-specific musculoskeletal pain in preadolescents. Prevalence and 1-year persistence. Pain 1997; 73:29–35. Mikkelsson M, Salminen JJ, Sourander A, Kautiainen H. Contributing factors to the persistence of musculoskeletal pain in preadolescents: a prospective 1-year follow-up study. Pain 1998;77:67–72. Øster J. Recurrent abdominal pain, headache and limb pains in children and adolescents. Pediatrics 1972;50:429–36. Passchier J, Orlebeke JF. Headaches and stress in schoolchildren: an epidemiological study. Cephalalgia 1985;5:167–76.
559
Salminen JJ. The adolescent back. A field survey of 370 Finnish school children. Acta Paediatr Scand 1984;315:1–122. Salminen JJ, Erkintalo M, Laine M, Pentti J. Low back pain in the young. A prospective three-year follow-up study of subjects with and without low back pain. Spine 1995;20:2101– 8. Salminen JJ, Erkintalo MO, Pentti J, Oksanen A, Kormano MJ. Recurrent low back pain and early disc degeneration in the young. Spine 1999;13: 1316–21. Savedra MC, Tesler MD, Ward JD, Wegner C. How adolescents describe pain. J Adolesc Health Care 1988;9:315–20. Sherry DD, McGuire T, Mellins E, Salmonson K, Wallace CA, Nepom B. Psychosomatic musculoskeletal pain in childhood: clinical and psychological analyses of 100 children. Pediatrics 1991;88:1093 –9. Takala EP, Viikari-Juntura E, Moneta GB, Saarenmaa K, Kaivanto K. Seasonal variation in neck and shoulder symptoms. Scand J Work Environ Health 1992;18:257–61. Troussier B, Davoine P, de Gaudemaris R, Fauconnier J, Phelip X. Back pain in school children. A study among 1178 pupils. Scand J Rehabil Med 1994;26:143–6. Vikat A, Rimpela M, Salminen JJ, Rimpela A, Savolainen A, Virtanen SM. Neck or shoulder pain and low back pain in Finnish adolescents. Scand J Public Health 2000;28:164–73.
Prognosis of non-specific musculoskeletal pain in preadolescents: A prospective 4-year follow-up study till adolescence Ashraf El-Metwallya,b,*, Jouko J. Salminenc, Anssi Auvinenb, Hannu Kautiainena, Marja Mikkelssona a
Department of Physical and Rehabilitation Medicine, The Rheumatism Foundation Hospital, Pikija¨rventie 1, 18120 Heinola, Finland b Tampere School of Public Health, FIN-33014 University of Tampere, Tampere, Finland c Department of Physical and Rehabilitation Medicine, University Hospital of Turku, P.O box 52, 20520 Turku, Finland Received 9 September 2003; received in revised form 11 March 2004; accepted 11 March 2004
Abstract Musculoskeletal pain is common in children but studies on the outcome and predictive factors for persistence/recurrence of these symptoms are scarce. A baseline cross-sectional survey of 1756 schoolchildren (mean age 10.8) identified 564 (32.1%) children with musculoskeletal pain. At baseline, these children were evaluated using a structured questionnaire and examined for hypermobility and physical fitness. The children were re-evaluated after one year, and four years (at adolescence) using the same pain questionnaire. At 1-year follow-up, 53.8% (95% CI 48.8 – 58.8) of the children reported pain persistence (persistent preadolescent musculoskeletal pain). At 4-year follow-up, 63.5% (95% CI 58.7– 68.1) of them had musculoskeletal pain. Neck was the site with most persistent/recurrent musculoskeletal pain. Those with persistent preadolescent musculoskeletal pain had approximately three times higher risk of pain recurrence (OR ¼ 2.90 [95% CI 1.9– 4.4]). In the univariate analysis, female gender, older age group (11 þ ), hypermobility, co-existence of psychosomatic symptoms, having high disability index, and reporting combined musculoskeletal pain at baseline predicted pain recurrence at adolescence. In the multivariate analysis, age, headache, hypermobility and having combined musculoskeletal pain were found as independent predictors. Statistically significant sex interactions were found for age, depressive feelings, waking up during nights and hypermobility. More psychosomatic symptoms predicted pain recurrence in girls than in boys, and hypermobility was a strong predictor in females only. Musculoskeletal pain in preadolescents is not a self-limiting phenomenon and more studies are still warranted to explore its determinants aiming to improve the long-term outcome of these symptoms. q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Musculoskeletal Pain; Preadolescents; Children; Hypermobility; Prognosis
1. Introduction Several recent studies have suggested a possible association between musculoskeletal pain in childhood and development of musculoskeletal disorders in adults (Harreby et al., 1995; Leboeuf-Yde and Kyvik, 1998; Salminen et al., 1999). Hence, a better understanding of determinants of these pain symptoms in children is important, not only for developing effective preventive strategies to the youth, but also for providing a better understanding of the origin of chronic pain in adults (Goodman and McGrath, 1991). * Corresponding author. Address: Tampere School of Public Health, FIN33014 University of Tampere, Tampere, Finland. Tel.: þ358-0504630701; fax: þ 358-38491516. E-mail address: [email protected] (A. El-Metwally).
Previous studies have shown that musculoskeletal pain is common in pre-adult populations. Prevalence varies by age and case definition of musculoskeletal pain. In a study conducted in Denmark, 38% of 16 – 19 year olds reported musculoskeletal pain during the past two weeks (Kamper-Jorgensen et al., 1992). In our previous study in ages 10– 12, 32% reported musculoskeletal pain during the preceding 3 months (Mikkelsson et al., 1997). Furthermore, biennial nationwide postal surveys (1985 – 2001), and annual classroom surveys (1996 – 2001) in Finland have shown that musculoskeletal pain symptoms are becoming more common in adolescents (Hakala et al., 2002). Current evidence indicates that some factors might be causally related to the onset of these complaints in schoolchildren. These include the child’s age and gender, anthropometric parameters, physical activity, after-school
0304-3959/$20.00 q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2004.03.021
A. El-Metwally et al. / Pain 110 (2004) 550–559
work, heavy backpack, smoking, psychological factors and family history of musculoskeletal pain (Balague´ et al., 1999). However, a prognostic factor, that can potentially predict the disease course subsequent to the onset, need not necessarily be a risk factor for the disease incidence (Laupacis et al., 1994). Studies on the natural course of musculoskeletal pain in childhood are scarce and most of the previous studies had the limitations of small, hospital-based study populations. However, in a prospective 3- and 9-year follow-up study of 14-year old children initially complaining of low back pain, 35% of them still reported recurrent LBP at 18 years of age and up to early adulthood. Early degeneration of lower lumbar discs was the most significant predictive factor for LBP persistence in this controlled study (Salminen et al., 1999). In another 9– 12 year follow-up study of 16-year-old Norwegian adolescents, 59% of baseline subjects had neck pain and 56% had lumbar pain. Muscular tenderness in the neck/shoulder area was the most significant risk factor for later cervical pain (Hertzberg, 1985). However, both studies were limited to a specific local pain in adolescence. Cooccurrence of different localized pain states in the same child has been shown in earlier studies (Øster, 1972; Salminen, 1984; Mikkelsson et al., 1997). Hence, further research is still needed using a younger age group, with a sufficient sample size and a wider case definition of musculoskeletal pain. This study is a continuation of a previous survey of a representative preadolescent population. The results on musculoskeletal pain and risk factors at baseline and at 1-year follow-up have been reported earlier (Mikkelsson et al., 1997,1998). We evaluated the same cohort 3 years later to determine to what extent musculoskeletal pain recurred at adolescence. We also investigated which baseline factors could contribute to recurrence of musculoskeletal pain at 4-year follow-up.
2. Materials and methods 2.1. The study population The initial study took place in Lahti, a town of 97 689 (2002) inhabitants in southern Finland. Nineteen of the 21 primary schools in the town took part in the study. The Steiner school, the hospital school, and the schools for hearing and physically disabled and the mentally handicapped were excluded because some of the methods used in this study were not suitable for their pupils. All pupils from the third and fifth grades and present at school on the day of the survey participated in the study. The initial sample consisted of 867 third-grade (mean age 9.8 [SD ¼ 0.4] years) and 889 fifth-grade schoolchildren (mean age 11.8 [SD ¼ 0.4] years), representing 82.9% of all schoolchildren in these grades in Lahti (Fig. 1). These children completed a pain questionnaire and were tested for hypermobility and
551
Fig. 1. Flow chart of the study.
cardio-respiratory fitness in March 1995. Of them 564 children (32%) reported musculoskeletal pain symptoms. These children constituted our study population and were re-evaluated after one year and four years to determine persistence/recurrence of preadolescent musculoskeletal pain and investigate factors contributing to it. 2.2. Instruments 2.2.1. Pain questionnaire A structured pain questionnaire was designed to asssess musculoskeletal pain during the previous three months. Musculoskeletal pain symptoms were classified according to body area (neck, upper limb, chest, lower limb, upper back, lower back, buttock) and frequency (pain seldom or never, once a month, once a week, more than once a week, almost daily). The 5-level frequency classification was adopted from the questionnaire used in the nationwide survey on health and health-related behaviors in schoolchildren by the WHO (King et al., 1996). Psychosomatic symptoms (headache, abdominal pain, depressive mood, day tiredness, difficulties in falling asleep, waking up during nights) were asked about with the same frequency categorization as for musculoskeletal pain. Each symptom was considered positive if it was present at least once a week. Disability due to pain was assessed by the following questions (1) do you have difficulties in falling asleep
552
A. El-Metwally et al. / Pain 110 (2004) 550–559
because of your pain or does your pain disturb your sleep (2) do you have difficulties while sitting during lessons (3) do you feel pain if you walk more than one kilometer (4) do you feel pain during physical exercise class (5) does your pain disturb your hobbies. Preadolescents were categorized into 3 groups according to their pain disability symptoms (no symptoms, 1 –2 symptoms and 3 –5 disability symptoms). Absence from school due to pain or aches was assessed to ascertain functional limitations. However, this question did not distinguish musculoskeletal pain from other types of pain (e.g. headache, abdominal pain). The amount of physical exercise was evaluated and children were categorized into 2 groups according to its frequency (less than 3 times a week and 3 or more times a week). The pain questionnaire included also questions about injury or trauma to the reported pain area, other diseases, and regular use of medications. During the design procedure, two versions of the questionnaire were tested in two different schools in Nastola, a neighboring community in Lahti, Finland and the test-retest reliability of the final version evaluated by repeating the survey at one-week interval (Mikkelsson et al., 1996). The test-retest reliability of the questionnaire in pain at least once a week was good (k 0.9). The concurrent validity of the pain questionnaire was examined by comparing it with interviews of 31 third- and 25 fifthgrade children. The observed agreement between pain questionnaire and interview technique was 86% (95% confidence interval [95% CI] 74 –94%) and k was 0.67 (Mikkelsson et al., 1997). 2.2.2. Hypermobility test Hypermobility test was conducted only at baseline, using Beighton’s method (Beighton et al , 1973). No stretching was allowed before the test. Intra and inter-observer reliability were measured earlier with a k coefficient of 0.75 and 0.78, respectively (Mikkelsson et al., 1996). Out of 403 children included in the analysis, 381 (94.5%) were examined for hypermobility. School absence on the examination date was the reason for non-participation. Beighton score of six was chosen as the cutoff point for hypermobility on the basis of the distribution of the results. 2.2.3. Shuttle run test As the hypermobility test, the shuttle run test (Leger and Lambert, 1982) was conducted only at baseline It measures maximal performance indoors and provides a valid and reliable index of cardio-respiratory endurance or maximal oxygen uptake (VO2max) (Committee of experts on sport research 1988). It was performed to all the study subjects except to children with acute musculoskeletal injury, acute respiratory infection, or other diseases inhibiting maximal physical strain. Children from schools that did not have room for the 20-m distance, needed to perform this test, were also excluded. Out of 403 children, who were included in the analysis, 368 (91.3%) underwent the shuttle run test.
2.3. Case definition of musculoskeletal pain Those who reported musculoskeletal pain in one or more area (s) at least once a week, and did not report injuring the pain site, fulfilled the case definition of musculoskeletal pain at baseline and constituted the study population for the follow-up study. The same case definition of musculoskeletal pain was used at the follow-up assessments. We have excluded children who reported injuring the pain area as, in this report, we are specifically investigating the outcome and contributing factors of recurrence of non-traumatic preadolescent musculoskeletal pain. It must be noted that we have only excluded children who reported directly injuring the musculoskeletal area with pain while overuse injuries were included in our case definition of musculoskeletal pain at both preadolescence and adolescence. 2.4. Follow-up The one-year follow-up was conducted in March 1996 with the same pain questionnaire as at baseline. At this follow-up, 515 (91.3%) of the children who had reported musculoskeletal pain at baseline filled in the pain questionnaire. At 4-year follow-up, 430 (76.2%) of the children in our original study population completed the pain questionnaire again. However, 27 children had incomplete data and were excluded. The final follow-up study population of 403 children with complete data was used in the analysis. These 403 children represented 71.5% of the original 564 subjects with musculoskeletal pain at baseline. Reasons for non-participation in the follow-ups included absence from school on the day of testing, changing schools, moving away and refusing to participate. Children who constituted our final study population were compared with those who were lost to follow-up or had incomplete data (Table 1). Both groups were similar with respect to almost all the baseline variables. The mean age (10.8 versus 10.9 years) and the mean VO2 max (34.4 versus 33.5 ml/kg/min) were comparable in the two groups. Furthermore, a multivariate logistic regression analysis was conducted to assess the impact of various baseline factors on the odds that a child would be lost to follow-up, and none of the baseline variables significantly influenced the rate at which children dropout. 2.5. Statistical methods The statistical analysis consisted of descriptive statistics with the most important values expressed with 95% CI. General linear models for the binomial family with log link function (risk ratio [RR]) were used to identify predictive factors for recurrence of musculoskeletal pain in adolescence. Musculoskeletal pain at 4-year follow-up was used as the dependent variable in all the models i.e. case was defined as recurrence of musculoskeletal pain at four years from baseline. The analysis included the following
A. El-Metwally et al. / Pain 110 (2004) 550–559 Table 1 Comparison of baseline data of included subjects and those not included in the analysis Variables
Included in the study
Lost at follow-up
Total, N Males, N (%)
403 194 (48.1)
161 79 (49.1)
Musculoskeletal pain, N (%) Neck pain Upper limb pain Chest pain Lower limb pain Upper back pain Lower back pain Buttock pain
187 (46.4) 64 (15.9) 68 (16.8) 219 (54.3) 70 (17.3) 47 (11.7) 25 (6.2)
73 43 24 85 37 23 13
Psychosomatic symptoms, N (%) Headache Abdominal pain Depressive mood
214 (53.0) 148 (36.5) 161 (40.0)
34 (56.5) 55 (34.2) 68 (42.2)
Difficulty in falling asleep Day tiredness Waking up during nights
147 (36.5) 228 (56.6) 110 (27.3)
72 (44.7) 98 (60.9) 53 (32.9)
Pain Disability symptoms, N (%) Pain disturbs sleeping Pain disturbs sitting during lessons Pain disturbs walking more than 1 km Pain disturbs physical exercise class Pain disturbs hobbies
119 (29.5) 62 (15.4) 129 (32.0) 105 (26.1) 90 (22.3)
51 26 47 44 52
Frequency of exercisea, N (%) Hypermobilityb, N (%) School absence due to pain, N (%)
202 (50.1) 30 (7.4) 101 (25.1)
78 (48.4) 7 (4.0) 43 (26.7)
(45.3) (27.0) (14.9) (52.8) (23.0) (14.3) (8.1)
(31.7) (16.1) (29.2) (27.3) (32.3)
553
3. Results At 1-year follow-up, 53.8% (95% CI 48.8 – 58.8) reported musculoskeletal pain at least once a week in at least one part of the body. At 4-year follow-up (at adolescence), 63.5% (95% CI 58.7– 68.1) had pain at least once a week in at least one musculoskeletal site. Musculoskeletal pain was reported by 74.7% (95% CI 68.5 – 80.0) of those who reported musculoskeletal pain at 1year follow-up and by 50.5% (95% CI 43.3 –57.6) of those who reported musculoskeletal pain at baseline only. Those who reported musculoskeletal pain at both baseline and 1year follow-up (persistent preadolescent musculoskeletal pain) had approximately three times (OR ¼ 2.9 [95% CI 1.9 – 4.4]) higher risk of musculoskeletal pain recurrence at adolescence compared to those who reported musculoskeletal pain at baseline only (Fig. 2). The most persistent/recurrent musculoskeletal pain site was the neck. Of 187 children reporting neck pain (as a single musculoskeletal pain site or combined with other musculoskeletal pain symptoms) at baseline, 90 (48.1% [95% CI 40.7 – 55.5]) had neck pain at 1-year follow-up and 98 (52.4% [95% CI 45.2– 59.5]) at 4- year follow-up (Fig. 3). At both 1-and 4-year follow-ups, children with regional musculoskeletal pain at baseline (with a single affected musculoskeletal site) had better prognosis than those with combined musculoskeletal pain (with several affected
a
Exercised more than 3 times a week. A child was categorized as hypermobile if his/her hypermobility score was six or higher. b
independent variables, age (below/above 11 years), psychosomatic symptoms (present/absent), disability index (0/1 – 2/3 –5), frequency of exercise (less/more than three times a week), Beighton score of hypermobility (below/above six), musculoskeletal pain location (regional/combined), school absence due to pain (never/once or more), and VO2 max (continuous). Regression models were first fitted with each of these variables separately, together with age, to estimate the age-adjusted predictive effect for each of them then a multivariate analysis was conducted using a backward stepwise method. All independent variables were initially included in the regression equation and elimination of variables, at each step, was based on the likelihood ratio test at 10% level of significance. Modification by sex was assessed by evaluating interactions between sex and other potential predictive factors. In all tests, a P value of less than 0.05 (two tailed) was considered statistically significant. All statistical analyses were performed using the SPSS (for Windows), version 10.0.
Fig. 2. Proportions of persistent/recurrent preadolescent musculoskeletal (MS) pain. 1-and 4-year follow-up assessments of schoolchildren in Lahti, Finland.
554
A. El-Metwally et al. / Pain 110 (2004) 550–559
Fig. 3. Persistence/recurrence proportions of preadolescent musculoskeletal (MS) pain symptoms at 1-year follow-up (1996) and 4-year follow-up (1999) of schoolchildren in Lahti, Finland.
musculoskeletal pain sites) (Fig. 4). Within those who had combined musculoskeletal pain at baseline and reported persistent/recurrent musculoskeletal pain, pain remained combined in the majority at both 1-year follow-up (66.7%) and 4-year follow-up (62.9%). On the other hand, within those who had regional musculoskeletal pain at baseline and reported persistent/recurrent musculoskeletal pain, pain was reported in several musculoskeletal sites by 35% of them at 1-year follow-up then by 51% of them at 4-year follow-up. Children who reported recurrent musculoskeletal pain at 4-year follow-up were older, had higher proportion of combined preadolescent musculoskeletal pain and hypermobility compared to children free of musculoskeletal pain
at follow-up. In addition, children with recurrent musculoskeletal pain had higher proportion of persistent preadolescent musculoskeletal pain and had more psychosomatic symptoms at baseline (Table 2). In the univariate analysis, children with combined musculoskeletal pain at baseline had a 24% higher ageadjusted risk of pain recurrence compared to those with regional musculoskeletal pain. Similarly, girls had a 24% higher risk of pain recurrence than boys and hypermobile children had a 29% higher recurrence risk compared to children with normal joint laxity. In addition to that, children aged 11– 13 years at baseline were more likely to develop musculoskeletal pain recurrence than those aged 9 – 10 and children with high subjective disability index were also at a higher risk. From the psychosomatic symptoms, headache, abdominal pain, and depressive feelings were predictive of musculoskeletal pain recurrence. In the multivariate analysis, only age (RR ¼ 1.28 [95% CI 1.10 – 1.49]), headache (RR ¼ 1.28 [95% CI 1.08– 1.51]), hypermobility (RR ¼ 1.35 [95% CI 1.08– 1.68]), and having combined musculoskeletal pain at baseline (RR ¼ 1.18 [95% CI 1.02 – 1.36]) remained statistically significant. Frequency of exercise, physical fitness and school absence due to pain did not predict the recurrence of musculoskeletal pain (Table 3). When analysed by sex, headache, abdominal pain and having combined musculoskeletal pain predicted pain recurrence among boys, and headache was the only statistically significant predictor in the multivariate analysis. The predictive factors in girls were age, abdominal pain, depressive feelings, waking up during nights and hypermobility. However, only age and depressive feelings and hypermobility were statistically significant in the backward step-wise regression. Statistically significant sex
Fig. 4. Proportions of persistent/recurrent musculoskeletal (MS) pain by number of involved sites at baseline (regional versus combined). 1-and 4-year followups of preadolescents in Lahti, Finland.
A. El-Metwally et al. / Pain 110 (2004) 550–559 Table 2 Baseline variables for adolescents with and without musculoskeletal pain at 4-year follow-up Variables
Musculoskeletal pain at adolescence
Age (mean)
10.9
Sex, N (%) Males Females
112 (43.8) 144 (56.2)
Musculoskeletal pain, N (%)a One area Two or more areas
No musculoskeletal pain at adolescence
Table 3 Relative risk (RR) and 95% confidence intervals (95% CI)) for recurrence of musculoskeletal pain from preadolescence till adolescence (403 children) Variables
Univariate analysis a RR (95% CI)
82 (55.8) 65 (44.2)
Sex Boys Girls
Reference 1.24 (1.07–1.44)
Age (years) 9– 10 11 –13
Reference 1.29 (1.11–1.51)
Reference 1.28 (1.10–1.49)
1.24 (1.06–1.43) 1.25 (1.09–1.43) 1.15 (1.00–1.32) 1.05 (0.90–1.21) 1.01 (0.87–1.17) 1.12 (0.96–1.30)
1.28 (1.08–1.51)
99 (67.3) 48 (32.7)
Pain at 1-year follow-up
162 (63.3)
55 (37.4)
Psychosomatic symptoms, N (%)a Headache Abdominal pain Depressive feelings Difficulty in falling asleep Day tiredness Waking up during nights
Psychosomatic symptoms Headache Abdominal pain Depressive feelings Difficulty falling asleep Day tiredness Waking up during nights
150 (59.1) 105 (41.8) 113 (44.7) 94 (37.3) 146 (57.0) 76 (30.0)
64 (44.1) 43 (29.5) 48 (33.1) 53 (36.1) 82 (56.2) 34 (23.4)
Disability index 0 1– 2 3– 5
Pain Disability symptoms, N (%) Pain disturbing sleeping Pain disturbing sitting Pain disturbing activity Pain disturbing walking Pain disturbing hobbies
Multivariate analysis RR (95% CI)
b
10.5
131 (51.2) 125 (48.9) b
555
c
Amount of exercise Less than 3 times a week More than 3 times a week
Reference 1.02 (0.86–1.21) 1.23 (1.02–1.49) Reference 1.00 (0.86–1.15)
(30.0) (17.2) (31.6) (25.8) (25.4)
42 (28.6) 18 (12.2) 48 (31.3) 39 (26.5) 25 (17.0)
Beighton score for hypermobility ,6 Reference 6– 9 1.23 (1.02–1.49)
Reference 1.35 (1.08–1.68)
Pain location at baselined Regional Combined
Reference 1.24 (1.08–1.42)
Reference 1.18 (1.02–1.36)
77 (30.0) 133 (52.0) 46 (18.0)
50 (34.0) 81 (31.3) 16 (10.9)
School absence due to pain Never absent Absence one day or more
Reference 1.08 (0.92–1.26)
VO2 max (per unit increase)
1.00 (0.98–1.01)
77 44 81 66 65
c
Disability index 0 1–2 3–5
Beighton score for hypermobility ,6 220 (90.5) 6–9 23 (9.5)
131 (94.9) 7 (5.1)
Frequency of exercise, N (%) Less than 3 times a week 127 (50.1) 3 or more times a week 123 (49.2)
74 (51.7) 69 (48.3)
School absence due to pain, N (%) VO2 max (mean)
67 (26.1)
34 (23.1)
34.6
34.1
a
Except for the variable “age”, all other figures were age adjusted. Using backward stepwise procedure. Only entered variables shown. c Number of pain disability symptoms. d Pain in one (regional) versus more than one (combined) musculoskeletal site at baseline. b
higher age-adjusted risk of recurrence within girls in the multivariate analysis. Out of 18 hypermobile girls who had preadolescent musculoskeletal pain at baseline, 17 reported musculoskeletal pain at 4-year follow-up.
a
Symptoms reported in a frequency of at least once a week. Those who had musculoskeletal pain at both baseline and 1-year follow-up (persistent preadolescent musculoskeletal pain). c Number of pain disability symptoms reported. b
interactions were found for age ðP , 0:001Þ; depressive feelings ðP ¼ 0:002Þ; waking up during nights ðP ¼ 0:049Þ and hypermobility ðP ¼ 0:037Þ: These factors predicted pain recurrence in girls but not in boys (Table 4). Hypermobility was associated with approximately a 30%
4. Discussion This population-based follow-up study of preadolescents with non-specific musculoskeletal pain shows that these symptoms are not as transient and self-limiting as previously assumed. Almost two thirds complained of musculoskeletal pain at 4-year follow-up. Children with regional musculoskeletal pain had better prognosis than
556
A. El-Metwally et al. / Pain 110 (2004) 550–559
Table 4 Relative risk (RR) and 95% confidence intervals (95% CI)) for recurrence of musculoskeletal pain from preadolescence till adolescence. Separate analysis for boys and girls Univariate analysis a RR (95% CI)
Multivariate analysis b RR (95% CI)
Boys
Girls
Boys
Age (years) 9– 10 11– 13
Reference 1.11 (0.87–1.42)
Reference 1.47 (1.21–1.78)
Psychosomatic symptoms Headache Abdominal pain Depressive feelings Difficulty falling asleep Day tiredness Waking up during nights
1.43 (1.12–1.81) 1.27 (1.00–1.61) 1.00 (0.78–1.29) 0.98 (0.75–1.26) 1.10 (0.86–1.41) 1.03 (0.79–1.35)
1.05 (0.89–1.24) 1.20 (1.03–1.40) 1.21 (1.03–1.42) 1.08 (0.92–1.27) 0.91 (0.78–1.07) 1.18 (1.01–1.37)
Disability indexc 0 1– 2 3– 5
Reference 1.25 (0.93–1.69) 1.37 (0.94–2.01)
Reference 0.91 (0.75–1.09) 1.13 (0.95–1.34)
More than 3 times a week
Reference 0.99 (0.78–1.27)
Reference 1.04 (0.89–1.26)
Beighton score for hypermobility ,6 6– 9
Reference 0.88 (0.50–1.56)
Reference 1.19 (1.00–1.41)
Pain location at baseline d Regional Combined
Reference 1.32 (1.04–1.66)
Reference 1.11 (0.95–1.31)
School absence due to pain Never absent Absence one day or more VO2 max (per unit increase)
Reference 1.05 (0.80–1.36) 1.01 (0.99–1.03)
Reference 1.13 (0.97–1.33) 0.99 (0.97–1.02)
Variables
Girls
Reference 1.40 (1.17–1.67) 1.43 (1.12–1.83) 1.24 (1.07 –1.43)
Amount of exercise Less than 3 times a week
a b c d
Reference 1.31 (1.18–1.46)
Except for the variable “age”, all other figures were age adjusted. Using backward stepwise procedure. Only entered variables shown. Number of pain disability symptoms. Pain in one (regional) versus more than one (combined) musculoskeletal site at baseline.
those with combined musculoskeletal pain. In addition to that, preadolescents who had regional musculoskeletal pain at baseline reported combined more than regional musculoskeletal pain at adolescence. This might indicate that musculoskeletal pain tends to be more generalized as the child grows up. It is difficult to compare these results with the descriptive results of the few previous follow-up studies. This is due to the differences in the children’s age at baseline, the length of follow-up and the case definitions of musculoskeletal pain. Moreover, the study subjects, in most of the previous prognostic studies, were drawn from hospitals or rheumatology clinics. Clinical study subjects may not represent the population at large due to referral biases, which is likely to overestimate the risk of recurrence, as documented in adult patients with persistent pain (Crook et al., 1989). Yet, our 4year recurrence proportion of musculoskeletal pain of roughly 65% is in agreement with the recurrence proportion
of 59% found in a clinic-based 9-year follow-up study (Flatø, et al., 1997). The high recurrence proportion in our study should be interpreted with caution due to the widecase definition and the short time recall of musculoskeletal pain. However, musculoskeletal pain is an intermittent phenomena and the pattern of recurrence rather than continuity may be a more realistic description of an individual’s experience of a chronic musculoskeletal problem (Deyo, 1993). Another possibility for our high recurrence figures would be that we evaluated musculoskeletal pain symptoms during winter when symptoms can be more frequent (Takala et al., 1992; Ehrmann Feldman et al., 2002). The 30% recurrence proportion of low-back pain in our study is somewhat lower than that of a follow-up study of 14-year-old children with low back pain (LBP) at baseline (Salminen et al., 1995). In their sample, 29% of the boys and 60% of the girls had recurrent LBP at 3-year follow-up, and 35% after 9 years from baseline (Salminen
A. El-Metwally et al. / Pain 110 (2004) 550–559
et al., 1999). In another population-based follow-up study, 27% of children with fibromyalgia at baseline still had similar symptoms at 30 months follow-up (Buskila et al., 1995). As mentioned above, the comparison of the studies is difficult. However, all these studies indicate that there are subpopulations among children with a tendency to develop chronic pain. Our univariate results show a high risk of musculoskeletal pain recurrence among girls, children who were older and children with high disability index at baseline. Furthermore, hypermobile children and children with psychosomatic symptoms and generalized musculoskeletal symptoms were at increased risk of recurrent pain. A strong association between psychosomatic symptoms and musculoskeletal pain in schoolchildren was also found in a recent cross-sectional study in a nationally representative sample of 11276 12 – 18-year-old Finnish adolescents (Vikat et al., 2000). In the multivariate analysis, age, headache, hypermobility and multiple painful musculoskeletal areas were the strongest predictive factors for pain recurrence. These results are in accordance with the previous cross-sectional studies, which have shown that prevalence of musculoskeletal pain in older children is higher than in younger (Salminen, 1984; Balague et al., 1988, 1994; Troussier et al., 1994; Burton et al., 1996) and that headache and musculoskeletal pain in children frequently co-exist (Sherry et al., 1991; Anttila et al., 2002). However, the crosssectional design strongly limits the conclusions of these earlier studies. The contributing factors of musculoskeletal pain recurrence are to a great extent similar to our previous results based on one-year follow-up. However, in this study we have found a statistically significant association between hypermobility at preadolescence and musculoskeletal pain recurrence at adolescence. This is in contrast to our previous findings, where hypermobility was not associated with preadolescent musculoskeletal at baseline nor was it predictive of musculoskeletal pain persistence at 1-year follow-up (Mikkelsson et al., 1996; Mikkelsson et al., 1997; Mikkelsson et al., 1998). An association between hypermobility and different musculoskeletal symptoms has been previously reported in a number of studies on schoolchildren (Beighton et al., 1973; Gedalia et al., 1991, 1993). However, we are not aware of any previous reports on the role of hypermobility in predicting recurrence of childhood musculoskeletal pain in adolescence. Our study shows differences in the factors predicting recurrent musculoskeletal pain between boys and girls. Age, depressive feelings, waking up during nights and hypermobility were predictive factors only in girls. Gender differences in the determinants of paediatric and adolescent chronic pain have been previously reported (Passchier and Orlebeke, 1985; Mirlijnet al., 2003). More psychosomatic symptoms predicted musculoskeletal pain recurrence in girls compared to boys. This is in line with earlier studies
557
where a relatively stronger association between psychosomatic-psychosocial factors and musculoskeletal pain in girls was found (Sherry et al., 1991; Merlijn et al., 2003). Joint hypermobility strongly predicted pain recurrence in girls but not in boys. Females show a greater joint mobility at all ages (Giannini and Brewer, 1982), and hypermobility syndrome is more prevalent among females (Beighton et al., 1973; Jessee el al., 1980). However, confirming the prognostic relevance of hypermobility in children needs a larger sample size of hypermobile children complaining of musculoskeletal pain at baseline in order to have enough statistical power for evaluating the difference between genders as regard to musculoskeletal pain recurrence. We are using the term persistence to describe those who reported musculoskeletal pain at 1-year follow-up and the term recurrence for those who reported musculoskeletal pain at 4-year follow-up. These might not be the correct terms for both. Occurrence of musculoskeletal pain in a frequency of at least once a week during the 3 months preceding the 1-year follow-up assessment does not necessarily indicate that the child had musculoskeletal pain with the same frequency during the whole preceding year, which might be assumed by using the term persistence. On the other hand, occurrence of musculoskeletal pain at 4-year follow-up might have been persistent musculoskeletal pain if the child had had pain during the whole 4-year follow-up period. Unfortunately, we did not have this information. Our questionnaire evaluated the occurrence of musculoskeletal pain only during the preceding 3 months. We have used this rather short time scale to avoid difficulties with recall of pain. Recall problems have been reported in several previous studies which have used a protracted recall time scale in collecting data about pediatric and adolescent pain (Goodman and McGrath, 1991; Savedra et al., 1988; Ehrmann Feldman et al., 2002). By using the term ‘psychosomatic symptoms’ we mean symptoms that are chiefly considered to be expressions of psychological stress and only rarely associated with any organic disease in schoolchildren. Depressive mood is certainly a psychological symptom. Childhood abdominal pain and headache are believed to be having a psychosomatic origin in the great majority of cases (Bury RG. 1987; Alfven G. 1986). However, it might be questionable whether difficulty in falling asleep, waking up during nights and day tiredness were manifestations of psychological stress or consequences of chronic pain. The rationale of excluding pain preceded by injury from our case definition of musculoskeletal pain is our belief that determinants and prognosis of musculoskeletal pain complaints related to trauma (falling down, sports injuries) might be different from non-traumatic pain complaints. Previous studies on young athletes have shown that most injuries in children’s sports are minor and self-limiting (Maffulli and Baxter-Jones, 1995; Maffulli and Bruns, 2000). Including children who
558
A. El-Metwally et al. / Pain 110 (2004) 550–559
reported pain symptoms precipitated by trauma at adolescence might have spuriously over-estimated our recurrence proportion at 4-year follow-up, as it is most probably un-related to the past history of non-traumatic musculoskeletal pain at preadolescence. To our knowledge, the present study is the first prospective follow-up study of preadolescents with musculoskeletal pain till adolescence. Strengths of the study are it’s population based design; fairly long prospective followup; children’s self-report of pain, as was previously recommended in pediatric pain research (Goodman and McGrath, 1991); and the use of a valid pain questionnaire which classifies musculoskeletal pain according to frequency and location aided by a diagram. In addition, pain interference with daily activities was assessed. However, pain intensity was not evaluated and we did not gather information about anatomic/structural factors (except for hypermobility) or socio-demographic factors. These factors might be related both to the onset and prognosis of musculoskeletal pain in children (Salminen et al., 1999; Aasland et al., 1997; Balague´, 1999). In conclusion, the current findings indicate that nonspecific musculoskeletal pain in preadolescents is not a selflimiting phenomenon. The persistence of musculoskeletal pain in preadolescence was strongly associated with pain recurrence in adolescence. As sex was a significant effect modifier for many baseline variables, future studies identifying determinants of musculoskeletal pain in children and adolescents should test for sex interactions. Although this study did not test for psychological variables (except for depressive feelings), the strong association between psychosomatic symptoms and musculoskeletal pain recurrence indicates that psychological issues should be evaluated with pediatric musculoskeletal pain to determine if psychological intervention should be incorporated.
Acknowledgements This study was funded by PATU Development Project and the Medical Research Fund of the Rheumatism Foundation Hospital. We thank Tuija Sulonen, RN, for assistance in collecting the data at follow-up. We also thank Riina Haataja, BA and Rahman Shiri, MD for commenting on the statistical analyses.
References Aasland A, Flatø B, Vandvik IH. Psychosocial factors in children with idiopathic musculoskeletal pain: a prospective, longitudinal study. Acta Paediatr 1997;86:740– 6. Alfven G., Psychological tension headaches among children_a survey. La¨kartidningen, 1986;45:3827–8 (in Swedish). Anttila P, Metsahonkala L, Aromaa M, Sourander A, Salminen J, Helenius H, Alanen P, Sillanpaa M. Determinants of tension-type headache in children. Cephalalgia 2002;22:401–8.
Balague´ F, Dutoit G, Waldburger M. Low back pain in school children. An epidemiological study. Scand J Rehabil Med 1988;20: 175 –9. Balague´ F, Nordin M, Skovron ML, Dutoit G, Yee A, Waldburger M. Nonspecific low-back pain among schoolchildren: a field survey with analysis of some associated factors. J Spinal Disord 1994;7:74– 9. Balague´ F, Troussier B, Salminen JJ. Non-specific low back pain in children and adolescents: risk factors. Review Eur Spine J 1999;8: 429 –38. Beighton P, Solomon L, Soskolne CL. Articular mobility in an African population. Ann Rheum Dis 1973;32:413 –8. Burton AK, Clarke RD, McClune TD, Tillotson KM. The natural history of low back pain in adolescents. Spine 1996;20:2323–8. Bury RG. A study of 111 children with recurrent abdominal pains. Aust Paediatr J 1987;23:117–9. Buskila D, Neumann L, Hershman E, Gedalia A, Press J, Sukenik S. Fibromyalgia syndrome in children-an outcome study. J Rheumatol 1995;22:525–8. Committee of experts on sports research. Eurofit. Handbook for the Eurofit tests of physical exercise. Rome: Council of Europe, Committee for the Development of Sport, 1988 p. 24. Crook J, Weir R, Tunks E. An epidemiological follow-up survey of persistent pain sufferers in a group family practice and specialty pain clinic. Pain 1989;36:49–61. Deyo RA. Practice variations, treatment fads, rising disability. Do we need a new clinical research paradigm? Spine 1993;18:2153 –62. Ehrmann Feldman D, Shrier I, Rossignol M, Abenhaim L. Risk factors for the development of neck and upper limb pain in adolescents. Spine 2002;27:523–8. Flatø B, Aasland A, Vandvik IH, Forre O. Outcome and predictive factors in children with chronic idiopathic musculoskeletal pain. Clin Exp Rheumatol 1997;15:569–77. Gedalia A, Press J:. articular symptoms in hypermobile schoolchildren; a prospective study. J Pediatr 1991;119:944 –6. Gedalia A, Press J, Klein M, Buskila D. Joint hypermobility and fibromyalgia in schoolchildren. Ann Rheum Dis 1993;52:494–6. Giannini EH, Brewer Jr EJ. Standard methodology for Segment I, II, and III Pediatric Rheumatology Collaborative Study Group studies. II. Analysis and presentation of data. J Rheumatol 1982;9: 114 –22. Goodman JE, McGrath PJ. The epidemiology of pain in children and adolescents: a review. Pain 1991;46:247 –64. Hakala P, Rimpela A, Salminen JJ, Virtanen SM, Rimpela M. Back, neck, and shoulder pain in Finnish adolescents: national cross sectional surveys. Br Med J 2002;325:743–6. Harreby M, Neergaard K, Hesselsoe G, Kjer J. Are radiologic changes in the thoracic and lumbar spine of adolescents risk factors for low back pain in adults? A 25-year prospective cohort study of 640 schoolchildren. Spine 1995;20:2298 –302. Hertzberg A. Prediction of cervical and low-back pain based on routine school health examinations. A nine- to twelve-year follow-up study. Scand J Prim Health Care 1985;3:247–53. Jessee EF, Owen Jr DS, Sagar KB. The benign hypermobile joint syndrome. Arthritis Rheum 1980;23:1053– 6. Kamper-Jorgensen L, Mollenvang Nielsen M., Born og unge. Kropsbevidsthet. Forebyggelsesinnsats. Rapport fra et seminar om forebyggelse af og muskel- og skeletsygdomme hos born og unge (in Danish). Dansk institutt for Klinisk Epidemiologi (DIKE). Kobenhavn: 1992. pp. 13 –17. King A, Wold B, Tudor-Smith C, Harel Y. The Health of Youth: A Cross-sectional Survey. Geneva: WHO; 1996. p. 68– 9. European Series No. 69. Laupacis A, Wells G, Richardson WS, Tugwell P. Users’ guides to the medical literature V. How to use an article about prognosis. J Am Med Assoc 1994;272:234–7. Leboeuf-Yde C, Kyvik KO. At what age does low back pain become a common problem? A study of 29,424 individuals aged 12-41 years. Spine 1998;23:228–34.
A. El-Metwally et al. / Pain 110 (2004) 550–559 Leger LA, Lambert J. A maximal multistage 20-m shuttle run test to predict VO2 max. Eur J Appl Physiol Occup Physiol 1982;49:1–12. Maffulli N, Baxter-Jones ADG. Vommon skeletal injuries in young athletes. Sports Med 1995;19:137 –49. Maffulli N, Bruns W. Injuries in young athletes. Eur J Pediatr 2000;159: 59–63. Merlijn VP, Hunfeld JA, van der Wouden JC, Hazebroek-Kampschreur AA, Koes BW, Passchier J. Psychosocial factors associated with chronic pain in adolescents. Pain 2003;101:33– 43. Mikkelsson M, Salminen JJ, Kautiainen H. Joint hypermobility is not a contributing factor to musculoskeletal pain in pre-adolescents. J Rheumatol 1996;23:1963 –7. Mikkelsson M, Salminen JJ, Kautiainen H. Non-specific musculoskeletal pain in preadolescents. Prevalence and 1-year persistence. Pain 1997; 73:29–35. Mikkelsson M, Salminen JJ, Sourander A, Kautiainen H. Contributing factors to the persistence of musculoskeletal pain in preadolescents: a prospective 1-year follow-up study. Pain 1998;77:67–72. Øster J. Recurrent abdominal pain, headache and limb pains in children and adolescents. Pediatrics 1972;50:429–36. Passchier J, Orlebeke JF. Headaches and stress in schoolchildren: an epidemiological study. Cephalalgia 1985;5:167–76.
559
Salminen JJ. The adolescent back. A field survey of 370 Finnish school children. Acta Paediatr Scand 1984;315:1–122. Salminen JJ, Erkintalo M, Laine M, Pentti J. Low back pain in the young. A prospective three-year follow-up study of subjects with and without low back pain. Spine 1995;20:2101– 8. Salminen JJ, Erkintalo MO, Pentti J, Oksanen A, Kormano MJ. Recurrent low back pain and early disc degeneration in the young. Spine 1999;13: 1316–21. Savedra MC, Tesler MD, Ward JD, Wegner C. How adolescents describe pain. J Adolesc Health Care 1988;9:315–20. Sherry DD, McGuire T, Mellins E, Salmonson K, Wallace CA, Nepom B. Psychosomatic musculoskeletal pain in childhood: clinical and psychological analyses of 100 children. Pediatrics 1991;88:1093 –9. Takala EP, Viikari-Juntura E, Moneta GB, Saarenmaa K, Kaivanto K. Seasonal variation in neck and shoulder symptoms. Scand J Work Environ Health 1992;18:257–61. Troussier B, Davoine P, de Gaudemaris R, Fauconnier J, Phelip X. Back pain in school children. A study among 1178 pupils. Scand J Rehabil Med 1994;26:143–6. Vikat A, Rimpela M, Salminen JJ, Rimpela A, Savolainen A, Virtanen SM. Neck or shoulder pain and low back pain in Finnish adolescents. Scand J Public Health 2000;28:164–73.