Pain. 2008 (Jul 15); 137 (2): 316–322 ~ FULL TEXT
Minna Stahl, Hannu Kautiainen, Ashraf El-Metwally, Arja Hakkinen,
Jari Ylinen, Jouko J. Salminen, Marja Mikkelsson
Department of Physical and Rehabilitation Medicine,
University Hospital of Turku,
P.O.Box 52, 20520
This study investigated the natural course of neck pain (NP) in 9-12-year-olds during a 4-year follow-up. Risk factors for the occurrence and persistence of weekly NP were explored separately for boys and girls. At baseline, 1756 schoolchildren completed a questionnaire eliciting musculoskeletal pain symptoms, other physical, and psychological symptoms and frequency of physical activity, and were tested for joint hypermobility. Symptoms during the preceding three months were asked using a five-level frequency classification. Re-evaluation was performed after one and four years using identical questionnaires. During follow-up, 24% reported none, 71% fluctuating, and 5% persistent weekly NP. The frequency of NP at baseline was linearly related to weekly NP during follow-up in both genders (P<0.001). Furthermore, a significant increasing linear trend towards a more persistent course of NP was seen in children with weekly other musculoskeletal and/or other physical and psychological symptoms at baseline. Among originally neck pain-free pre-/early adolescents, weekly other musculoskeletal pain symptoms (only in girls) and other physical and psychological symptoms (in both genders) predicted the occurrence of weekly NP during follow-up. In conclusion, neck pain in schoolchildren tends to fluctuate, but there also seems to exist a subgroup (5%) with persistent NP already in pre-/early adolescents, or even earlier. Co-occurrence of frequent other musculoskeletal symptoms and/or markers of psychological stress with frequent NP are risk indicators for a more persistent course, at least within next few years. Since adult chronic NP problems might originate in childhood, further studies are needed, including preventive interventions.
KEYWORDS: Neck pain; Children; Adolescents; Risk factors; Prognosis; Epidemiology
From the Full-Text Article:
Chronic neck pain (NP) is a common health problem
among adults. [3, 7, 29, 39] According to a Finnish study,
5% of men and 7% of women suffer from chronic NP. 
From the beginning of the 1990s, the prevalence of NP
has steadily increased among Finnish adolescents ,
and similar trends have also been reported in some other
Western countries.  This has raised speculations that
NP might become a significant health problem in the
near future. According to Brattberg , chronic musculoskeletal
pain problems may appear already at the age
of 8–14 years, perhaps even earlier. Although there is
some evidence that adolescents with neck symptoms
are at a higher risk of having NP in adulthood [19, 31],
similar prognostic studies in younger children are lacking.
Thus, there are still gaps in the literature regarding
factors that can predict the development and persistence
of neck symptoms in school-aged children. Such information
would be highly important for creating scientifically-
based primary and secondary prevention
programs for youth, which might have implications for
adults as well.
This study is a continuation of the previous, prospective,
population-based follow-up surveys, which showed
that NP was the most persistent and recurrent musculoskeletal
pain in school-aged children. [10, 24] In addition,
we have reported that newly-developed neck symptoms
in originally musculoskeletal pain-free children were
mainly self-limiting and that NP became more common
among girls than boys in adolescence.  The aim of
this study was to conduct an in-depth investigation
about the course of NP in different frequency groups
from pre-/early adolescence stage to mid-adolescence
and to explore possible risk factors for its occurrence
and persistence separately for boys and girls.
This study took place in Lahti, a city in Southern Finland
with a population of 94,827 (1995). Nineteen of the 21 primary
schools in the town accepted the invitation to participate. Due
to unsuitability of some of the study methods, the city’s special
schools were excluded: the Steiner school, the hospital school,
and the schools for hearing impaired, physically disabled, and
mentally handicapped children. The study was conducted in
March 1995, and all the pupils from the third and fifth grades
present on the survey day participated. The initial sample consisted
of 1,756 children, 867 from the third (mean age 9.8
[SD = 0.4]) and 889 from the fifth grade (mean age 11.8
[SD = 0.4] years), representing 83% of all schoolchildren in
these grades in the city of Lahti.
The pupils were re-evaluated after one and four years, in
March 1996 and 1999. There were 1268 (72%) schoolchildren
(597 boys and 671 girls) who completed a study questionnaire
on all three occasions, at baseline and at the two follow-ups.
These schoolchildren constituted our study population. The
number of dropouts during the follow-up period was 488
(264 boys and 224 girls) owing to absence from school on
the survey day, changing schools, moving away, and refusal
to participate. The study population and those who were lost
to follow-up were similar in age and NP frequency
On all three study occasions, the subjects completed identical
questionnaires including pain questionnaire designed to assess
musculoskeletal pain symptoms in schoolchildren and additional
questions about other symptoms and frequency of physical
activity. Joint hypermobility was also tested at baseline.
The pain questionnaire included site-specific questions of
musculoskeletal pain in seven different areas of the body (A:
neck, B: upper limb, C: chest, D: lower limb, E: upper back,
F: lower back, and G: buttock). Beside the pain questions,
there were body maps with front and back views, divided into
seven different body parts to indicate the anatomic areas A–G
of the question. For analysis, the body parts E, F, and G were
combined to represent “back pain”. Musculoskeletal pain
symptoms were asked by using a five-level frequency classification
(pain seldom or never, once a month, once a week, more
than once a week, almost daily) during the preceding three
months (i.e. from Christmas until the day of the study). For
analysis, pain frequencies once a week, more than once a week
and almost daily were combined and renamed as “at least once
a week/weekly” class. Based on the frequency of NP reported
during the follow-up, three different types of courses of NP
were identified: pain-free (no NP at any evaluation point), fluctuating
(NP frequency varied from no pain to once a month to
at least once a week), and persistent (NP at least once a week
at all three evaluation points).
The development of the pain questionnaire, its reliability and
concurrent validity had been measured earlier in a sub-sample of
the population. [23, 24] The main outcome variable of this study,
neck pain, was considered in three different frequency categories
(none/once a month/ at least once a week) and the test–retest
reliability in stability for this threefold neck pain classification
over a week time-interval was good (j = 0.87).
Additional questions evaluated six physical and psychological
symptoms: headache, abdominal pain, depressive mood,
daytime tiredness, difficulty falling asleep, and waking up during
the night. These symptoms were asked using the same frequency
categorization as for musculoskeletal pain symptoms.
For analysis, all symptoms appearing at least once a week were
The frequency of physical activity was elicited with the
question:”How often do you exercise vigorously for at least
half an hour?” The response alternatives were “not at all”,
“1–2 times”, “3–4 times”, and “5–7 times per week”. Based
on the frequency of the reported physical activity, children
were categorized into 2 groups (less than 3 times a week, and
3 or more times a week).
Joint hypermobility was assessed at baseline using the Beighton’s method
(score 0 to 9, one point for each side of the body for tests A–D and one point for test E):
(A) passive dorsiflexion of the little fingers beyond 90°,
(B) passive apposition of the thumbs to the flexor aspect of the forearm,
(C) hyperextension of the elbows beyond 10°,
(D) hyperextension of the knees beyond 10°, and
(E) forward flexion of the trunk, with knees straight, so that the palms of the hands rest easily on the floor. 
The intra-and inter-observer reliability of this
method had also been measured earlier in a sub-sample of the
population with j coefficients of 0.75 and 0.78, respectively.  Beighton score of 6 was chosen as the cut-off point for
hypermobility on the basis of the distribution of the results.
Variables with normal (Gaussian) distribution of the
descriptive values were expressed by mean and standard deviations
(SD); statistical comparisons between the groups were
made using the t-test or analysis of variance (ANOVA). Measures
with discrete distributions were expressed as counts (%)
and analyzed by v2, Fisher’s exact test or Cochran–Armitage–
Trend test with Monte Carlo P-values. Risk factors for
the occurrence of NP during the follow-up were analyzed using
generalizing estimating equations (GEE) models with an
exchangeable correlation structure. Hommel’s method was
applied to adjust levels of significance for multiple testing if
Courses of NP from pre-/early adolescence to midadolescence
The population was divided into three subgroups
depending on the self-reported NP frequency at baseline:
no NP (61%), NP once a month (24%), and NP
at least once a week (15%). Figure 1 shows the natural
courses of NP during the 4-year follow-up within these
three groups, and in the whole cohort: 304 [24% (95%
CI 22 to 27)] stayed pain-free, 896 [71% (68 to 73)]
had a fluctuating course of NP, and 64 [5% (4 to 6)] persistent
NP at least once a week.
Relationship between the frequency of NP at baseline and during the follow-up
Of those schoolchildren reporting no NP at baseline,
219 [28% (95% CI 25 to 32)] reported weekly NP during
the follow-up (at either one or both re-evaluation
points). Of those reporting NP once a month at baseline,
152 [49% (95% CI 44 to 55)] had weekly NP during the
follow-up. Of those reporting weekly NP at baseline,
128 [67% (95% CI 60 to 72)] also had weekly NP during
the follow-up. There was a significant linearity
(P < 0.001) between the different baseline NP frequency
groups and weekly NP during the follow-up in both genders
(Figure 2). Girls who reported weekly NP at baseline
were 5.9 (95% CI 4.0 to 8.7) times more likely to also
have weekly NP during the follow-up compared with
girls without NP at baseline. The corresponding odds
ratio for boys was 4.6 (95% CI 3.2 to 7.4).
Baseline factors related to the persistent course type
Comparisons were made between the twelve different
baseline variables and the three different NP course
groups separately for boys and girls (Table 1). In both
sexes, there was a significant increasing linear trend
towards a more persistent course of NP when having
weekly other musculoskeletal pain and/or other selfreported
physical and psychological symptoms at baseline.
Joint hypermobility and frequency of physical
activity in pre-/early adolescence were not related to
the later course of NP.
Factors predicting occurrence of weekly NP
Schoolchildren reporting no neck symptoms at baseline
(N = 769) were chosen for the GEE models to determine
factors predicting occurrence of weekly NP during
the follow-up. Analyses were performed separately for
boys and girls (Table 2). The occurrence of weekly NP
increased during the follow-up time in both genders.
Having other physical and psychological symptoms with
a frequency of at least once a week at baseline predicted
the occurrence of weekly NP in both genders. Furthermore,
having weekly other musculoskeletal pain symptoms
predicted the occurrence of weekly NP among
girls. The number of reported other musculoskeletal
pain (1 to 4) or physical and psychological symptoms
(1 to 6) increased the odds ratio, respectively.
In this study, for the first time ever, the long-term (4
years) prognosis and risk factors of NP have been evaluated
in a pre-/early adolescent to mid-adolescent population.
As has been noted before, NP, already quite
prevalent in preadolescents, becomes more prevalent in
adolescence with female over-representation. [9, 13, 16, 18, 22, 27, 32, 33, 37, 38]. This study confirms our earlier
findings (in an originally musculoskeletal pain-free
cohort) that NP is mostly a fluctuating phenomenon
among school-aged children.  However, there is a
subpopulation (5%) showing a persistent course of neck
symptoms beginning already at ages 9 to 12, perhaps
even earlier. Siivola et al.  reported that neck symptoms
in adolescence predicted NP in early adulthood.
Interestingly, according to Cote et al. , about 5% of
adults suffer from highly disabling chronic NP. These
findings raise a suspicion that life-long chronic non-specific NP problems have their origins in childhood. Thus,
further longitudinal studies extending from childhood to
adulthood are needed.
We have previously reported that the frequency of
NP fairly well reflects the intensity of pain.  Moreover,
the findings of the current study indicated that
the frequency of NP is an important predictive factor
for the persistence of the pain.
Since a gender difference becomes evident in the prevalence
rates of NP in adolescence, we were interested to
explore the role of possible earlier risk factors separately
among boys and girls. We found that self-reported frequent
other musculoskeletal pain and/or other physical
and psychological symptoms such as headache, abdominal
pain, day tiredness, depressive mood and sleep difficulties
were risk factors for both the occurrence and
persistence of weekly NP during the next four years.
The physical and psychological symptoms asked in our
study are known only rarely associated with any organic
disease in schoolchildren and chiefly considered expressions
of psychological stress and often called psychosomatic
symptoms. [1, 2, 6, 11, 35] Our risk factor findings
are in accordance with previous reports: co-occurrence
of other musculoskeletal pains with NP has been recognized
in adolescent populations. [16, 30, 33, 37] Grimmer
et al.  reported that any bodily pain was experienced
by more than 50% of 13–17-year-old adolescents during
the previous week and, further, that the prevalence of
headache was decreasing with a concomitant increase
in neck and upper back pain over the 5-year study period.
Psychosomatic symptoms have displayed the strongest
association with NP compared to all other studied
potential risk factors in earlier cross-sectional studies. [36, 37]
In the study by Diepenmaat et al. , an association
existed between NP, stress and depressive symptoms
among 12–16-year-old adolescents. Feldman
et al.  demonstrated that a lower mental health score
was a risk factor for the development of weekly NP in a
one-year follow-up among adolescents. Parallel findings
have been reported among school-aged children with
low back pain.  Mikkelsson et al. , for one, suggested
in their 1-year follow-up of pre-/early adolescents
that depressive symptoms and sleep problems may contribute
to regional NP becoming widespread. In the
study by Siivola et al.  psychosomatic stress symptoms
in adolescence predicted neck and shoulder pain
in early adulthood.
In our study, neither general joint hypermobility nor
the level of physical activity affected the course of NP in
schoolchildren. We must note that in using the Beighton’s
method, as we did to detect joint hypermobility,
the specific mobility of the neck cannot be determined.
Several studies have been conducted to investigate the
relationship between hypermobility and musculoskeletal
symptoms in schoolchildren, but none of them has concentrated
specifically on NP and the results have been
contradictory. [4, 10, 14, 15, 23–25, 28] Similarly, inconsistent
findings have been published regarding the association
between physical activity and NP. This might be
attributed to the large variation of methods used to measure
physical activity, which prevents proper comparisons. [13, 22, 27, 37]
The strength of this study is its prospective, population
based setting with a response rate of 72% during
follow-up.  Furthermore, the outcome measure
(NP) was based on self-reports, precisely defined, and
the frequency was evaluated. [12, 17] The fluctuation of
NP due to seasonal variation was precluded, since the
baseline study and both follow-ups were conducted in
wintertime.  In these data, girls were over-represented
among the participants included in the analysis
vs. those lost to follow-up. However, it is likely that this
had no impact on the results, since the analyses were
performed separately for both genders. Another weakness
is that the follow-up did not cover the entire follow-
up time, but only the period of three months at
one and four years. Yet, we considered it appropriate
to call NP persistent if its reported frequency was at
least once a week at all three evaluation points.
By solely relying on the self-report of symptoms from
our study population, we have most likely detected also
neck symptoms that are milder than those seen in clinical
settings as only small proportion of the children seek
help for their neck pain. This might have led to underestimation
of the strength of the risk association – i.e. the
true measurements of risk association, for more severe
non-specific neck pain, would be greater than what we
report in this paper. It always must be remembered that
children and adolescents with frequent neck pain seeking
medical help must be carefully examined and followed,
because there might also be few cases with
underlying severe disease. Once severe underlying causes
for neck pain have been excluded, it might be helpful to
ask about the risk factors found in this study in order to
distinguish children most likely to have self-limiting
neck symptoms from those who are at a higher risk of
having more persistent NP, and who might hence benefit
from early secondary preventive interventions.
In conclusion, our results suggest that neck symptoms
in school-aged children tend to fluctuate in the
majority of cases, yet there seems to be a subgroup
(5%) of children with persistent NP already at ages 9
to 12, perhaps even earlier. Co-occurrence of frequent
other musculoskeletal symptoms and/or markers of
psychological stress with weekly NP are risk indicators
for a more persistent course of NP from pre-/early
adolescence to mid-adolescence. Our findings raise a
suspicion that adult chronic NP problems might have
their origins in childhood and therefore further studies
in young age groups, including preventive interventions,
The Medical Research Fund of The Rheumatic
Foundation Hospital, the University Hospital of Turku
and the Central Hospital of Jyva¨skyla¨ supported this
study. We thank Mrs. Tuija Sulonen, RN, for assistance
in collecting the data at follow-up, Mr. Olli Heinonen,
Paavo Nurmi Keskus, for offering a desk and a computer
for the writing process of this article and Mrs.
Marja Vajaranta for revising the language.
Psychological tension headaches among children. A survey.
Lakartidningen 1986;45:3827–8. in Swedish.
Aro H, Paronen O, Aro S.
Psychosomatic symptoms among 14–16 year old Finnish adolescents.
Soc Psychiatry 1987;22:171–6.
Aromaa A, Koskinen S.
Health and functional capacity in finland: baseline results of the Health 2000 Health Examination Survey.
Helsinki, Finland: National Public Health Institute; 2002.
Beighton P, Solomon L, Soskolne CL.
Articular mobility in an African population.
Ann Rheum Dis 1973;32:413–8.
Do pain problems in young school children persist into early adulthood? A 13-year follow-up.
Eur J Pain 2004;8:187–99.
A study of 111 children with recurrent abdominal pains.
Aust Paediatr J 1987;23:117–9.
Cassou B, Derriennic F, Monfort C, Norton J, Touranchet A.
Chronic neck and shoulder pain, age, and working conditions: longitudinal results from a large random sample in France.
Occup Environ Med 2002;59:537–44.
Cote P, Cassidy JD, Carroll L.
The Saskatchewan Health and Back Pain Survey.
The Prevalence of Neck Pain and Related Disability in Saskatchewan Adults
Spine (Phila Pa 1976). 1998 (Aug 1); 23 (15): 1689–1698
Diepenmaat ACM, van der Wal MF, de Vet HCW, Hirasing RA.
Neck/shoulder, low back, and arm pain in relation to computer use, physical activity, stress, and depression among dutch adolescents.
El-Metwally A, Salminen JJ, Auvinen A, Kautiainen H, Mikkelsson M.
Prognosis of non-specific musculoskeletal pain in preadolescents: a prospective 4-year follow-up study till adolescence.
El-Sheikh M, Buckhalt JA, Mark Cummings E, Keller P.
Sleep disruptions and emotional insecurity are pathways of risk for children.
J Child Psychol Psychiatry 2007;48:88–96.
Fejer R, Kyvik KO, Hartvigsen J.
The prevalence of neck pain in the world population: a systemic critical review of the literature.
Eur Spine J 2006;15:834–48.
Feldman D, Shrier I, Rossignol M, Abenhaim L.
Risk factors for the development of neck and upper limb pain in adolescents.
Gedalia A, Press J.
Articular symptoms in hypermobile schoolchildren: a prospective study.
J Paediatr 1991;119:944–6.
Gedalia A, Press J, Buskila D.
Joint hypermobility and fibromyalgia in schoolchildren.
Ann Rheum Dis 1993;52:494–6.
Grimmer K, Nyland L, Milanese S.
Repeated measures of recent headache, neck and upper back pain in Australian adolescents.
Goodman J, McGrath P.
The epidemiology of pain in children and adolescents: a review.
Hakala P, Rimpela¨ A, Salminen JJ, Virtanen SM, Rimpela¨ M.
Back, neck and shoulder pain in Finnish adolescents: national cross-sectional surveys.
Prediction of cervical and low back pain based on routine school health examinations.
Scan J Prim Health Care
Jones G, Watson K, Silman A, Symmons D, Macfarlane G.
Predictors of low back pain in British school children: a population-based prospective Cohort study.
King A, Wold B, Tudor-Smith C, Harel Y,
The health of youth. A Cross-National Survey,
WHO Regional Publications 1996, Eur Ser 1996;69:68–9.
Kujala U, Taimela S, Viljanen T.
Leisure physical activity and various pain symptoms among adolescents.
Br J Sports Med 1999;33:325–8.
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.
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.
Mikkelsson M, Sourander A, Salminen JJ, Kautiainen H, Piha J.
Widespread pain and neck pain in schoolchildren. A prospective one-year follow-up study.
Acta Paediatr 1999;88: 1119–24.
Niemi S, Levoska S, Kemila¨ J, Rekola K,
Keinanen-Kiukaanniemi S. Neck and shoulder symptoms and leisure time activities in high school students.
Qvindesland A, Jonsson H.
Articular hypermobility in Icelandic 12-year-olds.
Picavet HSJ, Schouten JSAG.
Musculoskeletal pain in the Netherlands: prevalences, consequences and risk groups, the DMC3-study.
The adolescent back. A field survey of 370 Finnish school children.
Acta Paediatr Scand 1984;73:1–122.
Siivola S, Levoska S, Latvala K, Hoskio E, Vanharanta H, Keinanen-Kiukaanniemi S.
Predictive factors for neck and shoulder pain: a longitudinal study in young adults.
Smedbraten B, Natvig B, Rutle O, Bruusgaard D.
Self-reported bodily pain in school children.
Scand J Rheumatol 1998;27:273–6.
Stahl M, Mikkelsson M, Kautiainen H, Hakkinen A, Ylinen J, Salminen JJ.
Neck pain in adolescence. A 4-year follow-up of pain-free preadolescents.
Takala E-P, Viikari-Juntura E, Moneta G, Saarenmaa K, Kaivanto K.
Seasonal variation in neck and shoulder symptoms.
Scand J Work Environ Health 1992;18:257–61.
Tamminen TM, Bredenberg P, Escartin T, Kaukonen P, Puura K, Rutanen M, et al.
Psychosomatic symptoms in preadolescent children.
Psychother Psychosom 1991;56:70–7.
van Gent C, Dols J, de Rover C, Hira Sing R, de Vet C.
The weight of schoolbags and the occurrence of neck, shoulder and back pain in young adolescents.
Vikat A, Rimpelä M, Salminen JJ, et al.
Neck or Shoulder Pain and Low Back Pain in Finnish Adolescents
Scand J Public Health. 2000 (Sep); 28 (3): 164–173
Wedderkopp N, Leboeuf-Yde C, Andersen LB, Froberg K, Hansen HS.
Back Pain Reporting Pattern in a Danish Population-based Sample of Children and Adolescents
Spine (Phila Pa 1976). 2001 (Sep 1); 26 (17): 1879–1883
Webb R, Brammah T, Lunt M, Urwin M, Allison T, Symmons D.
Prevalence and predictors of intense, chronic, and disabling neck and back pain in the UK general population.
Return to PEDIATRICS
Return to NECK AND BACK PAIN