Table 2.

Quality assessment of four articles reporting on five studies on puberty and back pain

		Jansens[33](Netherlands)	Jansens[33](USA)	Wedderkopp[35]	LeResche[34]	Hulsegge[32]
1. Study sample	- Was the sampling likely to be suitable to obtain a group representative of the corresponding general population?	Yes (Probably)	Yes (Probably)	Yes (Probably)	Yes (Probably)	Yes (Probably)
	- If participation at BL or at FU < 80%, was response bias investigated?	Yes	Yes	Yes	Yes	No
	- Inclusion criteria, were age groups appropriate to study the whole period of puberty?	Yes	Yes	Yes	Yes	Not really
2. Data collec-tion	- Data for BP and puberty were collected independently by 2 different persons or by at least 1 questionnaire?	Yes	No	Yes	No	Yes
3. Puberty	- Clearly defined?	Yes	Yes	Yes	Yes	Yes
	- Description of how puberty stage was determined?	Yes	Yes	Yes	Yes	Yes
	-Reference provided for validity of test?	No	No	No	Yes	Yes
	-Reference provided for reliability of test?	Yes	Yes	No	No	No
4. Back pain	- Clearly defined?	Yes	Yes	Yes	Yes	Yes
	- Description of how BP was assessed?	Yes	Yes	Yes	Yes	Yes
	- Recall period (<= 1-month)?	No	No	Yes	No	No
	- Was BP at BL taken into account?	Yes	Yes	NA	NA	NA
5. Control for age and sex	- Was the association puberty- BP controlled for age?	Yes	Yes	Stratified analysis	Yes	Yes
	- for sex?	Yes	Yes	NA	Stratified analysis	Yes
6. Stat. analysis	- Was dose-response prevalence estimated in relation to puberty stages investigated?	Yes	Yes	Yes	Yes	No
	- Was dose-response of frequency or severity in relation to puberty stages investigated?	No	No	NA	NA	NA
	- Was dose-response tested for trend?	Yes	Yes	Yes	Yes	Yes
Score (0-17)%		14/17	13/17	12/14	12/15	10/15
		82%	76%	86%	80%	67%

(listed by the quality score).

BL = baseline, FU = follow up, BP = back pain, NA = not applicable.