The responses to two independent questions indicated that adolescents experience a low level of musculoskeletal pain with little effect on everyday life. On the other hand, our findings indicate that computer use of 14 or more hours/week (corresponding ≥ 2 hours/day) is related to moderate/severe computer-associated pain at all measured anatomic sites. Moreover, computer use of 14 or more hours/week was related to moderate/severe pain in the lower back and head, which affected the everyday lives of adolescents. Girls reported greater pain intensity and more inconvenience to everyday life due to pain at all anatomic sites than boys, and the prevalence rates increased with age.
To the best of our knowledge, this is the first study to evaluate pain intensity and level of perceived negative impact on everyday life related to computer use in adolescents. Our results suggest that computer use of 14 or more hours/week (estimated ≥ 2 hours/day) is related to severe or moderate pain. Computer use exceeding 2 hours/day is suggested to be a threshold for neck-shoulder pain [5
] neck-shoulder pain is associated with 1 or more hours/day spent on the computer [11
In our study, the pain intensity was lower than that reported in some previous studies [18
], but approximately the same level as in the Finnish follow-up study by Saarni et al. with 88 participants [17
]. Our results confirm earlier findings, that girls report more severe pain than do boys [18
In the present study, the intensity of pain and inconvenience to everyday life were measured using the well-documented visual analogue scale, VAS. The VAS has been extensively studied and has shown good acceptability, responsivity, and validity for most children aged 8 years and older [25
]. Before analysis of the data, pain intensity categories were defined based on the findings of Collins et al. [26
]. Musculoskeletal symptoms in five anatomical locations were evaluated by two independent questions. These anatomical sites had been related to computer use in previous studies. The first question measured musculoskeletal symptoms in general with no reference to computer use. The second one measured musculoskeletal pain that children themselves attributed to their computer use. Using two different questions gives a more reliable picture and a wider perspective to the association between musculoskeletal symptoms and computer use. The questions and their VAS categories were stated in different sections of the questionnaire; computer-associated pain was stated in the ICT-section and musculoskeletal pain in the section concerning health status.
There are some limitations to the present study. The convenience sample of the present study does not necessarily represent the entire population of that age, although it is unlikely that this would influence the studied relationships. Due to the cross-sectional design of the study, we are restricted to investigating the contemporary relations in our data, and thus causal inferences cannot be made based on this study alone. As the questions used in the study measured the respondents' perception of the intensity of computer-associated pain, it was a subjective measure. Thus it was left to the respondents to report whether these symptoms were due specifically to computer use. Being a questionnaire survey, the rate of occurrence of computer -associated and musculoskeletal symptoms, the pain intensity and inconvenience to everyday life were based on self-reports, and therefore memory bias is possible and differences between individuals' interpretations cannot be ruled out [27
]. Respondents reported the weekly duration of computer use in response to an open question. There may be misreporting and overestimation of the duration of adolescents' computer use especially if there are different procedures to measure computer time [10
]. In adults, the use of self-reports can lead to the misclassification of computer exposure for more than 80% of respondents [28
]. In a previous study of adults self-reported duration of computer use but not recorded computer exposure, was positively associated with upper extremity symptoms [29
]. A study using a computer-based program showed that the use of a computer mouse predicted acute neck-shoulder pain, but not chronic neck-shoulder pain [30
]. Overall, based on an overview of systematic reviews, computer work has been associated with upper extremity musculoskeletal symptoms, but the causal relationship between computer work and symptoms shows a more mixed level of evidence [31
]. We also do not know either if the mechanisms by which computer use elicits pain differ between adults and adolescents. Future studies with longitudinal designs are needed to examine the persistence of musculoskeletal symptoms, with accurate measurements of computer use and to evaluate the changes between symptoms and computer use.