Manuela Cattelan

Background: Many nonconclusive studies have been conducted on low back pain (LBP) in adolescents and associated factors. Objective: The aim was to assess the lifetime prevalence and associated factors of LBP in adolescents. Materials and Methods: A questionnaire was administered in high school students (14–19-yr-old participants) in Veneto region (Italy). The self-administered, structured questionnaire included anthropometric data; psychologic factors and lifestyle; presence, intensity, and family history of LBP; referral to professional health care for LBP; and a short version of the International Physical Activity Questionnaire. Results: A total of 6281 adolescents were recruited; 5204 questionnaires were included in the final analysis. A total of 2549 (48.98%) students reported one or more LBP episodes and 723 (13.89%) reported nonspecific disabling lumbar pain (i.e., no underlying pathology); 1040 (41.11%) subjects with LBP consulted a healthcare professional. A significant association emerged for LBP with sex (female), positive family history, time spent sitting or using electronic devices, sleep deprivation (<5 hrs/night), and low level of physical activity. Conclusion: In a large sample of adolescents, LBP lifetime prevalence is high and often associated with disabling pain and sedentary lifestyle, requiring professional care. These findings may support the development of prevention and treatment strategies of LBP in adolescents, reducing the risk of developing chronic pain.

Composite likelihood inference has gained much popularity thanks to its computational manageability and its theoretical properties. Unfortunately, performing composite likelihood ratio tests is inconvenient because of their awkward asymptotic distribution. There are many proposals for adjusting composite likelihood ratio tests in order to recover an asymptotic chi-square distribution, but they all depend on the sensitivity and variability matrices. The same is true for Wald-type and score-type counterparts. In realistic applications, sensitivity and variability matrices usually need to be estimated, but there are no comparisons of the performance of composite likelihood-based statistics in such an instance. A comparison of the accuracy of inference based on the statistics considering two methods typically employed for estimation of sensitivity and variability matrices, namely an empirical method that exploits independent observations, and Monte Carlo simulation, is performed. The results in two examples involving the pairwise likelihood show that a very large number of independent observations should be available in order to obtain accurate coverages using empirical estimation, while limited simulation from the full model provides accurate results regardless of the availability of independent observations. This suggests the latter as a default choice, whenever simulation from the model is possible.

Composite likelihood inference has gained much popularity thanks to its computational manageability and its theoretical properties. Unfortunately, performing composite likelihood ratio tests is inconvenient because of their awkward asymptotic distribution. There are many proposals for adjusting composite likelihood ratio tests in order to recover an asymptotic chi square distribution, but they all depend on the sensitivity and variability matrices. The same is true for Wald-type and score-type counterparts. In realistic applications sensitivity and variability matrices usually need to be estimated, but there are no comparisons of the performance of composite likelihood based statistics in such an instance. A comparison of the accuracy of inference based on the statistics considering two methods typically employed for estimation of sensitivity and variability matrices, namely an empirical method that exploits independent observations, and Monte Carlo simulation, is performed. The results in two examples involving the pairwise likelihood show that a very large number of independent observations should be available in order to obtain accurate coverages using empirical estimation, while limited simulation from the full model provides accurate results regardless of the availability of independent observations.

Summary: The study of the determinants of fights between animals is an important issue in understanding animal behavior. For this purpose, tournament experiments among a set of animals are often used by zoologists. The results of these tournament experiments are naturally analyzed by paired comparison models. Proper statistical analysis of these models is complicated by the presence of dependence between the outcomes of fights because the same animal is involved in different contests. This paper discusses two different model specifications to account for between-fights dependence. Models are fitted through the hybrid pairwise likelihood method that iterates between optimal estimating equations for the regression parameters and pairwise likelihood inference for the association parameters. This approach requires the specification of means and covariances only. For this reason, the method can be applied also when the computation of the joint distribution is difficult or inconvenient. The proposed methodology is investigated by simulation studies and applied to real data about adult male Cape Dwarf Chameleons.