First of all, we would like to complement Siedler and his colleagues for their innovative case–control study in patients with clinically established tendon lesions of the m. supraspinatus in order to verify a pathological dose–response relation for work-related risk factors of shoulder complaints, taking into account personal and sport-related confounders (Seidler et al. 2011).

We were surprised to read that lifting and carrying was considered as a potential risk factor or confounder and not pushing and pulling. As far as we know, there is no epidemiological evidence for an association between lifting and carrying and shoulder symptoms. However, several studies reported significant associations between pushing and pulling and shoulder symptoms (Van der Beek et al. 1993; Hughes et al. 1997; Hoozemans et al. 2002a, b; Harkeness et al. 2003; Smedley et al. 2003). Not taking into account pushing and pulling as a potential risk factor or confounder may partly explain the observed differences in odds ratio’s (ORs) between exposure in terms of occupational groups (Table 2; Seidler et al. 2011) and in terms of strenuous activities (Table 3; Seidler et al. 2011). For instance, for construction workers, packers and physically exposed service workers (f.i. nurses and refuse collectors), Seidler et al. (2011) observed significant adjusted ORs of 2.5, 5.0 and 1.9, respectively. These ORs are somewhat higher than for lifting and carrying, which may be caused by the fact that jobs that consist of manual materials handling often include pushing and pulling besides lifting and carrying (Baril-Gingras and Lortie 1995). For comparable jobs, Hoozemans et al. (2002a) reported significant prevalence rate ratios (PRRs) for shoulder symptoms between 2.2 and 4.9 for self-reported and observed exposure to pushing and pulling in their 1-year prospective cohort study among 829 workers. In this study, the PRRs were adjusted for working above shoulder level and lifting and carrying. These findings are supported by biomechanical studies on contact forces at the glenohumeral joint in jobs like service workers in distribution (Hoozemans et al. 2004) and refuse collectors (Kuijer et al. 2003). Therefore, we strongly recommend taking into account pushing and pulling when evaluating manual materials handling, especially in relation to shoulder symptoms (Kuijer et al. 2007).