Musculoskeletal diseases--a continuing challenge for epidemiologic research.

In this paper some quality issues of epidemiologic studies on work-related musculoskeletal diseases are discussed. The advantages and disadvantages of different types of epidemiologic studies are described, among them the rarely applied case-crossover design. Problems in the ascertainment of disease, as well as the assessment of exposure to physical load, are also brought up. The importance of understanding the pathomechanisms of the diseases under study is stressed.