Health symptom questionnaire responses were collected from upper secondary and high school students (n ¼ 245) before comprehensive repairs of moisture damage in the school. The questionnaire study was repeated 1 year (n ¼ 227), 3 years (n ¼ 256), and 5 years (n ¼ 233) after the repairs. The data were analyzed both in cross-sectional design including all respondents, and longitudinally including paired observations of those individuals who had responded both before and after the repairs. In

more »

... ion, the effect of intervention on health symptoms was analyzed using generalized estimating equations (GEEs), taking into account within-subject correlation between repeated measurements. Compared to the situation before the repairs, the situation after the repairs was significantly improved in most of the 20 symptoms studied among the cross-sectional study populations. However, improvement was not so clear in the paired analysis and GEE analysis among the students who responded to three repeated questionnaires. The results indicate that the repairs succeeded in the sense that new cases of symptomatic students were no longer expected. However, the reversibility of symptoms among the group of exposed individuals may need to be considered separately.