Whether an event x – for example the birth of a child – is causally related to a health problem y – such as later tooth loss – could hardly be investigated so far: For practical and ethical reasons, it is not possible to carry out clinical studies with random division into comparison groups, as is the case with many complex health research questions. To solve this problem to some extent, scientists from Heidelberg, Wuppertal, Munich, Worms and Nijmegen (Netherlands) have now transferred statistical methods from econometrics to medical research in a new approach. For example, they wanted to know what is behind the saying “Every child costs a mother a tooth”. After evaluating data from more than 34,000 people from the “Survey of Health, Ageing, and Retirement in Europe”, a data survey conducted in 14 European countries and Israel, they came to the conclusion that the birth of a child can actually lead to an above-average frequency of tooth loss. The results of the evaluation have now been published in the Journal of Epidemiology & Community Health.