Use of joinpoint regressions to evaluate changes over time in conscript height

Objectives Cohort variation in adult height expresses both the impact of socio‐economic change on human biology in a wide temporal perspective and social inequalities within populations. We aimed to test the use of joinpoint regressions to identify periods in which changes in height trends were statistically significant. Methods Data correspond to the height recorded in Madrid City (Spain) for 65 313 conscripts between 1936 and 1974 (cohorts from 1915 to 1953), a period of social and political turmoil. Secular trends in height were analyzed in eight districts with contrasting socio‐economic conditions, grouped in two categories, lower‐class and middle‐ and upper‐class. Trends in height were evaluated by quadratic regressions and by joinpoint regressions to identify the cut‐off years when trends changed significantly. Results Height increased in both socio‐economic categories of districts, more among conscripts from the lower‐class ones. However, results clearly show differences in trends according to district of residence. Whereas the increase in height in conscripts from the middle‐ and upper‐class districts was steady, it was slower in those from the lower classes, with declines in height during the Civil War and first years of the Franco dictatorship. Conclusions Joinpoint analysis reveals the association between urban living conditions and adult height, and that the disparities intensified during critical historical periods of Spain.