Data sources

Information on height at school ages was obtained from the Copenhagen School Health Records Register (CSHRR) [6]. The CSHRR is a population-based cohort containing computerized information on 372,636 children who were born 1930–1989 and who attended a private or public school in the Copenhagen municipality. From 2 April 1968 onwards, a unique personal identification number was assigned to all Danish residents [7]. Children who attended school in 1968 or later had their identification number recorded on their health card, while it was retrieved for children who left school prior to this time.

Information on adult height was obtained by linkage of children in the CSHRR to the Danish Diet, Cancer and Health (DCH) cohort [8] via the identification number. Individuals included in the DCH cohort were enrolled from December 1993 to May 1997. Inclusion criteria were being 50–64 years of age, born in Denmark, residing in the Copenhagen or Aarhus areas, and no diagnosis of cancer registered in the Danish Cancer Registry prior to enrollment. Of 79,728 men who were invited to participate, 27,178 accepted the invitation, filled in a questionnaire, and were interviewed about their diet and lifestyle [8].

Via the identification number, information on prostate cancer status was obtained from the Danish Cancer Registry. It was established in 1942 [9], and the use of multiple notifications from different data sources secures a high degree of completeness. Further, through manual quality controls, validation of information in complex cases, and a high proportion of morphologically verified tumors, the validity of information in the register is enhanced [9]. Prostate cancer was defined by the International Classification of Diseases 10 code C61. Vital status was obtained by linkage to the national Vital Statistics Register [7].

Childhood height was measured by school doctors or nurses, and these health examinations were mandatory and performed annually through 1983 and thereafter only at school entrance and exit unless the child had specific health requirements [6]. To account for increases in height during childhood and across birth years, height measurements were transformed to z-scores based upon internal sex-, age-, and birth cohort-specific references (five-year intervals) using the Lamda, Median, Sigma (LMS) method [10].

Height at age 50–65 years (referred to as adult height) was measured by a trained laboratory technician [8]. To facilitate comparisons of the strengths of the associations across ages despite different standard deviations (SD) for heights, adult height was transformed to z-scores based upon internal birth cohort-specific references (1930–1934, 1935–1939, 1940–1944, 1945–1947). Adult height was as childhood height normally distributed, and the z-scores were calculated as standard z-score = [(observed value − population mean)/SD] using all information available among DCH males identified in the CSHRR.

The starting population for this study was men included in the CSHRR with an identification number, who were born 1930–1947, and who did not die, emigrate, or become lost to follow-up before 1 May 1993 or age 50 years (entry date and age criteria in the DCH cohort). In the eligible population of 6797 men included in both the CSHRR and the DCH, 62 were excluded due to a diagnosis of any cancer prior to enrollment in the DCH study, seven men with missing information on adult height, and 857 without information on height either at 7 or 13 years of age were excluded. Follow-up started on December 1993 through May 1997 at ages 50–65 years, upon enrollment into the DCH study, and ended at the date of diagnosis of prostate cancer, death, emigration, loss to follow-up, or 31 December 2012, whichever came first.

The project was approved by the Danish Data Protection Agency. According to Danish law, ethical approval is not required for purely register-based studies.