br We observed an inverse
We observed an inverse association between height and HNC risk (2% lower in risk per 5 cm height increase), which is consistent with previous investigations on adult height and HNC risk within the INHANCE consortium that reported a 9% and 14% 21658-70-8 in the risk of HNC with every 10 cm increase in height among men and women . In the INHANCE study, every 10 cm increase in height was also associated with a lower HNC risk in Japan, but this association was only observed in women . One possible explanation for our results is the influence of childhood or adolescent nutrition status that might have influenced both adult height and cancer risk .
There are certain limitations in our study. First, there might be re-call bias due to the nature of a case-control study, and the weight and height data are self-reported. However, individuals might not ne-cessarily associate height or weight separately with the risk of HNC, thus recall bias might not be diﬀerential among the cancer cases and controls. Certainly for weight and height at age 20 for older individuals, recall error is possible, but we would not expect recall bias which would be diﬀerential recall for cases compared to controls. Secondly, there was limited statistical power in stratified analyses on various factors. Thirdly, data on human papilloma virus (HPV) infection were not available. We conducted stratified analysis by subsite since HPV is more strongly associated with oropharyngeal cancer, but the BMI results for oral cavity and oropharynx were similar to those for hypopharynx and larynx. In addition, there is no established association between HPV infection and BMI. Thus, we did not expect that HPV would act as a confounder in our study.
This is one of the few studies investigating BMI and HNC risk in an Asian population. Other strengths of our study include adjustment for betel quid chewing frequency and duration which are strong risk factors for HNC, especially for oral cavity cancer in Taiwan . We also ad-justed on other well-established risk factors, alcohol drinking and ci-garette smoking. Finally, we had BMI information from diﬀerent time points to investigate the impact of potential reverse causation where early signs of HNC might have impacted the patient’s diet. BMI at age 20 enabled us to assess the BMI association without the potential in-fluence of HNC symptoms prior to diagnosis.
In summary, our study provided evidence for an inverse association between BMI and HNC risk in an East Asian population. Being under-weight at age 20 was associated with a higher HNC risk among never cigarette smokers and never alcohol drinkers, whereas being under-weight at interview was associated with a much higher HNC risk among ever cigarette smokers and ever alcohol drinkers. Low BMI at a young age along with smoking and drinking may be factors that can be used to identify individuals at high risk for head and neck cancer in this po-pulation. Whether these results are specific to the generation of in-dividuals in our study needs to be studied further. Future studies on anthropometric measures such as waist-to-hip circumference in an East Asian population may further elucidate the possible association.
Dr. Yuan-Chin Amy Lee, Dr. Paolo Boﬀetta, Dr. Zuo-Feng Zhang and Dr. Mia Hashibe coordinated and designed the study. Yuji Chen and Shuang Li analyzed data. Yuji Chen drafted the manuscript. Dr. Yuan-Chin Amy Lee and Dr. Mia Hashibe also contributed to the manuscript writing. Dr. Qian Li, Dr. Chien-Jen Chen, Dr. Wan-Lun Hsu, Dr. Pen-Jen Lou, Dr. Cairong Zhu, Dr. Jian Pan, Dr. Hongbing Shen, Dr. Hongxia Ma, Dr. Lin Cai, Dr. Baochang He, Dr. Yu Wang, Dr. Xiaoyan Zhou, Dr. Qinghai Ji, Dr. Baosen Zhou, Dr. Wei Wu, Dr. Jie Ma and Dr. Min Dai collected the data and read the manuscript drafts. Dr. Daisuke Kawakita Cancer Epidemiology 60 (2019) 208–215
revised the manuscript critically. All authors read the manuscript and approved of it.
Conflicts of interest
This work was supported by the University of Utah School of Medicine, Department of Family and Preventive Medicine Marian Bishop Award, University of Utah Study Design and Biostatistics Center, with funding in part from the National Cancer Institute through Cancer Center Support P30 CA042014 awarded to Huntsman Cancer Institute, and the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant 8UL1TR000105 (formerly UL1RR025764).
Appendix A. Supplementary data
I. Holcatova, K. Kjaerheim, P. Lagiou, R.J. Lowry, T.V. Macfarlane, G.J. Macfarlane, B.E. McCartan, P.A. McKinney, F. Merletti, H. Pohlabeln, L. Richiardi, L. Simonato,