Global Advanced Research Journal of Medicine and Medical Sciences
July 2012 Vol. 1(6), pp. 154-162
Copyright © 2012 Global Advanced Research Journals
Full Length Research Paper
Age at cancer diagnosis in Malawi
Humphrey Misiri1,2 * and Abdi Edriss3
1Department of Biostatistics, Institute of Basic Medical Sciences, University of Oslo, Oslo, Norway.
2Department of Community Health, College of Medicine, Blantyre, Malawi.
3Bunda College, University of Malawi, Lilongwe, Malawi.
Accepted 28 June, 2012
Cancer which is one of the leading causes of death worldwide is emerging as a serious public health problem in Malawi due to the AIDS pandemic. Research has shown that HIV causes the syndrome of premature aging and accelerates carcinogenesis. The objective of this study was to describe age at cancer diagnosis and to fit the age distribution of childhood and adult cancer diagnosis in Malawi. We therefore fitted the normal, gamma, lognormal and inverse Gaussian probability distributions to the data for the 1996–2005 period from the Malawi National Cancer Registry and selected the model of best fit using the Akaike Information Criterion. Additionally, a finite mixture distribution of lognormals was also fitted to the data. According to the analysis for this study, the median ages at diagnosis are at most 42 years for AIDS-defining cancers and at least 46 years for non-AIDS defining cancers. Furthermore, the ages at childhood and adult cancer diagnosis follow lognormal distributions and the distribution of age at cancer diagnosis (all cancers) is a finite mixture distribution of lognormals with estimated mixing proportions equal to 0.071 and 0.929. The estimated means of the mixture distribution are 5.1 and 45.1 years and the corresponding estimated standard deviations are 1.211 and 2.842 years. This analysis suggests that age at cancer diagnosis in Malawi is relatively low and has a bimodal distribution. Therefore, to achieve maximum impact, cancer prevention and control activities should target the 15-50 year age range.
Keywords: cancer, diagnosis, AIC, finite mixture.