Estimating Age-specific Prevalence of Testosterone Deficiency in Men Using Normal Mixture Models

Yungtai Lo

doi:10.6339/jds.2009.07(2).463

Abstract

Testosterone levels decline as men age. There is little consensus on what testosterone levels are normal for aging men. In this paper, we estimate age-specific prevalence of testosterone deficiency in men using nor mal mixture models when no generally agreed on cut-off value for defining testosterone deficiency is available. The Box-Cox power transformation is used to determine which transformation is most appropriate for correcting skewness in data and best suits normal mixture distributions. Parametric bootstrap tests are used to determine the number of components in a normal mixture.

Highlights

Testosterone is a male sex hormone that helps maintain bone mass, fat distribution, male hair patterns, muscle mass and strength, sex drive and sperm production throughout male adult life
The age-specific prevalence of testosterone deficiency in men aged 50 to 59 with or at risk for human immunodeficiency virus (HIV) infection was estimated to be about 29% after adjusting for covariates. This estimated prevalence rate is lower than 61.9% estimated by using the cut-off point 325 ng/dl suggested by Harmen et al (2001), 55.4% using the cut-off point 300 ng/dl suggested by the recent endocrine society annual andropause consensus meeting, or 37.6% using the cut-off point 216 ng/dl suggested by Mohr et al (2005) for men in their 50s as the lower limit of the normal range for testosterone levels in these HIV at risk men
We found that injection drug use and use of psychotropic medications were associated with decreased total testosterone levels

Summary

Introduction

Testosterone is a male sex hormone that helps maintain bone mass, fat distribution, male hair patterns, muscle mass and strength, sex drive and sperm production throughout male adult life. Harman et al(2001) reported age-specific prevalences of hypogonadism of 12, 19, 28, and 49 percent for men in their 50s, 60s 70s, and 80s, respectively They defined hypogonadism as total testosterone levels < 325. Unlike the conventional approach for estimating prevalence, mixture models do not need a pre-specified cut-off value to classify each serum total testosterone. We use mixture models to account for individual testosterone levels that can arise either from a subpopulation with testosterone deficiency or from a subpopulation without testosterone deficiency, and to estimate the agespecific rate of testosterone deficiency in men in a particular age group when no generally agreed on cut-off value for that age group is available. We illustrate the use of the Box-Cox power transformation in mixture models to estimate age-specific prevalence of testosterone deficiency with a sample of serum total testosterone levels obtained from an HIV at-risk aging men’s prospective study.

Data and Model Description

Testing for the Number of Components

Adjustment for Covariates

Findings

Discussion