| dc.description.abstract | Introduction Various studies have identified TB-induced metabolome variations. However, in most of these studies, a large
degree of variation exists between individual patients.
Objectives To identify differential metabolites for TB, independent of patients’sex or HIV status.
Methods Untargeted GCxGC/TOF-MS analyses were applied to the sputum of 31 TB+and 197 TB- individuals. Univariate
statistics were used to identify metabolites which are significantly different between TB+and TB- individuals (a) irrespective of HIV status, and (b) with a HIV+status. Comparisons a and b were repeated for (i) all participants, (ii) males only
and (iii) females only.
Results Twenty-one compounds were significantly different between the TB+and TB- individuals within the female subgroup (11% lipids; 10% carbohydrates; 1% amino acids, 5% other and 73% unannotated), and 6 within the male subgroup
(20% lipids; 40% carbohydrates; 6% amino acids, 7% other and 27% unannotated). For the HIV+patients (TB+vs. TB-), a
total of 125 compounds were significant within the female subgroup (16% lipids; 8% carbohydrates; 12% amino acids, 6%
organic acids, 8% other and 50% unannotated), and 44 within the male subgroup (17% lipids; 2% carbohydrates; 14% amino
acids related, 8% organic acids, 9% other and 50% unannotated). Only one annotated compound, 1-oleoyl lysophosphaditic
acid, was consistently identified as a differential metabolite for TB, irrespective of sex or HIV status. The potential clinical
application of this compound should be evaluated further.
Conclusions Our findings highlight the importance of considering confounders in metabolomics studies in order to identify
unambiguous disease biomarkers | en_US |
