Premium
Multivariate relationships between epidemiologic risk factors and zoonotic infections among military personnel in the country of Georgia: A non‐linear canonical correlation analysis
Author(s) -
Akhvlediani Tamar,
Chitadze Nazibrola,
Chlikadze Rusudan,
Rostiashvili Nino,
Betashvili Medea,
Imnadze Paata,
Rivard Robert G.,
Nikolich Mikeljon P.,
Washington Michael A.,
Bautista Christian T.
Publication year - 2019
Publication title -
zoonoses and public health
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.87
H-Index - 65
eISSN - 1863-2378
pISSN - 1863-1959
DOI - 10.1111/zph.12632
Subject(s) - canonical correlation , leptospirosis , epidemiology , multivariate statistics , multivariate analysis , environmental health , veterinary medicine , categorical variable , medicine , demography , statistics , mathematics , sociology
Zoonotic diseases are endemic in the country of Georgia. Using the non‐linear canonical correlation (NCC) method, the aim of this study was to examine the relationship between thirteen epidemiological risk factors and seropositivity to five zoonotic infections (anthrax, Q fever, tularemia, leptospirosis, and Crimean‐Congo hemorrhagic fever [CCHF]) among Georgian military recruits during 2014–2016. According to this multivariate statistical technique, which is suitable for the analysis of two or more sets of qualitative variables simultaneously, two canonical variables were identified. These variables accounted for 68% of the variation between the two sets of categorical variables (“risk factors” and “zoonotic infections”). For the first canonical variable, there was a relationship among CCHF (canonical loading, which is interpreted in the same way as the Pearson's correlation coefficient, [cl] = 0.715), tick bites (cl = 0.418) and slaughter of animals (cl = 0.351). As for the second canonical variable, Q fever (cl = −0.604) and leptospirosis (cl = −0.486) were related to rodents inside and outside home (cl = −0.346) and sweeping in or around home (cl = −0.317). The NCC method allows researchers to obtain additional insights into the complex relationship between epidemiological risk factors and multiple zoonotic infections.