期刊名称：Pakistan Journal of Statistics and Operation Research
印刷版ISSN：2220-5810
出版年度：2020
卷号：16
期号：2
页码：265-277
DOI：10.18187/pjsor.v16i2.3075
语种：English
出版社：College of Statistical and Actuarial Sciences
摘要：Demographic and socio-economic (SE) factors are associated with modifiable, clinical factors and cardiovascular disease (CVDs). However, the inclusion of mediators in these relationships creates complex pathways which extend the roles of factors within the model. The traditional hierarchal logistic regression (HLR) is unable to estimate the transmittal effects of factors through different layers of factors of CVDs. The study aims to simultaneously estimate and validate the probable linear and non-linear relationships among factors of CVDs considering potential mediators. Four hundred sixty participants (312 males and 148 females) were selected through systematic sampling in this sex-matched case control (1:1 ratio) study conducted in the largest Cardiac Center of Pakistan. The information on demographic, SE, modifiable and clinical factors of CVDs was recorded. Warp partial least squares (PLS) based on warp 3 algorithm was used to estimate the simultaneous linear and non-linear path coefficients of the proposed model of study. The study found that demographic and SE factors played a significant role in shaping the modifiable factors which further transmit their impact to CVDs. However, this transmitted impact of modifiable factors on CVDs was mediated through metabolic syndrome abnormalities (MSA) except self-reported subjective stress (SSS). Sleep satisfaction and negative dietary habits were the mediators between the relationship of SSS and MSA. Physical activity is the strongest factor associated with CVDs status. The proposed path analyses, verifying the mediation role of MSA in the pathways of relationships which would help in identifying the risky group of population and guide in formulating the health promotion policies for the reduction of CVDs burden. Further, Warp 3 algorithm is the better option to estimate complex models containing linear and non-linear relationship in the same model.
关键词：Cardiovascular Diseases;Risk Factors;Path Analyses;Partial Least Squares;Warp 3 Algorithm;Mediation
