Limits...
A mechanistic model for atherosclerosis and its application to the cohort of Mayak workers

View Article: PubMed Central - PubMed

ABSTRACT

We propose a stochastic model for use in epidemiological analysis, describing the age-dependent development of atherosclerosis with adequate simplification. The model features the uptake of monocytes into the arterial wall, their proliferation and transition into foam cells. The number of foam cells is assumed to determine the health risk for clinically relevant events such as stroke. In a simulation study, the model was checked against the age-dependent prevalence of atherosclerotic lesions. Next, the model was applied to incidence of atherosclerotic stroke in the cohort of male workers from the Mayak nuclear facility in the Southern Urals. It describes the data as well as standard epidemiological models. Based on goodness-of-fit criteria the risk factors smoking, hypertension and radiation exposure were tested for their effect on disease development. Hypertension was identified to affect disease progression mainly in the late stage of atherosclerosis. Fitting mechanistic models to incidence data allows to integrate biological evidence on disease progression into epidemiological studies. The mechanistic approach adds to an understanding of pathogenic processes, whereas standard epidemiological methods mainly explore the statistical association between risk factors and disease outcome. Due to a more comprehensive scientific foundation, risk estimates from mechanistic models can be deemed more reliable. To the best of our knowledge, such models are applied to epidemiological data on cardiovascular diseases for the first time.

No MeSH data available.


Related in: MedlinePlus

Age dependent hazard and relative risk of hypertension.In both panels, the solid line refers to the mechanistic model while the empirical model is shown by a dashed line and with 95% confidence interval indicated in gray. The plots refer to a worker born in 1930, smoking and without higher education. A: Hazard of a worker with normal blood pressure. The kink after age 60 is related to a calendar year effect around the time of the transition from the Soviet Union to Russia. B: Ratio of the hazard of a worker with hypertension compared to the hazard of a worker with normal blood pressure.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5383300&req=5

pone.0175386.g003: Age dependent hazard and relative risk of hypertension.In both panels, the solid line refers to the mechanistic model while the empirical model is shown by a dashed line and with 95% confidence interval indicated in gray. The plots refer to a worker born in 1930, smoking and without higher education. A: Hazard of a worker with normal blood pressure. The kink after age 60 is related to a calendar year effect around the time of the transition from the Soviet Union to Russia. B: Ratio of the hazard of a worker with hypertension compared to the hazard of a worker with normal blood pressure.

Mentions: The age dependence of the hazards for both models is illustrated in Fig 3A. The agreement is very good for those ages in which many cases of first incidence of stroke occurred: The median cohort age for first stroke incidence is 64 years, the 2.5% and 97.5% quantiles are 43 and 83 years. Strong deviations exist for ages below 30 years where the hazard of the empirical model acquires very small values. While about two such young cases would be expected from the mechanistic model, only about one case is predicted by the empirical model. As there are no such young cases in the cohort, the empirical model fits the data slightly better (by 2 points in the deviance).


A mechanistic model for atherosclerosis and its application to the cohort of Mayak workers
Age dependent hazard and relative risk of hypertension.In both panels, the solid line refers to the mechanistic model while the empirical model is shown by a dashed line and with 95% confidence interval indicated in gray. The plots refer to a worker born in 1930, smoking and without higher education. A: Hazard of a worker with normal blood pressure. The kink after age 60 is related to a calendar year effect around the time of the transition from the Soviet Union to Russia. B: Ratio of the hazard of a worker with hypertension compared to the hazard of a worker with normal blood pressure.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5383300&req=5

pone.0175386.g003: Age dependent hazard and relative risk of hypertension.In both panels, the solid line refers to the mechanistic model while the empirical model is shown by a dashed line and with 95% confidence interval indicated in gray. The plots refer to a worker born in 1930, smoking and without higher education. A: Hazard of a worker with normal blood pressure. The kink after age 60 is related to a calendar year effect around the time of the transition from the Soviet Union to Russia. B: Ratio of the hazard of a worker with hypertension compared to the hazard of a worker with normal blood pressure.
Mentions: The age dependence of the hazards for both models is illustrated in Fig 3A. The agreement is very good for those ages in which many cases of first incidence of stroke occurred: The median cohort age for first stroke incidence is 64 years, the 2.5% and 97.5% quantiles are 43 and 83 years. Strong deviations exist for ages below 30 years where the hazard of the empirical model acquires very small values. While about two such young cases would be expected from the mechanistic model, only about one case is predicted by the empirical model. As there are no such young cases in the cohort, the empirical model fits the data slightly better (by 2 points in the deviance).

View Article: PubMed Central - PubMed

ABSTRACT

We propose a stochastic model for use in epidemiological analysis, describing the age-dependent development of atherosclerosis with adequate simplification. The model features the uptake of monocytes into the arterial wall, their proliferation and transition into foam cells. The number of foam cells is assumed to determine the health risk for clinically relevant events such as stroke. In a simulation study, the model was checked against the age-dependent prevalence of atherosclerotic lesions. Next, the model was applied to incidence of atherosclerotic stroke in the cohort of male workers from the Mayak nuclear facility in the Southern Urals. It describes the data as well as standard epidemiological models. Based on goodness-of-fit criteria the risk factors smoking, hypertension and radiation exposure were tested for their effect on disease development. Hypertension was identified to affect disease progression mainly in the late stage of atherosclerosis. Fitting mechanistic models to incidence data allows to integrate biological evidence on disease progression into epidemiological studies. The mechanistic approach adds to an understanding of pathogenic processes, whereas standard epidemiological methods mainly explore the statistical association between risk factors and disease outcome. Due to a more comprehensive scientific foundation, risk estimates from mechanistic models can be deemed more reliable. To the best of our knowledge, such models are applied to epidemiological data on cardiovascular diseases for the first time.

No MeSH data available.


Related in: MedlinePlus