Bakker, J. (Joanne) (2017) The predictive value of left ventricular hypertrophy on ECG for cardiovascular outcome in the Thai population. thesis, Medicine.
![[img]](https://umcg.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
BakkerJ.pdf Restricted to Registered users only Download (590kB) |
Abstract
Background: In Thailand, the prevalence of cardiovascular disease (CVD) increased excessively during the past decades and is the leading cause of death. Left ventricular hypertrophy (LVH) can be seen on electrocardiogram (ECG) and is associated with poor cardiovascular outcome, but no incremental predictive value is seen when putting LVH in a traditional predicting model. Most studies on this specific subject were performed in Western populations. However, because Asians have a higher prevalence of LVH, we will explore the predictive value of LVH in the Thai population. When CVD prediction is more accurate, treatment can be started earlier and CVD events can be prevented. Method: ECG data of 2350 Thai citizens, participating in the EGAT cohort study, are analyzed. LVH on ECG is defined following the Sokolow-Lyon criteria, the Minnesota criteria and the Cornell criteria. Outcomes are fatal and nonfatal CVD events, coronary heart disease (CHD), cerebrovascular accident (CVA) and all-cause mortality. Multivariate Cox regressions are used to explore the association between LVH and outcome and tests for equality of ROC area are used for comparing predictive value of the models. Results: A significant association between Sokolow-Lyon LVH, Minnesota LVH and total LVH and the outcomes CVD, CHD and CVA is seen. In the multivariate Cox regressions only Sokolow-Lyon criteria and Minnesota criteria were associated with CVD [HR (95% CI): 1.86 (1.18-2.94) and 1.98 (1.22-3.21) respectively] and CHD [HR (95% CI): 1.92 (1.06-3.47) and 2.63 (1.48-4.67), respectively]. No incremental predictive value is observed when putting LVH in the traditional model for predicting CVD or CHD (all p-values >0.05). For predicting CVD, a model with Sokolow-Lyon criteria and BMI instead of fasting blood sugar (FBS) and non-HDL, both have the same predictive value [ROC areas (p-value): 0.69 and 0.70 (0.40), respectively], but these models had better predictive value than a standard model without these variables (p-values: 0.04 and 0.02, respectively). Conclusion: Even though LVH is significantly associated with CVD and CHD outcome, there is no incremental predictive value when the LVH variable is put in the model with traditional risk factors. For predicting CVD, there seems to be an option to replace variables that require blood samples (FBS and non-HDL) by Sokolow-Lyon LVH criteria and BMI.
| Item Type: | Thesis (Thesis) |
|---|---|
| Supervisor name: | Faculty supervisor: and Meer, P. van der and Cardiology UMCG |
| Supervisor name: | Second supervisor: P. Vathesatogkit and Cardiology Ramathibodi Hospital and Location: Department of Cardiology and Mahidol University, Ramathibodi and Hospital Bangkok |
| Faculty: | Medical Sciences |
| Date Deposited: | 25 Jun 2020 11:06 |
| Last Modified: | 25 Jun 2020 11:06 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/2614 |
Actions (login required)
 |
View Item |