Cardiovascular and Cerebrovascular Disease Incidence Among 42785 Adults: The Thai Cohort Study, 2005-2013

Results: The cumulative incidence of CVD and/or CVA in males was more than three times that in females. CVD and/or CVA incidence was correlated with ageing, obesity (AOR: 1.67, 95% CI: 1.16-2.40) and previous diagnosis with diabetes (AOR: 3.09, 95% CI: 1.80-5.30), hyperlipidaemia (AOR: 1.54, 95% CI: 1.08-2.19), hypertension (AOR: 1.71, 95% CI: 1.13-2.59), chronic kidney disease (AOR: 2.35, 95% CI: 1.35-4.10), and depression/anxiety (AOR: 2.76, 95% CI: 1.64-4.63). Short sleep time was positively associated with CVD and/or CVA in the Thai Cohort Study. An inverse association between performing housework and the incidence of CVD and/or CVA was also identified. However, current smoking had a significant positive correlation with the incidence of CVD and/or CVA for participants.


Introduction
Cardiovascular disease (CVD) and cerebrovascular disorders (CVA) are the ischemic or haemorrhagic disorders of the heart, brain and systemic tissues caused by conditions including hyperlipidaemia, blood viscosity, atherosclerosis, and hypertension (World Health Organization [WHO], 2019a). CVD and CVA are serious threats to human health due to high morbidity and high mortality. CVD and CVA are the leading causes of death globally, and according to the report of the World Health Organisation (WHO), over 13 million CVD and CVA deaths occur in low and middle income countries and regions, accounting for about 75% of the total global CVD and CVA deaths (WHO, 2019b).
Thailand, as one of the developing countries in Southeast Asia, bears a huge burden of CVD and CVA, just like other middle-income nations around the world (Kaptoge et al., 2019). Stroke and ischaemic heart disease are two cardiovascular and cerebrovascular illnesses with the highest morbidity and mortality. From 2005 to 2017, the Disability-Adjusted Life Year (DALY) loss caused by stroke increased by about 100,000, while that led by university's bachelor degree or above.
The third category was health conditions such as obesity and overweight, diabetes, hyperlipidaemia, hypertension, chronic kidney disease and emotional health problems (depression and anxiety) that could increase the risk of CVD and/or CVA. Obesity and overweight were defined using participants' body mass index (BMI). The Asian criteria of BMI was used in the study: BMI less than 18.5 (underweight), BMI between 18.5 and 23.0 (normal), BMI between 23.0 and 25.0 (overweight), BMI more than 25.0 (obese) (Banks, Lim, Seubsman, Bai, & Sleigh, 2011).
The fourth category were the personal lifestyle factors variables. The frequency of diet consumption (deep fried food, roasted/smoked food, instant foods, soft drink and soybean products) was categorized into five groups: less than one time per month (1), 1-3 times per month (2), 1-2 times per week (3), 3-6 times per week (4), daily or more (5). Western fast food was classified three groups: less than one time per month (1), less than 3 times per month (2) and over 3 times per month (3). The intake of fruit and vegetables would be noted as standard serves each day.
As for exercise per week, there are four groups, namely, "strenuous exercise for more than 20 minutes", "moderate exercise for more than 20 minutes", "mild exercise for more than 20 minutes" and "walking for at least 10 minutes". We calculated in the adjusted exercise measure weighting to evaluate the weekly physical activities: "2 × vigorous exercises + 1 × moderate exercises + 1 × walking sessions per week", which was based on the International Physical Activity Questionnaire and the Active Australia Survey methodology (Banks et al., 2011;Thawornchaisit et al., 2013;Zhao et al., 2015). Housework was measured as another form of physical activity. The frequency of this was divided into 5 groups: never, 1-3 times per month, 1-2 times per week, 3-4 times per week and most days. The length of time of watching TV and/or computer and sitting for any purpose (the number of hours per day) was analysed in this study to evaluate sedentariness. The length of time of sleeping (the number of hours per day) was also measured.

Statistical Analysis
All analyses were performed using Stata 14.0. The 8-year CVD and/or CVA incidence and its 95% CI were computed separately for male and female participants. Differences between CVD and/or CVA incidence for risk factor groups were assessed using the χ 2 test. Statistical significance was defined as p value less than 5%.
Logistic regression was used to analyse the association between the health risk factors and the incidence of CVD and/or CVA from 2005 to 2013. Due to the stratified analysis of gender, the first step was to use bivariate logistic regression to analyse the association between health risk factors measured at baseline and CVD and/or CVA incidence over the following 8 years. Multivariate logistic regression models were used to assess the multivariate adjusted odd ratios (ORs). For each set of associations, we adjusted for a different set of confounders based on observations on our descriptive data and from other literatures. The details were included on the Suppl. Table 1.

Ethical Considerations
Ethical approval was provided by the Sukhothai Thammathirat Open University Research and Development Institute and Human Research Ethics Committee of the Australian National University. Table 1 describes the baseline characteristics of the Thai cohort study in 2005. It was clear that there were more women participants than the men. In terms of the marital status, the number of singles and cohabitants was similar, accounting for 49.8% and 50.2% of the total participants respectively. The ratio of single women was higher than that of men. Moreover, the female educational attainment (31.4% at university level) was higher than males (28.2% at university level) in 2005. However, as for the personal salary, around 16.5% of men earned over 20000 baht in one month while only less than 10% of women reached this level.

The Cumulative Incidence of CVD and/or CVA
As for diagnosed CVD and/or CVA in the Thai Cohort Study in 2013 (Table 2), the cumulative incidence of CVD and/or CVA among the whole participants was 0.52% and the incidence in males was more three times than that in females (0.84% vs. 0.26%; see the Table 2.). In men (p-trend= 0.008) and women (p-trend= 0.0196) as well as the total study population (p-trend= 0.0146), the incidence of such diseases increased with age. Furthermore, people living with a partner (male: 1.02%; female: 0.36%) were more susceptible to CVD and/or CVA than single persons (male:0.48%; female: 0.19%) in both genders. It was worth noting that participants who moved from the urban area to the countryside (UR) were more likely to suffer from these diseases (RR:0.84% vs. RU: 0.58% vs. UU: 0.63% The incidence of the diseases of interest was highest in people who consumed smoked food once or more times per day (0.82%). In addition, drinking soft drinks once or several times per day also increased incidence of CVD and/or CVA among the total participants (0.67%) (see the Suppl. Table 2.2.). However, the more vegetables the person ate, the lower incidence of CVD and/or CVA for this individual (p-trend= 0.0460). Moreover, the incidence of these diseases was also highest in people who smoked (current smoker vs. never:1.23% vs. 0.33%) or drank alcohol regularly (current regular vs. never: 1.19% vs. 0.37%) ( Table 2.).

The Association between Risk Factors and CVD and/or CVA among Participants
There was a linear positive relationship between age and CVD and/or CVA in males (p-trend= 0.0307) and total participants (p-trend= 0.028) after controlling for other variables (Table 3.). However, after controlling for variables including age, marital status and other disease diagnoses, there was no association between personal monthly income and CVD and/or CVA in the total members of the study. Additionally, monthly home income was not be a health risk factor of CVD and/or CVA in men or women separately.
Among all participants, obese individuals had a higher likelihood of CVD and/or CVA incidence when compared with those of normal body size  (Table 3.). However, most of the personal lifestyle factors such as physical activities and food intake habits (deep-fried food, instant food, soft drink, roasted/smoked food, soybean products, western food, vegetables, and fruits) did not have significant relationships with CVD and/or CVA (see the Suppl. Table 3.1. and the Suppl. Table 3.2.).
However, it was noteworthy that participants whose sleeping time averaged less than 6 hours had positive associations with CVD and/or CVA (AOR: 1.47, 95% CI: 1.07-2.01) after adjusting for other personal lifestyle elements, socio-demographic factors and health conditions. Moreover, housework (household cleaning or gardening work), one type of physical activity, was a protective factor against such diseases (Table 3).
It was obvious that current smoking increased by about 1.58 times the odds of catching CVD and/or CVA compared to those who did not smoke. Conversely, alcohol consumption had little remaining effect on these diseases (Table 3).

Discussion
Following economic and social development in Thailand, health problems for the population are gradually transferring from infectious diseases to chronic diseases (Khatab et al., 2019;Sleigh, Seubsman, Bain, & Thai Cohort Study Team, 2008;Thawornchaisit et al., 2013;Zhao et al., 2015). CVD and CVA are now common in this country with increasing incidence every year (Banks et al., 2011). This increased incidence is potentially being driven by many factors. We studied the relationship between CVD and/or CVA and risk factors and found that age, obesity, diabetes, hyperlipidemia, hypertension, chronic kidney disease, short sleep time, smoking and mental illnesses were closely related to the incidence of CVD and/or CVA. However, housework, as a form of incidental exercise, could protect people against CVD and/or CVA.

Underlying Diseases
Obesity is recognized as one of the major causes of CVD and CVA. The result of the study is consistent with the other longitudinal studies in Asia, which have reported that in obese patients the risk of incident CVD rose by from 27% to 97%. Obesity can directly affect the heart, causing ventricular hypertrophy and diastolic dysfunction. Moreover, obesity can also induce diabetes, high blood lipids and high blood pressure (Chen et al., 2019). These existing chronic conditions, as well as kidney disease are also associated with the development of CVD and CVA in this Thai population. Similar cohort studies among Asian populations have also reported these relationships (Ueshima et al., 2008;Wu et al., 2015). All of these conditions have vascular and cardiovascular effects which can increase CVD/CVA risk.
Renal failure in particular can affect heart function, leading to ischemic heart disease and possibly leading to stroke (Dad & Weiner, 2015). The result is supported by a cohort study in the USA (adjusted risk ratio (ARR): 2.3, 95% CI:1.2-3.4) (Bansal et al., 2017). Moreover, our results are also consistent with the systematic review of a cohort study of chronic kidney disease and the risk of stroke (adjusted risk ratio (ARR): 2.18, 95% CI: 1.68-2.84) (Masson et al., 2015).
The relationship between depression and CVD/CVA in this study is also worth noting. The American Heart Association suggests that depression or anxiety should be considered a major risk factor for coronary heart disease (Chaddha, Robinson, Kline-Rogers, Alexandris-Souphis, & Rubenfire, 2016). This may be because the prevalence of CVD/CVA risk factors such as smoking and alcohol consumption is higher in depressed patients (Dhar & Barton, 2016). Nevertheless, in our study, after controlling for smoking and drinking, mental illnesses are still positively associated with such diseases. This may be due to changes in neurological and endocrine function (Chaddha et al., 2016), however, the association between mental disorders and CVD and CVA is still under investigation.

Personal Lifestyle Factors
We found that current smoking was significantly associated with increased risk of developing CVD and/or CVA in male participants, which is consistent with much other research (Messner & Bernhard, 2014;Tan et al., 2018).
However, no association was found between alcohol consumption and diagnosed CVD and/or CVA. This may be due to the small number of people diagnosed with CVD and CVA, resulting in insufficient statistical power to detect relationships. However, in global longitudinal studies, alcohol has been shown to increase the risk of CVD and CVA development, namely, after controlling for other risk factors including age, SES, family disease history, and smoking, the larger amount of alcohol a person intakes, the higher the risk of stroke and coronary heart disease the individual gets (Wood et al., 2018).
Short sleep time was positively associated with CVD and/or CVA in the Thai cohort study. Furthermore, it is worth noting that people with sleep time less than 7 hours or more than 8 hours have a higher incidence of CVD and/or CVA than normal duration of sleeping times, which is consistent with a cross-sectional study in the USA (Buxton & Marcelli, 2010). Sleeping too much or too little could increase the probability of suffering from CVD and CVA (Buxton & Marcelli, 2010). However, the cause of these phenomena warrants further study.
Sedentary time did not have any effect on CVD and/or CVA in this cohort study. However, physical activity in the form of housework protected participants against CVD and/or CVA; no relationship was found with purposeful physical activity. Housework reducing the risk of CVD and/or CVA is consistent with other research among Asian Indian populations (Nag & Ghosh, 2013). Exercise may improve the function of blood vessels and inhibit arteriosclerosis, thereby reducing the risk of these illnesses (Nystoriak & Bhatnagar, 2018).

Socio-Demographic Factors
We also note that the influence of age on the incidence of such diseases is consistent with the results of many similar studies. For example, a cross-sectional study in Thailand (2008Thailand ( -2009 showed that the older age of people was, the higher the risk of stroke and coronary heart disease (Khatab et al., 2019). In TCS, the cumulative incidence of CVD and/or CVA in males was more three times than that in females, which is similar with the worldwide CVD study that found men in Southeast Asia are at higher risk of developing such diseases than women (Roth et al., 2017). It is a known factor that males' incidence of CVD and CVA is higher than females. Nevertheless, as TCS participants may not represent the total Thai population, the results found here will not reflect information about the incidence of these diseases in males being higher than that in females nationally (Roth et al., 2017;Thawornchaisit et al., 2013;Thawornchaisit et al., 2015).
In many studies in various settings socio-economic status in terms of income education and assets are associated with CVD and CVA risk. This relationship was not found in our cohort study. This may be due to the fact that our participants were all Open University students and displayed less extremes of high and low socio-economic status. As well, socio-economic status indicators are also often proxies for differences in biological and personal lifestyle risk factors, for example, higher income TCS members also have a higher average age.
However, low-SES is often associated with more risk taking behavior in terms of smoking and alcohol and also poorer diets, which increases the risk of coronary heart disease and stroke, compared with other people (Psaltopoulou et al., 2017).

Strengths and Limitations
In this study, we found the relationship between risk factors and diagnosed CVD and CVA. The TCS is a relatively mature and comprehensive study of health transitions in Thailand. This is the first study of the relationship between CVD and CVA and related factors from 2005 to 2013 in TCS.
Although the sample size of this study is large, the population coverage is relatively wide, and the research duration is relatively long, there are still some limitations in our study.
Firstly, our participants were all open university students at baseline, which means that they may not be able to represent the entire Thai population. Also, the participants are younger and more educated and may not manifest CVD and CVA until later in life. This means that incidence may be lower in the TCS than the general population. Thus, TCS participants may thus represent the next generation of people in Thailand (Thawornchaisit et al., 2013;Thawornchaisit et al., 2015). As well, despite the cohort population not being representative of the Thai national population, the associations found between risk factors and disease outcomes in the cohort are still valid and applicable. Moreover, in this 8-year study, we inferred the risk factors associated with CVD and CVA based on the characteristics of the participants in the baseline (2005), which may cause some bias because participants may change their habits within 8 years, resulting in inaccurate results. Furthermore, this study relies on self-reporting, which may lead to recall bias. This kind of bias belong to systematic error, and it is impossible to eliminate the error by increasing the sample size, thus it may have a certain impact on the results presented here.

Conclusion
In summary, older age, obesity and some disease factors such as hyperlipidemia, hypertension, diabetes, chronic kidney disease, and depression/anxiety were closely related to increased incidence of CVD and/or CVA in men and women. There was a positive relationship between short sleep time and CVD and/or CVA in the Thai cohort study.
Although the purposeful physical activities had less effect on these diseases, housework had an inverse correlation with the incidence of CVD and/or CVA among the participants. However, current smoking is still the main risk factors for incidence of CVD and/or CVA.

Recommendations
In addition to controlling smoking and drinking, the Thai government should devote attention to those diagnosed with mental illnesses such as depression/anxiety and other chronic co-morbidities to help relieve the burden of CVD and CVA in the Thai population. Currently, Thailand has a relatively complete diagnostic system for depression (Sharan, Sagar, & Kumar, 2017). Furthermore, the authorities should also advocate that the incidental house and garden work type exercise as just as beneficial as intentional exercise, and urge people to actively participate in physical exercise, thereby reducing the incidence of CVD and CVA. Wu, C., Hu, H., Chou, Y., Huang, N., Chou, Y., & Li, C. (2015). High blood pressure and all-cause and cardiovascular disease mortalities in community-dwelling older adults. Medicine, 94 (47)   Note. AOR*: Adjusted Odd ratios from logistic regression models of CVD and/or CVA controlling adjusted for socio-demographic factors, health conditions, personal lifestyle factors. AOR bolding: The results of AOR were statistical significance.