Risk Occurred Probability and Inuencing Factors of Stroke in Followed-Up Hypertension Patients

Objective To explore the risk occurred probability and inuencing factors of stroke in followed-up hypertension (HP) patients through the analysis of long-term followed-up cohort data. Methods The method of followed-up observation cohort was used to collect the information of 168417 followed-up hypertension patients from 2002 to 2020 in Jiading district in Shanghai. Kaplan-Meier method was used to analyze the risk occurred probability of stroke complications in long-term followed-up HP patients, and the inuencing factors were analyzed by Cox proportional risk model.


Introduction
Stroke is an acute cerebrovascular disease caused by many factors, and it is the rst cause of death in China and the second largest cause of death in the world [1][2][3]. Hypertension, diabetes, dyslipidemia, cardiovascular disease and unhealthy lifestyle are all risk factors of stroke, among which hypertension is recognized as the main modi able risk factor of stroke [4][5][6]. The guidelines of the American Stroke Association also believe that controlling the risk factors of stroke is an effective prevention and control strategy for stroke [7]. Hypertension patients were prone to falling stroke, but not all hypertension patients would have stroke, and previous community population observation results showed that different hypertension patients have different epidemic characteristics and prevalence probability of stroke. In the followed-up population of hypertension patients, ischemic stroke was the most important type, rather than hemorrhagic stroke [8].
In order to better observe and explore the risk probability and risk factors of stroke in hypertension patients, the data of a long-term follow-up cohort of hypertensive patients in Jiading district in Shanghai was used in this study, and the risk probability and in uencing factors of stroke in followed-up hypertension patients was analyzed and explored. We hope that these results can provide some help for the prevention and control of stroke in community in the future. The Follow-up Management System of hypertension have been formally established and implemented in Jiading district in Shanghai China since from 2002. The Hypertension Follow-up Management System Database records the medical information of all registered and followed-up hypertension patients. According to the guidelines and policies of hypertension prevention and control, the hypertension patients in communities were registered and followed up, family doctors and public health personnel in the community health service center are responsible for the registration, follow-up and management of hypertension patients in the community, the center for disease control and prevention is responsible for the formulation of technical scheme, the training of personnel and the quality control of follow-up data. All patients were essential hypertension.
All medical followed-up information of patients must be truthfully recorded in the system database by follow-up doctors.

Data sources
All patients were from the Hypertension Follow-up Management System in Jiading district in Shanghai China, and the system database recorded the medical information of all registered and followed-up hypertension patients. The hypertension patients in community were registered, investigated, led and followed-up regularly by community health workers or family doctors. A total of 168 417 hypertension patients were entered into the database of the observation as the deadline of September 30, 2020. All about stroke occurred in followed-up patients must be recorded in detail. Stroke must be diagnosed and con rmed by a senior hospital. This study was approved by Science and Technology Commission, Health Commission and CDC in Jiading district in Shanghai. All methods were carried out in accordance with relevant guidelines and regulations.

Data Collection
The registered information and followed-up records of hypertension patients were recorded in the system, and collected from the system mainly include the date of birth, gender, ID number, residential address, occupation, education level, family history, date of establishment, termination date and reasons for termination, blood pressure and blood pressure at each follow-up, history of stroke (type of stroke, diagnostic date and diagnostic hospital), lifestyle (smoking, drinking and physical activity) and so on.
The date of occurrence of stroke in hypertension patients must be later than the date of registration. If the date of onset of stroke was earlier than or close to the date of registration, this patient data would not be included.

De nition And Classi cation
The diagnosis of stroke was based on the clinical diagnostic criteria (corresponding clinical symptoms plus positive results of head CT or MRI) [9,10,11]. The speci c diagnosis was made by the secondary and tertiary hospitals, and then the relevant information was collected by family doctors in the community. Stroke mainly included ischemic stroke, hemorrhagic stroke and unclassi ed stroke in this study. Ischemic stroke include transient ischemic attack, cerebral infarction. Hemorrhagic stroke included subarachnoid hemorrhage, intracerebral hemorrhage and other non-traumatic intracranial hemorrhage.
The measurement and classi cation of blood pressure and the diagnosis of hypertension were carried out according to the standard of Chinese guidelines for the prevention and treatment of hypertension [11][12][13]. The diagnosis of hypertension was implemented by advanced clinical hospitals. The hierarchical management of hypertension patients was divided into high-risk, medium risk and low-risk

Statistical analysis
According to the deadline, all the recorded data in the hypertension follow-up management system before the deadline were exported to the Microsoft Excel database, and then the corresponding logic check, data screening and conversion were carried out. Finally, the sorted database was imported into the SPSS statistics software package (IBM SPSS statistics version 21) for statistical analysis. The mean and standard deviation of age, blood pressure and other quantitative data were calculated. The number and frequency (%) were calculated according to gender, number of patients, etc. The chi-square (X 2 ) test was used to compare the data between different groups. The risk probability was analyzed by Kaplan-Meier method, and the related in uencing factors were analyzed by Cox regression. Bilateral test, P < 0.05 for the difference was statistically signi cant.

Baseline demographic characteristics
In this study, 11143 cases had developed stroke before the deadline among 168417 followed-up hypertension patients in Jiading district in Shanghai China. The complicated ratio of stroke among followed-up hypertension patients was 6.62%. Among them, 5636 cases of stroke occurred in male hypertension patients, and the complicated ratio of stroke was 6.87%. There were 5507 cases of stroke in female hypertension patients, and the complicated ratio of stroke was 6.37%. Male was higher than female (x 2 = 17.091, p < 0.001).
Among these observational hypertension patients, no stroke occurred in patients under 30 years old (the complicated ratio was 0), 21 cases of stroke occurred among the patients aged 30 ~ 39 years old (the complicated ratio was 0.68%), 250 cases of stroke occurred among the patients aged 40 ~ 49 years old (the complicated ratio was 1.70%), 1417 cases of stroke occurred among the patients aged 50 ~ 59 years old (the complicated ratio was 3.29%), 3629 cases of stroke occurred among the patients aged 60 ~ 69 years old (the complicated ratio was 5.93%), 3800 cases of stroke occurred among the patients aged 70 79 years old (the complicated ratio was 11.79%), 2026 cases of stroke occurred among the patients aged 80 years old and above (the complicated ratio was 14.63%). The older the patients, the higher complicated ratio of stroke (x 2 = 424.276, p < 0.001).
2118 cases of stroke occurred among smoking patients (the complicated ratio was 5.97%), 255 cases of stroke occurred among occasionally smoking patients (the complicated ratio was 5.21%), 976 cases of stroke occurred among patients who used to smoke but now don't smoke (the complicated ratio was 9.92%), 7794 cases of stroke occurred among never smoking patients (the complicated ratio was 6.59%).
There were signi cant differences in the complicated ratio of stroke among different smoking patients (x 2 = 213.188, p < 0.001).
8966 cases of stroke occurred among never drinking patients (the complicated ratio was7.00%), 1353 cases of stroke occurred among occasionally drinking patients (the complicated ratio was 5.40%), 622 cases of stroke occurred among frequently drinking patients (the complicated ratio was 6.04%), 222 cases of stroke occurred among patients who drink every day (the complicated ratio was 4.04%). There were signi cant differences in the complicated ratio of stroke among different drinking patients (x 2 = 149.764, p < 0.001).
4534 cases of stroke occurred among normal weight patients (the complicated ratio was 5.30%), 316 cases of stroke occurred among thin patients (the complicated ratio was 4.66%), 6293 cases of stroke occurred among overweight or obese patients (the complicated ratio was 18.46%). There were signi cant differences in the complicated ratio of stroke among different body mass index patients (x 2 = 149.764, p < 0.001).

Risk Probability Of Stroke In Hp Patients
To explore the change of risk occurred probability of stroke in hypertension patients during period of hypertension, the hypertension time (unit is year) was selected as the horizontal axis of time, and the occurrence of stroke (including ischemic stroke, hemorrhagic stroke and unclassi ed stroke) was de ned as the event state, and Kaplan-Meier method was used to analyze the risk occurred probability of stroke.
The calculated risk probability of each time point was made as a longitudinal axis and a dynamic curve was drawn. See Fig. 1. The result of Fig. 1 showed that the risk probability of stroke in HP patients was not xed during the whole hypertension period, but uctuates. There were 4 peak onset periods, which were in 8 years (peak value 0.042, 4.2%), 15 years (peak value 0.140, 14.0%), 22 years (peak value 0.060, 6.0%) and 26 years (peak value 0.139, 13.9%). The highest risk of male patients was in 26 years (peak value 0.231, 23.1%), and the second peak was in 15 years (peak value 0.156, 15.6%). The highest risk of female patients was in 15 years (peak value 0.129, 12.9%), and the second peak was in 26 years (peak value 0.087, 8.7%).
If cumulative occurred probability was taken as the vertical axis (see Fig. 2), the result showed that cumulative risk occurred probability of stroke in male and female patients would continue to increase with the time extension of hypertension years, and the proportion of increase was not equidistant. The occurred probability was higher in men than that in women (x 2 = 113.570, p < 0.001). The total cumulative risk probability of stroke in hypertension patients was 0.789 (78.9%), and male was 0.910 (91.0%), female was 0.707 (70.7%).
The result of Fig. 3 showed that the cumulative risk probability of stroke in different blood pressure (BP) patients would continue to increase with HP time. The occurred probability was obviously higher in grad 3 (SBP ≥ 180mmHg or/andDBP ≥ 110mmHg) patients than that in grad 2 (SBP:160 ~ 179mmHgor/andDBP:100 ~ 109mmHg) and grad 1 (SBP:140~ 159mmHg or/and DBP:90 ~ 99mmHg) patients and (x 2 =189.139, p < 0.001). Figure 4 showed that the cumulative risk probability of stroke in different body mass index (BMI) patients would continue to increase with hypertension time. The risk probability was higher in thin (BMI < 18.5) patients than that in normal weight (BMI: 18.5 ~ 23.9) and overweight (BMI ≥ 24.0) patients (x 2 = 136.724, p < 0.001). The risk probability of stroke in normal weight patients would suddenly increase after hypertension 22 years and so on, and the reason was unknown.
The result of Fig. 5 showed that the cumulative risk probability of stroke in different family history (FH) patients would continue to increase with hypertension time. But the occurred probability wasno difference between hypertension patients with positive and negative family history of stroke(x 2 = 2.432, p =0.119).

In uencing Factors Of Stroke
In this study, 10 factors including sex, age, hypertension duration, smoking habit, drinking habit, physical activity, body mass index, blood pressure grad and family history (including family history of hypertension and stroke) were selected. Cox regression analysis was used to explore the in uence of these factors on the occurrence of stroke complications. From the results of Cox regression analysis, these factors have different degrees of in uence (p < 0.05). The major in uencing factors of stroke complications were age (RR = 2.917, p < 0.001), body mass index (RR = 1.450, p < 0.001), family history of stroke (RR = 1.386, p < 0.001), blood pressure grad (RR = 1.148, p < 0.001), hypertension duration (RR = 1.071, P < 0.001) and family history of hypertension (RR = 1.051, P < 0.001). The lower limit value of RR of sex, smoking habit, drinking habit and physical activity were less than 1. See Table 1.

Discussion
The main harm of hypertension lies in its complications. In the incidence spectrum of complications in hypertension patients, the constituent ratio from high to low was cerebrovascular diseases (49.61%), diabetes mellitus (34.94%), heart damage (13.93%), kidney diseases (1.04%), fundus damage (0.30%) and peripheral vascular diseases (0.04%). Cardiovascular and cerebrovascular diseases were the main complications in patients with hypertension in community [14]. In this study, the cumulative incidence rate of stroke in hypertensive patients was 6.62%, the male was 6.87% and the female was 6.37%.. The incidence rate of stroke in hypertensive patients was much higher than that incidence rate in general population [1,3,15]. Therefore, It's an important link in community stroke prevention and control that strengthen the blood pressure control of patients with hypertension and prevent the occurrence of stroke in patients with hypertension. What needs to be explained here again is that Jiading District government began to implement the free medication policy for hypertension patients with rural medical insurance as early as around 2000. Later, due to the cancellation of rural household registration and the implementation of urban population management policy in Shanghai, the free medication policy gradually faded out, but it is still open to poor households. The vast majority of patients were taking antihypertensive drugs in followed-up hypertension patients in Jiading Shanghai, only a few patients were not willing to accept drug treatment. Assuming these patients did not receive antihypertensive treatment, the incidence of stroke might be higher.
At present, there are many reports on the risk probability analysis or prediction of stroke with hypertension, and the methods used include Logistic Regression analysis, Framinghan evaluation method and ESRs score method [16][17][18][19]. The results of different methods are different. Kaplan-Meier method was used in this study, which is also called product limit method. It is not only suitable for small sample data, but also suitable for large sample data. In addition, this method can take into account the objects exiting the observation queue or incomplete censored data, so it is more applicable. Through the analysis of long-term observation data, it was found that with the extension of hypertension years, the cumulative risk probability of stroke in hypertension patients would continue to increase, and the increasing distance was not equidistant. The total cumulative occurred probability of stroke in hypertension patients was 0.789 (78.9%), and male was 0.910 (91.0%), female was 0.707 (70.7%). Male was higher than female. The observation results remind us that hypertension patients must control their blood pressure as soon as possible. With the extension of the course of disease, the risk of stroke will increase, and nally reach a very high risk probability, especially male patients.
The risk occurred probability of stroke in hypertension patients was not xed during the whole hypertension years, but uctuates. There were 4 onset peaks, which were in 8 years (peak value 0.042, 4.2%), 15 years (peak value 0.140, 14.0%), 22 years (peak value 0.060, 6.0%) and 26 years (peak value 0.139, 13.9%). Why did the risk probability of stroke in hypertension patients show four different peaks instead of one-way change, it might be the result of long-term con ict between antihypertensive treatment and vascular damage accumulation in hypertension patients, and it might be accompanied by other reasons, which needs further and more study. It is di cult to observe the dynamic changes of hypertension patients in the natural state in practical work based on medical ethics. The highest risk probability of male patients was in 26 years (peak value 0.231, 23.1%), and the second peak was in 15 years (peak value 0.156, 15.6%). The highest risk of female patients was in 15 years (peak value 0.129, 12.9%), and the second peak was in 26 years (peak value 0.087, 8.7%). If hypertensive patients did not receive antihypertensive treatment, the risk probability of stroke might be higher, and the peak time point might be more advanced.
Stroke is a group of cerebrovascular circulation disorders caused by various reasons, manifested as focal neurological de cit, and even accompanied by disturbance of consciousness. Because of its sudden onset, it is also called cerebrovascular accident. There were many risk factors for stroke, such as age, heredity, hypertension, heart disease, arrhythmia, diabetes, hyperlipidemia, smoking, drinking, obesity, high salt, high animal oil diet, excessive physical activity, etc. [9,[20][21][22][23][24][25][26][27]. The risk of stroke was higher in the hypertension and diabetes population. From the results of Cox regression analysis, these factors have different degrees of in uence (p < 0.05). According to the results of the long-term observation of registered hypertension patient, the major in uencing factors of stroke complications were age (RR = 2.917), body mass index (RR = 1.450), family history of stroke (RR = 1.386), blood pressure grad (RR = 1.148,), hypertension duration (RR = 1.071) and family history of hypertension (RR = 1.051). These results were consistent with the above literature reports. The in uence of sex, smoking habit, drinking habit and physical activity on stroke complications was statistically signi cant, the lower limit value of RR of were less than 1. This also means that the direction of action of these factors was protective factors or opposite effect. The reason might be affected by the assignment of data classi cation, for example, male was assigned as 1 and female was assigned as 2 in sex classi cation, but the risk of male was actually high than that of female. It might also be caused by other factors. This needs further study, but it is undeniable that sex, smoking habit, drinking habit and physical activity are the in uencing factors of stroke, which must be paid attention to in community prevention and control.
In short, hypertension patients were prone to stroke, and the risk probability would continue to disproportionately increase during observation, and the total cumulative risk probability could nally reach 78.9%. Male was higher than female. Age, BMI, family history (stroke and HP), blood pressure grad and duration of HP were related to the occurrence of stroke complications. Early control of patients' blood pressure and prevention and control of hypertension related risk factors is an important part and means of community stroke prevention and control.

Declarations
This study was approved by Science and Technology Commission, Health Commission and CDC in Jiading district in Shanghai. All methods were carried out in accordance with relevant guidelines and Page 10/17 regulations.