# Cardiovascular Diseases Risk (incl. Heart Attack & Stroke Risks)

Future health risks evaluate the user’s risk of developing a condition in the intervening time between the present day and some duration of time into the future.

## Description 

This risk prediction model estimates an individual’s likelihood (in percent) of developing cardiovascular disease (specifically their **first heart attack** or **stroke**) within the next 10 years. It does not apply to people who have already had a heart attack or stroke.

## Additional Information Required

• Body Mass Index (kg/m2 ) (calculated from height and weight) 

• Systolic Blood Pressure (mmHg)

Estimates are more accurate when the user also answers the following questions in their user profile: 

1. Are you taking medication for high blood pressure? (yes/no) 
2. Are you currently a smoker? (yes/no) 
3. Do you have diabetes? (yes/no)

## Participants 

Risk prediction models were derived from and tested on the same sample of Chinese adults (18+ years of age).

## Additional Data Collection 

Subjects were also asked whether they have previously had a heart attack, and whether they have previously had a stroke. The participant’s systolic blood pressure was also measured (taken as the average of 3 measurements). Only participants without prior heart attack or stroke were included in the study.

Data collection took place at baseline (Year 0) and cardiovascular disease status was assessed 10 years later.

## Modeling Approach 

**Cox regression** - the same methodological approach used in deriving risk prediction equations in the Framingham Heart Study. It derives prediction functions based on data from a prospective study that collected risk factor information and cardiovascular disease status at baseline and then tracked the occurrence of cardiovascular disease events for at least 10 years. Participants with a prior heart attack or stroke were removed from analysis. 

We derived mathematical risk prediction equations in a manner similar to the cardiovascular disease prediction equations derived in the Framingham Heart Study (<https://framinghamheartstudy.org/fhs-risk-functions/cardiovascular-disease-10-year-risk/>). Specifically, we used sex-specific cox proportional hazard regression to relate risk factors (predictors) to the incidence of a first cardiovascular disease event.&#x20;

We derived two versions of each equation: one using the minimal set of risk factors above and another using the full set.

## Model Performance&#x20;

### Internal validation&#x20;

The area under the curve (AUC) for the full model was 0.72 (n=4500).&#x20;

### External validation&#x20;

The China-PAR equation is the gold standard for predicting risk in the Chinese population, but it requires the user to know their cholesterol values and thus it is not practical for most people.&#x20;

The correlation between the current model’s predictions (without cholesterol) and the ChinaPAR model's predictions (with cholesterol) was examined using an alternative Chinese dataset.&#x20;

The relation between these two values overall, in males only and in females only is depicted in Figure 4. The respective Pearson correlations are r=0.68, r=0.61 and r=0.74.

<figure><img src="/files/jC4ctzupqrOG1KbKnXzG" alt=""><figcaption></figcaption></figure>

*Figure 4: Validation against China-PAR equation; Scatter plot and Pearson r for: (a) Males and females, (b) Males only, and (c) Females only*

## Secondary Models: Heart attack and stroke&#x20;

Heart attack and stroke-specific models were derived in a similar manner.

Predictions are displayed to the user as a percentage likelihood of having diabetes. A percentage of 7.25% or higher suggests a risk of cardiovascular diseases, 2.39% or higher indicates a risk of heart attack, and 4.79% or higher suggests a risk of stroke.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.bitdoctor.ai/clinical-measurement-reports/cardiovascular-diseases-risk-incl.-heart-attack-and-stroke-risks.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.