Your First Health Connect Integrated App

1. Introduction


What is Health Connect?

Health Connect is a health data platform for Android app developers. It provides a single, consolidated interface for access to users' health and fitness data, and consistent functional behavior across all devices. With Health Connect, users have a secure health and fitness data storage on-device, with full control and transparency over access.

How does Health Connect work?

Health Connect supports more than 50 common health and fitness data types and categories, including activity, sleep, nutrition, body measurements, and vitals like heart rate and blood pressure.

How Health Connect Works

With user permission, developers can securely read from and write to Health Connect using standardized schemas and API behavior. Users get full control over their privacy settings, with granular controls to see which apps are requesting access to data at any given time. The data in Health Connect is stored on-device and encrypted. Users also get the ability to shut off access or delete data they don't want on their device, and the option to prioritize one data source over another when using multiple apps.

Health Connect Architecture


The following is an explanation of the key aspects and architectural components of Health Connect:

  • Client app: To integrate with Health Connect, the client app links the SDK into their health and fitness app. This provides an API surface to interact with the Health Connect API.
  • Software Development Kit: The SDK enables the client app to communicate with the Health Connect APK.
  • Health Connect APK: This is the APK that implements Health Connect. It contains both its Permissions management and Data management components. The Health Connect APK is made available directly on the user's device, making Health Connect device-centric instead of account-centric.
  • Permissions management: Health Connect includes a user interface through which apps request the user's permission to display data. It also provides a list of existing user permissions. This allows users to manage the access they have granted or denied to various applications.
  • Data management: Health Connect provides a user interface with an overview of recorded data, whether it's a user's step counts, cycling speeds, heart rate, or other supported data types.

What you'll build

In this Codelab, you're going to build a simple health and fitness app integrated with Health Connect. Your app will do the following:

  • Obtain and check user permissions for data access.
  • Write data to Health Connect.
  • Read aggregated data from Health Connect.

What you'll learn

  • How to setup your environment to support Health Connect integration development.
  • How to obtain permissions and execute permission checks.
  • How to contribute health and fitness data to the Health Connect platform.
  • How to benefit from the on-device data storage.
  • How to validate your app with Google-provided developer tools.

What you'll need

  • Latest stable version of Android Studio.
  • Android mobile device with Android SDK version 28 (Pie) or higher.

2. Getting set up

Prepare the Health Connect app

The Health Connect app is responsible for handling all the requests that your application sends through the Health Connect SDK. These requests include storing data and managing its read and write access.

Access to Health Connect depends on the Android version installed on the phone. The following sections outline how to handle several recent versions of Android.

Android 14

Starting Android 14 (API Level 34), Health Connect is part of the Android framework. Because this version of Health Connect is a framework module, there's no setup necessary.

Android 13 and lower

On Android 13 (API Level 33) and lower versions, Health Connect is not part of the Android framework. Therefore, you need to install the Health Connect app from the Google Play Store. Scan the QR code to install Health Connect.


Get the sample code

The sample directory contains the start and finished code for this Codelab. In the Project view on Android Studio, you will find two modules:

  • start: The starter code for this project where you will make changes to this to complete the codelab.
  • finished: The completed code for this codelab where it is used to check your work.

Explore the ‘start' code

The Codelab sample app has basic UIs built using Jetpack Compose with the following screens:

  • WelcomeScreen: It is the app's landing page that shows different messages depending on the Health Connect availability whether it is installed, not installed, or not supported.
  • PrivacyPolicyScreen: It explains the app's use of permissions which is shown when the user clicks the Privacy policy link in the Health Connect permissions dialog.
  • InputReadingsScreen: It demonstrates reading and writing simple weight records.
  • ExerciseSessionScreen: It is where users insert and list out exercise sessions. When clicking on the record, it takes the user to ExerciseSessionDetailScreen to show more data associated with the session.
  • DifferentialChangesScreen: It demonstrates how to get a Changes token and get new changes from Health Connect.

HealthConnectManager stores all the functions that interact with Health Connect. In this Codelab, we will guide you step by step to complete the essential functionalities. The <!-- TODO: strings within the start build have corresponding sections in this Codelab where sample codes are provided for you to insert into the project.

Let's start by adding Health Connect to the project!

Add Health Connect client SDK

To start using the Health Connect SDK, you need to add a dependency to the build.gradle file. To find the latest version of Health Connect, check the Jetpack library release.

dependencies {
    // Add a dependency of Health Connect SDK
    implementation ""

Declare Health Connect visibility

To interact with Health Connect within the app, declare the Health Connect package name in the AndroidManifest.xml:

<!-- TODO: declare Health Connect visibility -->
   <package android:name="" />

Run the start project

Once they are all set, run the start project. At this point, you should see the welcome screen showing the text "Health Connect is installed on this device" and a menu drawer. We will add the functionalities to interact with Health Connect in the subsequent sections.

d54773774e4dc9f.png 462cd7b6cf553ad.png

3. Permission control

Health Connect recommends developers to restrict permission requests to the data types that are used in the app. Blanket permission requests reduce user confidence in the app and could lower user trust. If a permission is denied more than twice, your app gets locked out. As a result, permission requests will no longer appear.

For the purpose of this Codelab, we only need the following permissions:

  • Exercise Session
  • Heart Rate
  • Steps
  • Total Calories Burned
  • Weight

Declare permissions

Every data type your app reads or writes needs to be declared using a permission in AndroidManifest.xml. Starting from version alpha10, Health Connect uses the standard Android permission declaration format.

To declare the permissions for the data types required, use the <uses-permission> elements and assign their respective names with the permissions. Nest them within the <manifest> tags. For the full list of permissions and their corresponding data types, see List of data types.

<!-- TODO: declare Health Connect permissions -->
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>
  <uses-permission android:name=""/>

Declare an intent filter in AndroidManifest.xml to handle the intent that explains how your app uses those permissions. Your app needs to handle this intent and display a privacy policy describing how user data is used and handled. This intent is sent to the app once the user taps the Privacy policy link in the Health Connect permissions dialog.

<!-- TODO: Add intent filter to handle permission rationale intent -->
    <action android:name="" />

Now reopen the app to see the declared permissions. Click Settings from the menu drawer to get to the Health Connect Settings screen. Then, click App permissions where you should see Health Connect Codelab in the list. Click Health Connect Codelab to display a list of data types for read and write access on that app.

fbed69d871f92178.png 1b9c7764c1dbdfac.png

Request permissions

Apart from directly taking users to Health Connect Settings to manage permissions, you can also request permissions from your app via the Health Connect APIs. Note that users may change permissions at any time, so be sure your app checks if required permissions are available. In the Codelab project, we check and send permission requests before reading or writing data.

HealthConnectClient is an entry point to the Health Connect API. In HealthConnectManager.kt, obtain a HealthConnectClient instance.

private val healthConnectClient by lazy { HealthConnectClient.getOrCreate(context) }

To start the request permissions dialog within your application, build a set of permissions for the required data types first. You must request permissions to the data types that you only use.

For example, in the Record weight screen, you only need to grant read and write permissions for Weight. We have created a permission set in InputReadingsViewModel.kt as shown in the following code.

  val permissions = setOf(

Then, check if the permissions have been granted before launching the permission request. In HealthConnectManager.kt, use getGrantedPermissions to check if the permission of the required data types are granted. To launch the permission request, you have to create an ActivityResultContract by using PermissionController.createRequestPermissionResultContract(), which should be launched when the required permissions are not granted.

  suspend fun hasAllPermissions(permissions: Set<String>): Boolean {
    return healthConnectClient.permissionController.getGrantedPermissions().containsAll(permissions)

  fun requestPermissionsActivityContract(): ActivityResultContract<Set<String>, Set<String>> {
    return PermissionController.createRequestPermissionResultContract()

In the Codelab sample app, you may see the Request permissions button shown on the screen if you have not granted the permissions to the required data types. Click Request permissions to open the Health Connect permission dialog. Allow the required permissions and get back to the Codelab app.

626eedcec23659ce.png 6df6cf0e5c4a1a9e.png

4. Write data

Let's start writing records into Health Connect. To write a Weight record, create a WeightRecord object with the weight input value. Note that the Health Connect SDK supports various units classes. For example, use Mass.kilograms(weightInput) to set users weight in kilograms.

All data written to Health Connect should specify zone offset information. Specifying zone offset information while writing data provides time zone information when reading data in Health Connect.

After creating the weight record, use healthConnectClient.insertRecords to write the data into Health Connect.

* TODO: Writes [WeightRecord] to Health Connect.
suspend fun writeWeightInput(weightInput: Double) {
   val time =
   val weightRecord = WeightRecord(
       weight = Mass.kilograms(weightInput),
       time = time.toInstant(),
       zoneOffset = time.offset
   val records = listOf(weightRecord)
   try {
      Toast.makeText(context, "Successfully insert records", Toast.LENGTH_SHORT).show()
   } catch (e: Exception) {
      Toast.makeText(context, e.message.toString(), Toast.LENGTH_SHORT).show()

Now let's run the app. Click Record weight and enter a new weight record in kilograms. To verify if the weight record is successfully written to Health Connect, open the Health Connect app in Settings and go to Data and access > Body measurements > Weight > See all entries. You should see the new weight record written from Health Connect Codelab.

Write exercise sessions

A session is a time interval during which a user performs an activity. An exercise session in Health Connect can include anything from running to badminton. Sessions allow users to measure time-based performance. This data records an array of instantaneous samples measured over a period of time such as a continuous heart rate or location samples during an activity.

The following example shows you how to write an exercise session. Use healthConnectClient.insertRecords to insert multiple data records associated with a session. The insertion request in this example includes ExerciseSessionRecord with ExerciseType, StepsRecord with steps count, TotalCaloriesBurnedRecord with Energy and series of HeartRateRecord samples.

   * TODO: Writes an [ExerciseSessionRecord] to Health Connect.
  suspend fun writeExerciseSession(start: ZonedDateTime, end: ZonedDateTime) {
          startTime = start.toInstant(),
          startZoneOffset = start.offset,
          endTime = end.toInstant(),
          endZoneOffset = end.offset,
          exerciseType = ExerciseSessionRecord.EXERCISE_TYPE_RUNNING,
          title = "My Run #${Random.nextInt(0, 60)}"
          startTime = start.toInstant(),
          startZoneOffset = start.offset,
          endTime = end.toInstant(),
          endZoneOffset = end.offset,
          count = (1000 + 1000 * Random.nextInt(3)).toLong()
          startTime = start.toInstant(),
          startZoneOffset = start.offset,
          endTime = end.toInstant(),
          endZoneOffset = end.offset,
          energy = Energy.calories((140 + Random.nextInt(20)) * 0.01)
      ) + buildHeartRateSeries(start, end)

   * TODO: Build [HeartRateRecord].
  private fun buildHeartRateSeries(
    sessionStartTime: ZonedDateTime,
    sessionEndTime: ZonedDateTime,
  ): HeartRateRecord {
    val samples = mutableListOf<HeartRateRecord.Sample>()
    var time = sessionStartTime
    while (time.isBefore(sessionEndTime)) {
          time = time.toInstant(),
          beatsPerMinute = (80 + Random.nextInt(80)).toLong()
      time = time.plusSeconds(30)
    return HeartRateRecord(
      startTime = sessionStartTime.toInstant(),
      startZoneOffset = sessionStartTime.offset,
      endTime = sessionEndTime.toInstant(),
      endZoneOffset = sessionEndTime.offset,
      samples = samples

5. Health Connect Toolbox

Introduction to the Health Connect Toolbox

The Health Connect Toolbox is a companion developer tool to help you test your app's integration with Health Connect. It can read and write data directly to Health Connect, allowing you to test your app's CRUD operations.

In this Codelab, we are going to use the Health Connect Toolbox to test the reading and writing functionalities you just implemented.

Set up Health Connect Toolbox

Extract the ZIP folder to get the APK files. Then to install the Toolbox APK on a connected device, use adb. Traverse to the folder where the APK is located and run the following command:

$ adb install HealthConnectToolbox-{Version Number}.apk

The first time you open the Health Connect Toolbox app, you're taken to the permission settings under Apps > Special app access > Display over other apps. This permission allows Health Connect Toolbox to show an overlay on top of other apps so that you can test reading and writing data without leaving the app you're developing.

To manage read and write permissions for testing, you can open the Health Connect app from either the main screen of the Toolbox app or directly go to the permission flow.


Read and Write Health Records

Health Connect Toolbox supports reading and writing all Health Connect data types. In the last session of the Codelab, you successfully wrote Weight and Exercise Session records to Health Connect. Let's check if you can read the data from the Health Connect Toolbox.

Before reading and writing to Health Connect, permission from the user must be obtained. The same is true for the Health Connect Toolbox. First, accept the permission request from Health Connect Toolbox. Next, click the search icon 1f407c55884bb8c3.png from the overlay menu to open the dialog, select the data type such as Weight, and click READ HEALTH RECORD. You should see the records from the Codelab sample app that you just wrote to Health Connect.

To insert a record into Health Connect, click on the edit icon 10c524823c596aea.png from the overlay menu to open the dialog. Then, select the data type. Let's insert a Weight record from the Toolbox. In the next session, we will show you how to read the record via Health Connect API and display the data in your app.


6. Read data

Now that you have written Weight and Exercise Session records with both the Codelab sample app and the Toolbox app, let's use the Health Connect API to read those records. First, create a ReadRecordsRequest and specify the record type and the time range to read from. The ReadRecordsRequest can also set a dataOriginFilter to specify the source app of the record you want to read from.

     * TODO: Reads in existing [WeightRecord]s.
    suspend fun readWeightInputs(start: Instant, end: Instant): List<WeightRecord> {
        val request = ReadRecordsRequest(
            recordType = WeightRecord::class,
            timeRangeFilter = TimeRangeFilter.between(start, end)
        val response = healthConnectClient.readRecords(request)
        return response.records
   * TODO: Obtains a list of [ExerciseSessionRecord]s in a specified time frame.
  suspend fun readExerciseSessions(start: Instant, end: Instant): List<ExerciseSessionRecord> {
    val request = ReadRecordsRequest(
      recordType = ExerciseSessionRecord::class,
      timeRangeFilter = TimeRangeFilter.between(start, end)
    val response = healthConnectClient.readRecords(request)
    return response.records

Now let's run the app and check if you can see a list of weight records and exercise sessions.

a08af54eef6bc832.png 3b0781389f1094a1.png

7. Read differential data

The Health Connect Differential Changes API helps to track changes from a specific point of time for a set of data types. For example, you want to know if users have updated or deleted any existing records outside your app so that you can update your database accordingly.

Reading data with Health Connect is restricted to applications running in the foreground. This restriction is in place to further strengthen user privacy. It notifies and assures users that Health Connect does not have background read access to their data and that data is only read and accessed in the foreground. When the app is in the foreground, the Differential Changes API enables developers to retrieve changes made to Health Connect by deploying a Changes token.

In HealthConnectManager.kt there are two functions which are getChangesToken() and getChanges(). We will add Differential Changes APIs to these functions to get data changes.

Initial Changes token setup

The data changes are retrieved from Health Connect only when your app requests them with a Changes token. The Changes token represents the point in commit history from which differential data will be taken.

To get a Changes token, send a ChangesTokenRequest with a set of data types you want to track the data changes. Keep the token and use it when you want to retrieve any updates from Health Connect.

   * TODO: Obtains a Changes token for the specified record types.
  suspend fun getChangesToken(): String {
    return healthConnectClient.getChangesToken(

Data update with Changes token

When you want to get changes from the last time your app synced with Health Connect, use the Changes token you got earlier and send a getChanges call with the token. The ChangesResponse returns a list of observed changes from Health Connect such as UpsertionChange and DeletionChange.

   * TODO: Retrieve changes from a Changes token.
  suspend fun getChanges(token: String): Flow<ChangesMessage> = flow {
    var nextChangesToken = token
    do {
      val response = healthConnectClient.getChanges(nextChangesToken)
      if (response.changesTokenExpired) {
        throw IOException("Changes token has expired")
      nextChangesToken = response.nextChangesToken
    } while (response.hasMore)

Now let's run the app and go to the Changes screen. First, enable Track changes to get a Changes token. Then insert weight or exercise sessions either from the Toolbox or from the Codelab app. Go back to the Changes screen and select Get new changes. You should see the upsertion changes now.

f3aded8ae5487e9c.png 437d69e3e000ce81.png

8. Aggregate data

Health Connect also provides aggregated data through aggregate APIs. The following examples show you how to get cumulative and statistical data from Health Connect.

Use healthConnectClient.aggregate to send AggregateRequest. In the aggregate request, specify a set of aggregate metrics and the time range you want to get. For example, ExerciseSessionRecord.EXERCISE_DURATION_TOTAL and StepsRecord.COUNT_TOTAL provide cumulative data, while WeightRecord.WEIGHT_AVG, HeartRateRecord.BPM_MAX, and HeartRateRecord.BPM_MIN provide statistical data.

     * TODO: Returns the weekly average of [WeightRecord]s.
    suspend fun computeWeeklyAverage(start: Instant, end: Instant): Mass? {
        val request = AggregateRequest(
            metrics = setOf(WeightRecord.WEIGHT_AVG),
            timeRangeFilter = TimeRangeFilter.between(start, end)
        val response = healthConnectClient.aggregate(request)
        return response[WeightRecord.WEIGHT_AVG]

This example shows how to get associated aggregation data for a specific exercise session. First, read a record using healthConnectClient.readRecord with a uid. Then, use the startTime and the endTime of the exercise session as time range and the dataOrigin as filters to read associated aggregations.

   * TODO: Reads aggregated data and raw data for selected data types, for a given [ExerciseSessionRecord].
  suspend fun readAssociatedSessionData(
      uid: String,
  ): ExerciseSessionData {
    val exerciseSession = healthConnectClient.readRecord(ExerciseSessionRecord::class, uid)
    // Use the start time and end time from the session, for reading raw and aggregate data.
    val timeRangeFilter = TimeRangeFilter.between(
      startTime = exerciseSession.record.startTime,
      endTime = exerciseSession.record.endTime
    val aggregateDataTypes = setOf(
    // Limit the data read to just the application that wrote the session. This may or may not
    // be desirable depending on the use case: In some cases, it may be useful to combine with
    // data written by other apps.
    val dataOriginFilter = setOf(exerciseSession.record.metadata.dataOrigin)
    val aggregateRequest = AggregateRequest(
      metrics = aggregateDataTypes,
      timeRangeFilter = timeRangeFilter,
      dataOriginFilter = dataOriginFilter
    val aggregateData = healthConnectClient.aggregate(aggregateRequest)
    val heartRateData = readData<HeartRateRecord>(timeRangeFilter, dataOriginFilter)

    return ExerciseSessionData(
      uid = uid,
      totalActiveTime = aggregateData[ExerciseSessionRecord.EXERCISE_DURATION_TOTAL],
      totalSteps = aggregateData[StepsRecord.COUNT_TOTAL],
      totalEnergyBurned = aggregateData[TotalCaloriesBurnedRecord.ENERGY_TOTAL],
      minHeartRate = aggregateData[HeartRateRecord.BPM_MIN],
      maxHeartRate = aggregateData[HeartRateRecord.BPM_MAX],
      avgHeartRate = aggregateData[HeartRateRecord.BPM_AVG],
      heartRateSeries = heartRateData,

Now let's run the app and check if you can see the average weight on the Record weight screen. You can also see the detailed data of an exercise session by opening the Exercise sessions screen and choosing on one of the exercise session records.


9. Congratulations

Congratulations, you've successfully built your first Health Connect integrated health and fitness app.

The app can declare permissions and request user permissions on data types used in the app. It can also read and write data from the Health Connect data storage. You've also learned how to use the Health Connect Toolbox to support your app development by creating mock data in the Health Connect data storage.

You now know the key steps required to make your health and fitness app a part of the Health Connect ecosystem.

Further reading