Apple Health

Collected data

• Your activity

  activity

  Readings such as steps; walking, running, cycling, swimming, wheelchair, rowing, paddling, skating and snow-sports distances; flights climbed; active and resting energy; exercise, move and stand minutes and stand hours; running and cycling speed, power, cadence and form; workout effort; VO2 max, and notifications of low cardio fitness. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your body measurements

  body_measurements

  Height, weight, body mass index, lean body mass, body fat percentage, waist circumference, and the wrist temperature your Apple Watch measures while you sleep, which Apple also uses to estimate your menstrual cycle. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your reproductive health

  reproductive_health

  Cycle tracking and reproductive health: menstrual flow and spotting, cycle irregularity notifications, cervical mucus, basal body temperature, ovulation, progesterone and pregnancy test results, sexual activity, contraceptive use, pregnancy, lactation, menopause, and bleeding after menopause or during or after pregnancy. It also includes three reproduction-related symptoms, breast pain, pelvic pain and vaginal dryness, which are kept here rather than with your other symptoms so that they are shared only if you share this. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your hearing and sound exposure

  hearing

  Environmental sound levels, headphone audio levels, the sound reduction your headphones applied, and notifications of loud environments and loud headphone audio. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your vital signs

  vital_signs

  Heart rate, resting and walking heart rate, heart rate variability, heart rate recovery, AFib history, blood oxygen, body temperature, blood pressure and respiratory rate, and notifications of a high, low or irregular heart rhythm and of possible hypertension. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your lab and test results

  lab_results

  Blood glucose, insulin delivery, inhaler use, lung function (forced vital capacity, FEV1 and peak expiratory flow), electrodermal activity, peripheral perfusion index, and the number of times you fell. Health records that a doctor, clinic or hospital shared with Health are not included. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your sleep

  sleep

  Sleep stages and time in bed, breathing disturbances while you sleep, sleep apnea notifications, and your sleep duration goal. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your mindful minutes

  mindfulness

  Your mindful minutes: when each mindfulness session began and ended. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your nutrition

  nutrition

  What you logged eating and drinking: energy, carbohydrates, fat, protein, sugar and fibre, vitamins and minerals, caffeine and water. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your alcohol consumption

  alcohol_consumption

  Blood alcohol content and the number of alcoholic drinks you logged. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your mobility

  mobility

  Walking steadiness and its notifications, walking speed, step length, walking asymmetry and double support time, stair ascent and descent speed, and six-minute walk distance. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your symptoms

  symptoms

  Symptoms you logged, such as headache, fatigue, fever, coughing, nausea, dizziness, mood changes, sleep changes, skin and hair changes and bladder incontinence. Breast pain, pelvic pain and vaginal dryness are in Reproductive health instead. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your other Health data

  other

  Everything that belongs to none of the other areas: UV exposure, time in daylight, underwater depth, water temperature, toothbrushing and handwashing, and any kind of Health data this version does not recognise yet, such as one Apple adds in a later iOS release, whatever it measures. Nothing is dropped for having an unfamiliar type, and the import receipt says how many such readings arrived. Each reading carries its type, value and unit as the export states them; its start, end and creation dates; the time-zone offset where it was recorded, which shows when you were in another time zone; and whether it was typed in by hand. The names of the apps and devices that recorded it are not included, nor anything that identifies a particular device, so no reading carries the name someone gave their phone or watch; the device’s maker, model, and hardware, firmware and software versions are kept, because a reader needs to know whether a measurement came from a watch or a chest strap. Readings are told apart by their type and their start, end and creation dates only: two readings of the same type with the same start, end and creation dates arrive once, whatever their values, units or devices, and the import receipt counts the second as a duplicate. On a very large export this stops once 400,000 different readings across all the health areas have been remembered, and the import receipt of each area affected says so. Every import brings in everything in your export again: a reading you already have is updated rather than duplicated, and a reading you delete in Health is not removed here.

• Your workouts

  workouts

  Workout type, with duration in minutes, energy in kilocalories and distance in kilometres, each converted from whatever unit the export stated, plus start and end dates, the time-zone offset where it began, which shows when you were in another time zone, and up to 500 pause or segment markers and up to 500 per-metric statistics per workout, which summarise readings taken during it, such as its average, minimum and maximum heart rate; a workout with more of either has that list cut at 500, and the import receipt names which list under fields_unavailable. On iPhones running iOS 16 or later, a workout can also carry separate detail for each activity within it; that per-activity detail is not included, and the activities‘ statistics are used only to add up the workout’s totals where the workout’s own do not. Such exports typically state no total distance or energy for a workout, so those are added up from the workout’s own statistics, or from its activities’ where those record a kind the workout’s own lack, never from both; where each records a kind of distance the other lacks, the total is left empty and the import receipt says so. A quantity that is not a number, or whose unit could not be converted, is left empty rather than guessed, and the import receipt says so. Application and source names are excluded, as is the workout’s weather block and any GPS route, because weather on a known date reconstructs location. The device’s maker, model, and hardware, firmware and software versions are kept; its name and identifiers are not. Workouts are told apart by their type, start and end only: two of the same type with the same start and end arrive once, whatever their duration, distance, energy or device, and the import receipt counts the second as a duplicate; on a very large export this stops once 100,000 different workouts have been remembered, and the workouts import receipt says so. Every import brings in everything in your export again, and a workout you already have is updated rather than duplicated; a workout you delete in Health is not removed here.

• What was imported

  coverage_diagnostics

  A short receipt for each kind of data in each import: whether everything in your export was read, how many records arrived and the earliest and latest dates among them, how many duplicates were discarded, how many entries could not be read or were of a kind this version does not recognise, the names of any fields left empty, and the date your export was taken. The receipts alone therefore show, for each kind of data, whether you have any, how many, and over what dates: for example, whether your export holds reproductive-health records, how many, and from when to when.

Available in

Confirmed against the published Vana releases below. An app can include an older connector than the latest upstream release.

Build only. This connector is published upstream but is not enabled in the Vana releases checked for this directory.