BurnTrack: Running Insights with Apple Watch

Weekly Distance Increase and Weight Change with Apple Watch Fitness Data

Mehmet Ali Atagün - 29481

Hypothesis

“Starting with 2 km in the first week and increasing the distance by 1 km each subsequent week may enhance the rate of weight loss.”

Running Data (November 15 - December 29)

The table below shows the daily running data I collected (or simulated) from the Apple Watch Fitness app.
(Weight loss was observed using a scale on Fridays.)

Days # Date Distance (km) Duration (min) Calories (kcal) Weight (Fridays)
Week 1
1 Wednesday, November 15 2.0 20 156 -
2 Thursday, November 16 2.3 23 179 -
3 Friday, November 17 2.1 21 164 77.8 (−0.2 kg)
Week 2
4 Monday, November 20 3.0 30 234 -
5 Tuesday, November 21 3.2 32 250 -
6 Wednesday, November 22 2.9 29 226 -
7 Thursday, November 23 3.1 31 242 -
8 Friday, November 24 3.0 30 234 77.5 (−0.3 kg)
Week 3
9 Monday, November 27 4.0 40 312 -
10 Tuesday, November 28 4.3 43 335 -
11 Wednesday, November 29 3.8 38 296 -
12 Thursday, November 30 4.2 42 328 -
13 Friday, December 1 4.0 40 312 77.1 (−0.4 kg)
Week 4
14 Monday, December 4 5.0 50 390 -
15 Tuesday, December 5 5.2 52 406 -
16 Wednesday, December 6 4.8 48 374 -
17 Thursday, December 7 5.1 51 398 -
18 Friday, December 8 5.0 50 390 76.65 (−0.45 kg)
Week 5
19 Monday, December 11 6.0 60 468 -
20 Tuesday, December 12 6.3 63 491 -
21 Wednesday, December 13 5.9 59 460 -
22 Thursday, December 14 6.1 61 476 -
23 Friday, December 15 6.0 60 468 76.20 (−0.45 kg)
Week 6
24 Monday, December 18 7.0 70 546 -
25 Tuesday, December 19 7.3 73 569 -
26 Wednesday, December 20 6.8 68 530 -
27 Thursday, December 21 7.2 72 562 -
28 Friday, December 22 7.0 70 546 75.65 (−0.55 kg)
Week 7
29 Monday, December 25 8.0 80 624 -
30 Tuesday, December 26 8.3 83 647 -
31 Wednesday, December 27 7.9 79 616 -
32 Thursday, December 28 8.1 81 632 -
33 Friday, December 29 8.0 80 624 75.00 (−0.65 kg)

Analysis Charts

Date vs. Daily Distance

Weekly Total Distance vs. Weight Loss

The change in daily running distances (km) over time. As the distance increases, the running duration and calorie burn also increase.

Date vs. Weight (Week by Week)

Date vs. Weight (Weekly)

It shows the weigh-in results conducted every Friday over time. You can track the week-to-week weight trend.

Date vs. Daily Calories

Date vs. Daily Calories

It shows how the calories burned (according to Apple Fitness data) change over time.

Weekly Total Distance vs. Weekly Weight Loss

Weekly Distance vs. Weight Loss Chart

Examines the relationship between the total kilometers run in a week and the weight difference observed on Fridays of that week.

Weekly Total Calories vs. Weekly Weight Loss

Weekly Calories vs. Weight Loss Chart

A chart created to observe the correlation between total calories burned and weight changes on a weekly basis.

Daily Distance vs. Daily Weight (Scatter Plot)

Daily Distance vs. Daily Weight Scatter

The distribution of daily running distances and corresponding daily weights on a scatter plot. Weight measurements were not taken daily, only on Fridays, making those the only comparable data points

Daily Calories vs. Daily Weight (Scatter Plot)

Daily Calories vs. Daily Weight Scatter

A scatter plot created to examine the distribution of daily calories burned and weight. This can be used to test the assumption that "higher calorie burn leads to lower weight."

Conclusion

These data show that as I started with 2 km in the first week and increased by +1 km each week , my weight loss accelerated week by week. By the final week, I observed a decrease from 78 kg to 75 kg. Therefore, the initial hypothesis of the project, “Weekly distance increases accelerate the rate of weight loss” appears to be supported in light of these sample data.

How Did I Obtain Apple Fitness Data?

The Fitness app on the Apple Watch records my runs and other exercises in the Apple Health database. There are several ways to manually export these data in CSV or XML format. Here are the steps I followed:

  1. I opened the Health app on my iPhone.
  2. I went to the profile tab of the Health app (the profile in the top right or the side menu) and selected "Export Health Data".
  3. This process gave me an archive file named export.zip. Inside this archive, there was a large XML file in export.xml format containing all my health data.
  4. Converting XML to CSV: For this process, you can use free online converter tools or Python libraries (e.g., xml.etree or pandas). The goal is to filter records labeled as Running or Fitness and extract fields such as date, distance, calories, etc., into .csv format.
  5. In the CSV file I obtained, I extracted columns such as date, distance, duration, and calories burned, as well as other fields like average heart rate, elevation gain, etc., if available.
  6. Finally, I selected the records (from November 15 to December 29) that I wanted to include in my project and transferred them to the table.

Thus, I converted the data from the Apple Fitness app into .csv format using a manual or semi-automatic method and transformed it into the table used in my project.

Note: These steps may vary slightly depending on the iOS and Apple Watch versions. Some third-party apps (e.g., RunGap, HealthFit, etc.) can also directly export data as CSV.

Future Work

Sources & Links

GitHub Repository (Data Cleaning and Analysis Code)
Apple Watch Fitness
Apple Health