I’m interested in understanding the benefits of what I call “micro-tracking” for their health: tracking information such as diet, heart rate, exercise routine, etc. at a granularity finer than a day.
Starting last year, and likely expanding further into this year, I am using a more “macro-tracking” approach.
For instance, for supplements, this macro-tracking simply involves tracking the start and end date of consumption of each supplement bottle. This means roughly one new data entry per month. The corresponding “micro-tracking” approach would be to record each time I take a supplement, possibly with other information such as the time of day, relation with meal, etc.
Similarly, for food, I do record all purchases of restaurant food in my activity tracker (albeit not in a computable format). I am now thinking of adding information on the specifics of all food purchases (from both grocery stores and restaurants) in a computable format. This is a macro-tracking approach. The corresponding micro-tracking approach would mean recording each meal, including information such as time of day, quantity of various foods in the meal, etc. (For the most part I do not share food with others, and my food waste is near-zero, so purchases = consumption; I can record exceptions separately).
Disadvantages I see of micro-tracking:
Time: It looks like micro-tracking adds a nontrivial daily overhead of tracking work. Time is very important to me. Even more so, time that I need to spend daily, when I might be pressed for time on other tasks, is even more important.
Difficulty with micro-quantification (for food): It’s an extra hassle to quantify exactly how much food I am consuming. When consuming cooked food, I am usually helping myself from a large bowl of prepared food and may go from 70% to 65% of it or something like that. The food isn’t perfectly mixed either, so some days I might end up having more of the tomatoes part and other days I might end up having more of the spinach part.
Aggregation effort: Once all the micro-quantities are entered, I need to aggregate them to extract meaningful data. If I enter data at a more aggregated level, this is somewhat easier.
Advantages I see of micro-tracking, and why these did not ultimately convince me:
-
Better correlational analysis of day-to-day fluctuations: If I were micro-tracking my moods, physiological measurements, and physical reactions as well as my food and supplement intake, I might be able to identify what patterns of food or supplement intake correlate with what moods. People I know who micro-track tend to have reasons like this.
This is a pretty good reason for some people. People who suffer from allergic reactions, stomach issues, or large mood swings could probably benefit from such diagnostic data very concretely. In my case, I haven’t had major issues of this sort frequently; the few rare times I do have such issues, I manually record in my notes folder along with details specific to the situation. I also don’t trust my self-reflection to assign quantifiable and consistent measures of things like my mood.
-
Having micro-tracking permits some sort of aggregations that wouldn’t be possible just with macro-tracking, such as information on time of day, time gaps, etc.: For instance, rather than just know how much of a Vitamin D supplement I took in the last three years, I can learn how much of it I took in the mornings versus at night, and what the day-to-day fluctuation in intake was. Possibly, such details matter a lot for assessing the impact of the supplement.
I agree that this is a potential benefit; however, as of now I am not collecting enough fine-grained data on the other side to meaningfully correlate with. It seems to me that combining total consumption with some general information on when I usually consume supplements is enough.
I’m curious to hear what people think I’m missing, or any other insights people want to share!
UPDATE 2024-11-21: These are the practices I settled on:
-
Starting March 2021 (a little over a month after this post), I have been recording all my food purchases here. I also entered nutritional information for most of the foods I purchased, which allows me to do crude calculation of nutrient consumption.
-
Starting June 2024, I have been recording something that’s intermediate between food purchase and consumption: the preparation or opening of food. For a raw vegetable or something like rice or lentils, this corresponds to when I add it to a cooked meal; for a packet or bottle, this corresponds to when I open it to start consumption. The information can be found here.
This has proved to be a better proxy than purchase tracking (though I’m still doing purchase tracking as well). The main advantage is that preparation or opening is more granular and closer to the time of actual consumption. This became particularly relevant for foods that I purchase in larger quantities, like rice, lentils, and yogurt.
-
I’ve also written a bunch of verification queries for both purchases and preparations and openings. These queries run automatically as part of the make commands I run to reload data after any data entry. This allows me to catch cases where my nutritional profile and food choices are deviating meaningfully from what I expect; in some cases, this information leads me to take action, whereas in others, it’s an expected consequence of situational factors I am already aware of.
-
I’m thinking of eventually making an interface for easier historical comparison of food consumption, but the verification queries (see preceding point) are good enough for now so I may not get around to making the interface in the near future.
It has been unambiguously helpful for my Apple Watch to inform me that my sleep quality is detectably higher when I exercise, even if that exercise is just a brisk 1-2 mile walk. I generally agree subjectively feel better when the watch tells me I’ve slept well. Connecting “go for your walk” to “feel noticeably better tomorrow” is much more motivating than going for a walk due to nebulous long-term health reasons. None of this would happen if the watch wasn’t automatically tracking my sleep (including interruptions and sleeping heart rate), and my daily activities.
I spent a lot of time thinking about self tracking and speaking with other people in the Quantified Self community. My conclusion is that for most people microtracking only makes sense when it’s focused on solving specific issues that the person cares about.
My approach to solving productivity/motivation/energy problems is to focus on variables with the greatest weight. The big three are diet, sleep, and exercise; so if I’m experiencing slowdowns the first thing I do is make sure those three variables are in good shape. Then I work down the list of variables with descending weight: stress, loneliness, boredom, etc., your variables may vary. Sure there is a difference between taking 1000mg and 2500mg of Vitamin C a day; but it’s not statistically relevant (unless you have some serious health issues, like scurvy).
Consider that you may need to spend not just a lot of time actually entering and analyzing data, but also on refining your model for what data to collect and how to analyze it, considering what if anything to do with it, researching what health metrics actually matter for your well-being, questioning whether this project is worth your time, addressing failures to implement it consistently, fielding questions about it from others, considering alternative projects to achieve similar results, and thinking about it when you’re not even using it.
I use Bitesnap and MFP to avoid most of your problems with diet tracking. Measuring exact weight of each ingredient in something I cook is still a hassle. For heart rate I recommend uECG. Many tools are being developed to track exercise in great depth such as mbientlabs’ wearable accelerometry and computer vision pose estimation.
Even if user gets good time saving equipment the daily time expenditure is still non trivial. The benefits could however be great! The current biggest problem is that no automated analysis software yet exists. For more see the Kialo debate:
https://www.kialo.com/everyone-should-health-track—self-quantify-49787
I’ve found micro tracking of my time over a week long period really beneficial. I don’t think the same would be true for diet as there is so many variables so look at, picking any one thing to look at would be privileging that hypothesis to an unreasonable degree. That said if there is something specific you are looking for, where you would expect short term fluctuations to be important, then I could see the value.
I suppose I would also see the value if you were running a trial on yourself. Toss a coin at the start of the week if heads each X every day, if tails then don’t. Measure something you expect to change over such a short period of time (e.g. sleep, subjective energy levels)