Tracking Nutrition with Cronometer and WHOOP
My two-month journey to unlocking 3 + 1 key changes for a more balanced diet
Introduction
For over 15 years, I have measured various things about my life, including workouts, sleep, and biometrics in general with an ever-evolving stack of apps and wearables that I call The Quantifying Stack.
As I aim to keep the stack lean, effortless, and affordable, there’s especially one important area of life that doesn’t easily fill the criteria: nutrition. There’s just no way around it, no sensor to measure what goes into our mouths. At a minimum, we have to take a picture of everything we eat and outsource the analysis to someone else. Oftentimes, the picture alone is not enough because many things are difficult to identify correctly based on a picture alone, so we have to add descriptions. And of course, we have to pay for the analysis — a lot.
Or we can take a hit in effortlessness and do it ourselves, though only for a while, just enough to gain insight into what is going well and what should change.
Enter Cronometer
On this notion, I decided to try out Cronometer, which is a nutrition-tracking app that focuses on accuracy. To me, the most intriguing point was this:
Your picture of overall health and wellness includes a lot more than calories, proteins, fats and carbs. Yes, of course we give you this info but we also toss in a lot more! Cronometer gives you detailed information on 82 micronutrients as well so you can keep an eye on all the vitamins and minerals essential for living a healthy, energetic life.
The fact that micronutrients are supported so extensively differs from my past experiences with different apps, which was one reason to choose Cronometer this time.
Add WHOOP
The other one is the direct integration with WHOOP, a wearable I’ve been wearing now for over a year and really like.
Regarding this experiment, the most important feature of WHOOP is the journal, which can be used to track various behaviors including calories, macros (proteins, carbohydrates, fats), and doses of dairy. These entries are then linked with measured biometrics, including Resting Heart Rate (RHR) and Heart Rate Variability (HRV), to reveal correlations between the logged behaviors and recovery scores derived from the measured biometrics.
Naturally, all the exercise information from WHOOP also plays a role in the burned calories estimation.
Note: I do not have an affiliation with either Cronometer or WHOOP, meaning they have not asked me to write this post nor do they know I’m writing it. All information contained here reflects only my personal experiences, good or bad, with these products as accurately as I could convey.
With that said, if you want to get into WHOOP, you can use my personal referral link to get one month free for you and me: https://join.whoop.com/6AA19A.
Getting Started
The integration between the apps was easy to activate.
After the Cronometer app was installed, I created an account and activated the integration from the menu, under Connect Apps & Devices / Devices / WHOOP by logging into the WHOOP account and switching every category on.
Armed with these tools, my main goal was to find answers to the following questions:
- Are there any nutritional imbalances I should correct?
- If so, how?
Disclaimer: Please note that I am a software engineer and not a nutrition expert. Therefore, any interpretations and planned actions I mention below may not be accurate and should not be considered recommendations of any kind. It’s important to keep in mind that these are only my personal results, which will differ for each individual. The purpose of sharing this information is to highlight the possibilities of using nutrition-tracking tools to improve one’s diet and well-being.
So What Did I Find Out?
My situation is not alarming, but there are things I could adjust.
As around half of my meals are vegetarian or even vegan, I suspected that my protein intake may be too low and it seems to be true: on average only 68% of the target was reached.
On the other hand, even though I think I’m not eating that much junk, the average fat intake still soars at 134%.
Thirdly and maybe most interestingly, my micronutrient levels are generally pretty good but there are some deficiencies, especially regarding antioxidants and vitamins.
Into The Details
At this point, I should elaborate that most of the following targets are in their default values in Cronometer:
- BMR is based on Mifflin St. Jeor Equation
- The macro ratio is as follows: 25% Protein, 45% Net Carbs (Total Carbs-Fiber-Sugar Alcohols), 30% Fat
- All micronutrient targets are at their defaults.
Also, as my stack includes WHOOP and I have also a slight weight loss goal in mind, I fixed the Activity Level and Energy Target as follows:
- Activity Level: Lightly Active + WHOOP exercises on top (as I’m not tracking e.g. commutes by bus as exercises)
- Energy Target: 100% goal has a deficit of 125 kcal compared to the maintenance value.
Furthermore, all of the Cronometer values and diagrams presented below are analyzing the period from May 22 to Jul 21, 2023.
Additionally, I used the paid Gold version of the Cronometer to get the maximum benefit and diagrams out of the app during this time.
Without any further ado, let’s look at some of the numbers more closely.
Energy Summary and Macronutrients
From here we can notice that I seem to have consumed nearly a perfect amount of total calories with a slight deficit even compared to the goal with the defined deficit. This is pretty much in line with my weight going down by around 1 kg during this time, which may indicate that the estimations of both consumed as well as burned calories are pretty close to the truth — or that many wrongs make it right and I have just been lucky.
To me, these estimations and ratios seem pretty solid, so I’m going to come up with some adjustments to achieve a greater balance regarding macros.
Micronutrients
The following information about micronutrients is not going to be as accurate, because not every logged item has all the micronutrient data available. Specifically, for this exact set of data, Cronometer states that
Based on the data-quality of the 1672 items logged we have a confidence score of 70.2%.
Missing data is especially true for all the custom foods I have added because the nutrition labels here in Finland are pretty scarce regarding micronutrients. However, I still think the results are at least directional if nothing else.
Let’s look at the big picture of micronutrients.
Here we can see that in general the values are pretty good and the lowest scores are in the categories of Antioxidants, Bone Health, Immune Support, and Vitamins. Upon closer inspection, the low values of Bone Health and Immune Support are mostly due to some low vitamin levels combined with low vitamin D levels, which I do not consider a big problem during sunny summertime, because we get vitamin D also from the sun through biosynthesis.
Consequently, let’s focus on the Vitamins and Antioxidants categories.
Vitamins
Based on this, I should try to increase the intake of especially folate and vitamins A, E, and K. Maybe I should also resume taking a vitamin D supplement as the fall is approaching.
Do note also the lower confidence score of 65,5%.
Antioxidants
Here we discover the biggest deficiencies yet, regarding beta-carotene, lutein+zeaxanthin, and lycopene. Similarly, the accuracy here is even lower than before, 64,4% to be exact.
Despite the lower accuracy of these sections, I believe I could benefit from increasing the levels of some of these micronutrients either by incorporating some new foods into my diet or by adding suitable supplements into the mix.
Bonus: Magnesium
I’ve been taking a magnesium supplement (375 mg) about a third of the days, as I have been inspecting its effects on my sleep and recovery with the help of the WHOOP journal. In addition to somewhat improving my recovery scores and sleep, it seems the supplementation has not gone too far and I should keep doing it, maybe even more regularly but with smaller doses than before.
The Tracking Process
Before I get into the part of turning these insights into concrete actions, I want to say some things about the tracking process itself.
Generally, the Cronometer app worked well during these two months and only ‘crashed’ on rare occasions caused by network or server problems. The food selection also seems to be extensive, especially for foods available in North America, but quite good also regarding the foods of Northern Europe.
However, I did also add quite a few new items during these 2 months.
Adding Missing Foods
Adding missing items was straightforward enough, even though the “automatic nutrition label reader” didn’t work most of the time, probably due to European nutrition labels and the non-English language used in them. Many of the foods I added were later incorporated also into the official Cronometer database, as I opted-in for receiving an email notification upon this happening.
However, from the micronutrient accuracy point of view, the devil lies in all the missing food items I added, because this information is almost non-existent in the Finnish nutrition labels. So if the tracking focus is more on the micro side, for fairly generic items e.g. sausages, it could actually be more beneficial to find another corresponding sausage item that contains all the micronutrient information instead of adding a new one, even though the values would not be precise.
Regardless, for more special local foods that have no counterpart in the database, the micronutrient data is going to be inevitably lost, which reduces the reliability of the entire analysis to some extent.
Guesstimation
Another thing to reduce the reliability of the results is the so-called “guesstimation” method I mostly used.
As I aim to keep self-quantification practical, there’s no way I’d carry a food scale with me wherever I go — not even temporarily. Furthermore, I wouldn’t cease going to restaurants either, because then the results would not reflect my actual everyday nutrition. Thus, most of the time I was estimating (or guessing) the amounts of food I consumed.
To minimize the error in my estimations, I tried to recalibrate my spider senses every once in a while by first guessing the weight of a certain amount of specific food and then actually measuring it. Most of the time I was pretty close, sometimes only 2 grams away. However, there’s no denying that this method certainly adds a layer of inaccuracy to the process in favor of practicality.
Furthermore, regarding restaurants, I always included some additional oil (5–10g) in the greasy meals I logged to account for the fact that they often use a lot of oil in cooking.
Off Days
I want to be honest with you: there were some days I didn’t log everything I ate or drank! These were mostly special days that involved social gatherings and maybe even some alcohol; again, practicality over perfection.
However, there were not many of these days and one great feature of Cronometer is the possibility to mark fully logged days as complete and include only these days in the charts. This is also what I did, thus the off days are not skewing the results in any direction, which is nice.
Turning Data Into Action
Now that we have gone through the relevant data, it’s time to form some insights and turn them into concrete actions that I will try to incorporate into my daily routine. To increase my chances of succeeding, I will try to change only three things at this time.
Increasing Protein Intake
The benefits of having enough protein in a diet are often emphasized especially when talking about resistance training or losing weight.
Generally speaking though, the recommended minimum amount of protein is only 0.8 grams per kilogram, which I seem to have achieved easily. However, for more active individuals, the recommended amount is something between 1.2 to 1.8 grams per kilogram. This translates to a value between 107 and 160 grams at my current weight, which means that I could likely benefit from increasing my protein intake a bit, albeit the deficit is necessarily not as big as Cronometer indicates with the default targets.
In addition to my Cronometer data suggesting I get too little protein, my WHOOP data indicates that doses of dairy may be moderately beneficial for my recovery. Therefore, it seems like a good idea to add some milk curd to my diet. Additionally, as my vitamin and fiber intakes were also somewhat lacking, I could mix berries into the curd, killing many birds with one stone.
Therefore,
I will start eating breakfasts more regularly by making smoothies containing milk curd and berries my go-to breakfast. To vary things a bit, boiled eggs and omelets shall also be included in the breakfast menu. Additionally, protein curd will be the top snack choice during the day.
Reducing Fat Intake
This one is a slimy bastard that’s not easy to tame completely but maybe anything is better than nothing.
First of all, I don’t want to give up the occasional pizza day or other craving days for that matter. I’ve learned in the past that eliminating these small joys of life completely is not the best method for me in the long run, provided that these days don’t become too frequent.
What I could do instead is to replace some of the typical meals, which include things like meatballs and kinds of pasta, with salads containing enough lean protein.
Thus,
I will replace 2 warm meals per week with salad including lean protein.
Fixing Micronutrient Deficiencies
As stated before, the most significant deficiencies I had were beta-carotene, lutein+zeaxanthin, and lycopene. Regarding the first two, Cronometer suggests that leafy greens like spinach, kale, and lettuce are good sources. Additionally, carrot is another excellent source of beta-carotene. Lycopene sources include tomatoes, tomato sauce, watermelon, and guava.
Based on these,
I will increase my intake of lettuce, tomatoes, carrots, and watermelon because these are all available at the salad table of most lunch places I go to. Additionally, I will use more spinach, e.g. with breakfast omelets.
As a low-hanging bonus fruit, I will include a multivitamin supplement that targets especially vitamins A, D, E, and K as well as folate into my diet.
Calories, Macros, and WHOOP Recoveries
Although it was not the main focus of this stint, I also wanted to see how different amounts of calories and macros correlate with my WHOOP recovery scores and the results are somewhat counterintuitive.
The first thing we notice is that in general all of these behaviors have a correlation of only 0–2% with my recovery, which is classified as negligible. This means that whether I’ve logged these behaviors or not doesn’t have any statistically meaningful correlation with my recovery. This is also the way it should be, given that I aimed to capture a realistic picture of my everyday nutrition despite logging it.
(Who knows, maybe the slight negative correlation is due to the extra effort required to track these foods…)
Delving deeper, the insights suggest that the amount of calories itself doesn’t have any noteworthy correlation with recovery but this is not the case with individual macronutrients. Instead, it seems that for all macronutrients, less seems to be slightly better, or more significantly, more is worse.
What could this mean? Bad data?
After thinking about this for a while, and considering the lifestyle choices I’ve made during the tracking period, I think I might have a hunch of what might be going on here.
Because I didn’t try to specifically change the ratios of macros during the typical days, the higher amounts in each category likely often represent some kind of special occasion. For example, on the days I have consumed more protein, I have likely also done resistance training or we have been grilling and eating sausages and other heavy stuff, consequently having worse recovery on the next day.
Similarly, on the days I’ve consumed most carbohydrates, I’ve likely been eating candy or other unhealthy sweet things, which has likely also affected my recovery. Same thing with fat — maybe grilling, maybe junk food, maybe both.
It’s also good to remember the classic ‘correlation is not causation’ -statement regarding these results, even though the WHOOP journal labels behaviors as either ‘helping’ or ‘hurting’ recovery.
All in all, I think I have to revisit these correlations after I make changes to my diet on typical days so that the special occasions won’t skew the results too much in any direction.
However, maybe there is one interpretation to be made out of this already: enough calories with balanced macros is good, whereas an excessive amount of anything is bad.
Conclusion
To summarize, my diet seems to have the following problems:
- Too little protein, only 68% of the default target on average.
- Too much fat, up to 134% of the default target on average.
- Some micronutrient deficiencies, especially regarding vitamins A, D, E, and K, folate, beta-carotene, lutein+zeaxanthin, and lycopene.
To combat these, I will make the following adjustments:
- I will start eating breakfasts more regularly by making smoothies containing milk curd and berries my go-to breakfast. To vary things a bit, boiled eggs and omelets shall also be included in the breakfast menu. Additionally, protein curd will be the top snack choice during the day.
- I will replace 2 warm meals per week with salad including lean protein.
- I will increase my intake of lettuce, tomatoes, carrots, and watermelon because these are all available at the salad table of most lunch places I go to. Additionally, I will use more spinach, e.g. with breakfast omelets.
Bonus: I will include a multivitamin supplement that targets especially vitamins A, D, E, and K as well as folate into my diet.
Some Final Thoughts About Cronometer
In general, I found Cronometer useful for tracking nutrition intake for this limited period of time. Especially the comprehensive micronutrient information was interesting to see for the very first time. Furthermore, the more whole and unprocessed foods the diet consists of, the more accurate the micronutrient results will also be due to more complete nutritional information available for unprocessed items, which itself can motivate to include healthier foods into the diet.
However, despite the various features Cronometer contains to make tracking easier, the tracking process is still a lot of work. Therefore, I think the cons of the process will outweigh the pros for most users after a couple of months unless they are targeting a very specific amount of consumed calories or continuously making changes to their diet. Instead, I would estimate typically a month or two of tracking every now and then provides enough insight into the state of one’s nutrition to make the necessary adjustments for the better.
Next Steps
I have already started introducing the planned changes into my routine but as I’m going to be vacationing for some time now, I will not track my diet at this time. However, I am planning to do another month of tracking later this year and report back in a follow-up post about how much did these changes accomplish in the end.
Be sure to follow me and my new publication The Quantifying Stack to not miss the follow-up post and other posts about practical self-quantification tools and methods.
Thanks for reading!
What do you think of these results and my interpretations? Do they make sense or not? Did I miss something? What would you like to see in the follow-up post? Let me know in the comment section below!
