Azure Grant

Azure Grant
QS Labs Associate Editor

QS18 Preview: Map Your Ovulatory Cycle with Continuous Body Temperature

Ovulation Cartoon


Surprisingly little of the attention and funding turned to personalized, predictive, preventative medicine has focused on the female reproductive system: pregnancy onset cannot be quickly identified, menopause onset and trajectory remain entirely mysterious, and adverse reactions to tools like hormonal birth control are difficult to anticipate. Importantly, there are no automated, cheap, high-accuracy methods for predicting ovulation in the diverse population of cycling people. Existing methods are based on once-daily measurement of basal body temperature (BBT). These are prone to sampling error, unreliable for people with irregular ovulatory cycles, and make use of only one minute out of the 1440 available as data points per day. Although some pricey new devices aim to tackle the pitfalls of BBT via more frequent measurement, they mostly hold their analytical techniques as company secrets, making their estimations difficult to understand or trust.

So, if you are trying to get pregnant, trying to not get pregnant, or trying to understand anything else specific associated with your cycle, maybe you share my more general question: What can we learn about ourselves from continuously tracking body temperature across the cycle?

As part of my graduate research, and in collaboration with the Quantified Self community, I’m organizing a “participatory research project” on ovulatory cycling. To capture the uniqueness in individuals’ cycles, and to work towards predicting features of interest, we’ll measure continuous skin temperature, as well as luteinizing hormone (LH) and a range of other metrics. From my prior research, I’ve seen that mammals (including humans) have distinctive patterns of body temperature around ovulation, but that there are wide individual differences in the way temperature changes across the cycle. It’s my goal to characterize both the uniqueness of individuals and the common patterns preceding ovulation. I will of course release the analytical tools I build open-source. In this project, I’ll be requesting permission to work towards publishing results containing de-identified data from participants, but keep in mind that data sharing is completely optional; you’re still invited to join if you want to keep your data completely private.

We recently tested the plausibility of this style of participatory research in a year-long project exploring blood cholesterol variability. Thanks to research support from Amgen, participants had access to testing instruments that allowed them to measure their own cholesterol. We measured our blood cholesterol as often as hourly, and met every week or two to discuss our projects. We showed the work at QSPH 18, participants made a range of personal discoveries, and we generated two manuscripts, currently in review, to report our results.

If you think you might be interested in participating, we’re going to have a breakout discussion at QS18 called “Map Your Ovulatory Cycle With Continuous Body Temperature”. There, we’ll outline the proposed structure of the project and get a sense of what questions would be most interesting to ask with this data. I’d like to see if we can identify ovulation and menses onset, but there are many other possible topics. Are you interested in menopause? Pregnancy? Effects of disturbed sleep? Are you a non-cycling person who is interested in continuous, non-invasive biomarkers? If you can make it to QS18 in Portland, please come to our session and bring your questions!

Posted in Group Experiments, Personal Informatics, QS18 | Tagged , , | Leave a comment

Meetups This Week



This week there are two QS meetups happening. On Thursday, Gary Wolf will be hosting a somewhat unusual meetup at the National Archive in The Hague. Self trackers and archivists will get together to discuss current trends and issues around personal data and quantified self, including archiving methods and data privacy.  The Hong Kong meetup will also get together on Thursday to share what they’ve learned from their own genomic data.

To see when the next meetup in your area is, check the full list of the over 100 QS meetup groups in the right sidebar. Don’t see one near you? Why not start your own! If you are a QS Organizer and want some ideas for your next meetup, check out the myriad of meetup formats that other QS organizers are using here.

Thursday, November 16th

Amsterdam, Netherlands

Hong Kong, China


Posted in Meetups | Tagged , | Leave a comment

Quantified Self Guide to Tracking Cholesterol and Triglycerides

What is a QS guide?

The purpose of a QS guide is to make it easy for you to start tracking a new metric. Searching for the right device, head-scratching over how to use the thing, and figuring out what experiment to try first can be a huge time sink. Our goal is to offer a worked example of all of these steps with the device(s) we found to be the best on the consumer market. While the most sophisticated tools for physiological measurements are offered through professional laboratories, our guide is – of course – meant to help you with your own, DIY self-tracking projects. It’s not an extensive review of every option, but it will lead you from purchasing, to validating, to syncing the data, to doing a first experiment. As you go through the process yourself, much of your learning will come from building a mental model of how your own physiology works through additional reading and experimentation. Don’t shy away from that work – a QS project may not answer your question expediently, but it has the potential to teach you a lot.

The Guide to Tracking Cholesterol and Triglycerides will discuss two home lipid trackers: the CardioChek Plus and the Cholestech LDX. I will give an in-depth review of my experience with the former. The guide will touch on the science of the tests and devices, their accuracy and precision, and suggest a first experiment to try.

A little about blood lipids

While we normally think of cholesterol and triglycerides as risk factors for heart disease, there’s actually much more to them. In fact, it turns out that the basic functions of lipid components — including total cholesterol, triglycerides, HDL-c and LDL-c, not to mention their roles in heart health — are an active area of research and the center of an ongoing controversy. What lipid measurements can certainly do is reflect how your body is handling ingested animal products and fats. If you’d like to learn more, we put together an animation that goes a little deeper into the physiology.

40450Option 1: CardioChek Plus

What It Does:  The CardioChek Plus measures blood lipids including total cholesterol, HDL-c and triglycerides using test strips. The device itself is battery powered and about the size of a Game Boy Color (and it makes similar sounds!). Each sample requires 40ul of blood and takes a few minutes. The major limitation of the device is its range of operation (it won’t report results if your lipids are very high or very low -and different lots of test strips have different operating ranges. Be sure to read the documentation before you purchase). It is an FDA approved, CLIA waved, testing system for clinical and paraclinical use.

Cost: New units retail for ~$800-$1000, but units appear on eBay for around $400. The tests cost ~$15 each and come in bundles of 15. Additionally, you will need rubbing alcohol, 2.8 mm lancets40 ul capillary tubes for blood collection and gauze wipes. Cost of extra supplies comes to about $200 for 15 tests.

Getting Your Data: The CardioChek stores data locally and has a limited memory. We recommend transferring the raw data by hand (3 numbers per test) to a personal spreadsheet.

Accuracy, Precision and Supporting Research: Finding information about the accuracy and precision of a new device can be non-trivial. Confirming what you learn can be even harder. We’ve had several months to figure out measurement validity for home lipid testing, and it’s a little complicated. At present, there is measurable variability (~13% is acceptable) in results obtained from clinical laboratory tests (Quest, LabCorp) as well as those from para-clinical tests like the CardioChek Plus. Chris Hanneman has written a great report that comments on the not-very-useful way validations are reported by glucose meter companies – and we acknowledge that the same is true here. The company that produces the device, PTS Diagnostics, reports numerous validations at the bottom of this page under resources, but we’ve averaged the basics across these many reports to produce a summary table. 

Accuracy is a measure of how close a measured value is to the true value of the measurement (obtained via some gold-standard device). For accuracy, PTS diagnostics reports 18% error for total cholesterol (averaged across reported tests on the website in this document), 8% for HDL-c, and 13% for triglycerides.

At least one academic group has published a validation of this device: Gao et al., 2016 . They reported 3% error for total cholesterol, 7.1% error for HDL-c,  and 7.6% error for triglycerides.

Precision is a measure of agreement between two measures which should be identical. In other words, it’s a measure of how much noise the device adds to the signal. During our own testing we measured the precision of the CardioChek Plus; you can view our results here. We actually found the CardioChek to be more precise than the company reports (so far).


Cholestech LDX Analyzer OnlyOption 2: Cholestech

What it Does: The Cholestech LDX also measures total cholesterol, HDL-c and triglycerides. However, the device is larger (shoebox sized) and less mobile than the CardioChek Plus — it must be plugged into an outlet and calibrated in each new location.

Cost: New units retail for ~$2000, but used units can be easily found on eBay for around $50-$100 each.Note: make sure units have ROM pack version 3.40 or higher, and calibrate the used device.

Getting Your Data: Similarly, we recommend transferring the raw data by hand (3 numbers per test) to a personal spreadsheet.

Supporting Research: Whitehead, 2014 offers accuracy and precision measurements. Bias was 11.6% for total cholesterol, and 12.9% for HDL-c.  The authors reported %CV of 2-3.5% for HDL-c and total cholesterol (pretty good!) – with the caveat that the venous blood samples they compared are less likely to introduce measurement error in comparison to the finger prick samples used at home.

My Experience, and What I Tried First

I only had the opportunity to use the CardioChek Plus, but my comments should apply to both devices. Setting up the device is trivial, but testing requires a few practice trials.  The main challenge is the amount of blood required; it’s forty microliters (µl) which is equivalent to 2 large drops of blood. For some people I worked with, this was easy. But for others like myself, running the sampling hand under a hot tap is necessary to get the blood flowing. On top of this, the blood needs to be collected and deposited on the test strip within a couple of minutes to get an accurate reading. If this sounds a little off-putting, don’t worry too much- one becomes a blood collecting ninja fairly quickly. The pay of is in the ability to learn what my lipids are doing in near-real-time.

A First Experiment

While preparing for the project I wondered how fast my lipids really changed. I knew that seasonal, ovulatory cycle, and daily changes in lipids had been reported in the literature, but I wasn’t able to find any examples of how individual ambulatory humans varied hour by hour. The dynamic actions of these compounds on short timescales are less well characterized than changes on the timescale of years, but they are likely to contain useful health information. Because of all of this, I decided to measure my lipids every hour from the time I woke up, to the time I went to sleep. I won’t go into everything I saw here, but I will share one picture.

Azure cholesterol fig

I’m 22 and in good health, yet across a single day I saw my total cholesterol nearly 50 mg/dL (almost plunking me into the at-risk for CVD category). Even more interesting, these changes seemed to occur at regular 3 h intervals, gradually climbing higher until they peaked around 8 pm. I learned that these changes might actually tell me more about my health than any one of those measurements alone could have.  If you’re interested in getting a general sense of what your lipids are doing before you dive into more complex tests, I highly recommend setting a date with your CardioChek Plus or Cholestech LDX for some hourly measurements. Want a more in-depth argument for why you should try this first? Check out this animation.

This guide may have revealed that blood lipids are more complicated than you thought. But there’s no need to be overwhelmed — explore the metric, and you’ll build a deep understanding of your lipids in the process.      

Posted in QS Guide | Tagged , | Leave a comment

Meetups This Week

Screen Shot 2016-11-30 at 11.10.36 AM

Two meetups this week are happening on opposite sides of the U.S. If you’re interested in what happened at our conference in Amsterdam – or alternately what one man’s physiological response to seeing a bear was – then the Portland meetup is the place to be.

If you are on the East Coast, the D.C. meetup group has a great lineup of talks from students and faculty of John’s Hopkins University this Saturday. The meetup will host an introduction to tracking your metabolism and circadian rhythms by Tom Woolf, tracking your productivity with the OmniTrack platform by Eun Kyoung Choe, and insights from running and HR data.

To see when the next meetup in your area is, check the full list of the over 100 QS meetup groups in the right sidebar. Don’t see one near you? Why not start your own! If you are a QS Organizer and want some ideas for your next meetup, check out the myriad of meetup formats that other QS organizers are using here.

Tuesday, October 17


Saturday, October 21

Washington D.C.

Posted in Meetups | Tagged , | Leave a comment

Meetups This Week


Three meetups this week featuring show & tell talks, and in Manchester a fitness & diabetes risk assessment.

To see when the next meetup in your area is, check the full list of the over 100 QS meetup groups in the right sidebar. Don’t see one near you? Why not start your own! If you are a QS Organizer and want some ideas for your next meetup, check out the myriad of meetup formats that other QS organizers are using here.


Tuesday, October 3rd

Los Angeles, California

Manchester, London

Thursday, October 5th

Austin, Texas



Posted in Meetups | Tagged , | Leave a comment

Thomas Blomseth Christiansen: Over-Instrumented Running

Some More Instrumentation Thomas Blomseth Christiansen


“When in doubt, add more instrumentation.”

Have you ever felt that some parts of your life should remain unquantified? Perceived quality of your poetry, or perhaps duration of arguments with your significant other? Until recently, I kept running in my ‘not to be quantified’ bucket. It was such a meditative alone time that I didn’t want to risk disturbing it through observation. I tracked the kilometers I ran and nothing more.  That changed at the finish line of a recent 50k, where I learned that my wearable had overestimated the race distance by over 8 k! I was pretty miffed, and I turned to this talk for inspiration.

Thomas inhabits the far end of the quantified running spectrum. His talk from QS17 is a fun watch, and the project page is here.

What did he do? He started out  disappointed by hitting the wall midway through a marathon. This is common enough, but Thomas’ response to running a painfully positive split was to code his own negative split plan generator, rubber band split-plan sticky notes to his arm, and set out to run at a concrete pace. His project evolved to include an olympic swimming coach, many new devices, a metronome – and ultimately mastery of the art of pacing.

Posted in Conference, QS17, Videos | Tagged , , , , , | Leave a comment

Mark Moschel: Blood Ketones During Regular Fasting

MarkMoschelWine-1Here’s proof that clarity and creativity are what make great data visualization. Mark’s illustrations show what he learned by combining mulitple-day fasts, ketone and glucose measurements and…wine.

He has generated several interesting personal insights, including some not yet published on: correlation of felt energy levels to blood ketone levels, the inverse relationship between ketones and glucose, and the ceiling effect of too-high ketones. I can’t find any publications on the wine-effect, so there may be a novel discovery in there as well. Check out Mark’s QS project page here.

I periodically become fascinated with ketosis, so the talk inspired me to revisit the topic of ketosis and prolonged fasting in women. The debate about the issue is intense, and there are relatively few publications that address women specifically. Have women in QS tried a similar experiment? What was your experience? We’ve started a forum post here on the topic.

Posted in Conference, QS17, Videos | Tagged , , , , | Leave a comment

Meetups This Week


omg is this benjyWe have three meetups happening this Wednesday! If you’re in Hong Kong, check out an intro to self-tracking. If you’re in London, book a spot quickly (there are only two left) and head to Camden for QS talks and a walk to the pub. And if you’re in the Bay Area, the SF Women’s meetup is getting together for the first time in a while: bring something to share and join!

Wednesday, September 27

London, England

Hong Kong, China

QSXX – San Francisco, California

To see when the next meetup in your area is, check the full list of the over 100 QS meetup groups in the right sidebar. Don’t see one near you? Why not start your own! If you are a QS Organizer and want some ideas for your next meetup, check out the myriad of meetup formats that other QS organizers are using here.


Posted in Meetups | Tagged , | Leave a comment

Kyrill Potapov: Tracking Productivity for Personal Growth

Eddie_KyrilPotapov“Once one of Eddie’s leaves wilts, that’s it. A record of my failures right there among all the green leaves.”

In a show&tell talk that is as sweet as it is clever, Kyrill asks how his grandchildren might one day learn about him through digital family heirlooms and offers this unique project as an example.

Kyrill recently acquired his grandfather’s shaving razor and was struck by the connection he felt through the evidence of ownership: the darkened areas, worn edges and other traces of use.

Reflecting on his own mostly computer-based work, Kyrill noted how little of a physical trail he leaves in the world. Could his time and productivity data leave a mark on anything? Does he have a physical object, like his grandfather’s razor, that is indirectly shaped by his toil, besides a dirty keyboard?

Kyrill explored this idea by connecting the time-tracking service RescueTime to a light placed in a box with a house plant that he named Eddie. When he spends time on things he finds personally fulfilling, like working on his PhD, the light turns on and the plant grows. When he’s caught up in other activities, the leaves yellow and die.

The arrangement adds a new dimension to his productivity data. Every couple of days, Kyrill opens the box to water the plant. This ritual provides an opportunity to take stock on how he has been using his time, based on the condition of the plant. Embodied in this living organism is his failures to stay on task and focus on what’s important. Distractions take on a new threat. Rather than just endangering his goals, they now threaten the health of Eddie.

Although Kyrill won’t be able to leave a houseplant to his descendants, it’s a worthwhile meditation on how different modes of presenting personal data can have a profound difference in the way it engages one’s emotions.

You can watch Kyrill’s talk at his QS Project page. You can read about how Kyrill  connected RescueTime to a lamp here.

Posted in Conference, QS17, Videos | Tagged , , , | Leave a comment

Jakob Eg Larsen: Tracking Sleep and Resting Heart Rate


Jakob Eg Larsen has tracked his sleep and resting heart rate (RHR) for the past four years. His 7 minute talk is far better watched than read about: it’s a great illustration of data validation, longitudinal tracking, and data assisted self-awareness.

Briefly, by tracking his RHR over a long period of time, Jakob has developed an intuition for connections between his RHR and physiological state. He’s able to use the data to tune his self-awareness, but still keep a safety net when unexpected RHR elevations might portend a flu. To boot, the years of data across the Fitbit Blaze, Oura ring and Basis are one of the most extensive within-individual comparisons I can find anywhere of these devices.

You can watch the full video of Jakob’s talk at his QS Project page.

Posted in Conference, QS17, Videos | Tagged , , | Leave a comment