Tag Archives: individualized nutrition

PREDICT: The Science – Part 1

Overview of the ZOE Scientific Project for Researchers & Clinicians
A ScienceKara Summary

I’m technically neither a researcher nor a clinician, but let’s see if we can crack the code, eh?

What is PREDICT 1?

PREDICT 1 was the first of – potentially – many PREDICT studies to come in the future. Specifically, it was a multiple test-meal challenge study that 1,100 participants completed over 14 days. Scientists analyzed the participants’ metabolic response to standardized meals. With this data, researchers can drive the development of an algorithm for predicting metabolic responses. So in theory, the data from the PREDICT studies, with thousands of participants could help PREDICT scientists provide predictive insight on metabolic responses for millions of people interested in improving their health through diet changes.

There are five biological markers PREDICT 1 scientists analyzed as a response to standardized meals:

  • Postprandial glucose (this marker most of all, at least for PREDICT 1)
  • Triglycerides
  • Insulin
  • Inflammatory markers
  • Self-reported hunger

Participants in the PREDICT 1 study were 1,100 healthy people mostly from King’s College in London, UK. 60% of these participants were made up of twins from the UK Twins Study, which is spearheaded by PREDICT scientist Tim Spector, PhD. A Harvard researcher team led a validation cohort at Massachusetts General Hospital, and ZOE researchers completed an additional three-week sub-study of 100 UK participants to investigate metabolic responses to different types of standardized meals.

Researchers finished gathering data in May 2019. PREDICT 2 launched in June 2019. This time, PREDICT would be a non-hospital-based study, but the two PREDICT studies’ sets of data would still be able to be “seamlessly combined.”

What is PREDICT’s end goal?

The ZOE report describes two doors that could be opened by the data and insights gained by the PREDICT studies.

  1. “Data-scaling” – releasing a product outside of a clinical study in 2020
    • Customers undergo similar protocol as the one I’m following as a PREDICT 2 
    • Customers consent to sharing data anonymously
    • ZOE can make money AND gather data
  2. “Traditional science”
    • Short- and long-term nutritional intervention studies to test the accuracy of any predictions made via experimental algorithms and (later) the efficacy of ZOE diets under controlled conditions
    • Additional PREDICT studies to continue to gather data, but in a lab setting; this door provides a chance to collect data unique to the lab setting versus at data collected at home (like what I’m doing with PREDICT 2)

The introduction to the report outlines five key hypotheses that provide a foundation for PREDICT. Here’s the gist.

  1. Variability: People respond to food in unique ways and it can be predictable.
  2. Postprandial: Scientists need to study how the body responds and changes after eating a meal (“postprandial”). Measuring “traditional fasting markers” is not enough.
  3. Predict: Nutritional responses are directly linked to health conditions, and nutritional responses can be measured. Thus, health conditions can be predicted and addressed.
  4. Algorithm: The more data scientists can analyze from individual participants in studies like PREDICT, the more accurate machine-learning will be to build predictive tests.
  5. Complex: Science needs hands-on studies like PREDICT – not just traditional observational studies – to obtain insight needed for true predictive algorithm technology.

What’s taking so long?

Why haven’t scientists figured out how to predict individual responses to food yet? The variability and complexity of the factors involved is overwhelming; ZOE scientists list “background exposures” like:

  • Time of day
  • Recent exercise
  • Previous meals 

It takes time, resources, and diligent volunteers to make a study like PREDICT function and produce useful results. We all have the same letters in our genetic code (AGTC); the words are just written in different ways. Millions of different ways.


I started to count the number of times the ZOE scientists used the term “postprandial” and lost count. Needless to say, “postprandial” would be blowing up a word cloud made up of the text in this report. The necessity for postprandial study expressed by the scientists in this report is what led to both PREDICT 1 and PREDICT 2. These types of studies are a huge undertaking, needing to standardize meals for participants and control factors like fasting time without the participant coming to a clinic every day. Plus, the participants in PREDICT 2 apply and activate their own glucose monitors and collect, package, and mail their own saliva, stool, and blood samples! Not to toot my own horn, but I’m really doing science a favor with my dedicated and studious participation in this study (wink)!


PREDICT 2 Study Experience: Day 2 Reflection

This post is the fourth of a series on my participation in the PREDICT 2 study. Click the links below to read earlier parts of the series:

Well, I made my first major blunder as a PREDICT 2 study participant: I forgot to put my activity tracker back on my wrist after taking a shower! I realized what I had done after arriving to work. With my commute being 45 minutes, I knew it wouldn’t be feasible to return home to retrieve the tracker. I went to the “Help” feature on the study app and tapped “Chat with us.” I sent a message explaining that I had forgotten to put the tracker on that morning, and I asked if it would be okay to remain as inactive as possible before returning home at the end of the work day.

The study team member who answers in the morning, Haya, thanked me for letting her know and that she would make a note of it. She said to resume normal activity, log any exercise in the study app, and to put the motion tracker back on when I got home. Crisis averted! 

Day 2: Return of the Muffins

I had my day 2 breakfast at 8:03am, consisting of three standardized muffins, looking and tasting the same as the muffins from day 1. I also had a cup of coffee, which I logged in the study app. 

It was just breakfast muffins today, so I was free to eat whatever I wanted for lunch and dinner. Today for lunch I had a chicken breast (~10oz) roasted with thyme, salt, and pepper and a simple ratatouille (1 cup) made up of eggplant, zucchini, and tomatoes seasoned with thyme, salt, pepper, garlic, and lemon. I had the same dish in the same servings last night, so I used the study app’s meal-copying function to add the same item as last night’s dinner. 

It’s definitely more difficult to log made-at-home meals than I anticipated. Obviously it would be easier to scan a barcode or look up an exact meal off of a restaurant menu. But in a normal week, I cook most nights, and I want my habits during the study days to be as close to my normal as possible so my results represent my usual responses to food.

I read an interesting article in the blog on joinzoe.com about the muffins used in PREDICT. While as a participant I am blind to the details, there are three different muffins I could be consuming:

  • High-carb, low-fat
  • Low-fat, high-carb
  • Average fat and carb

All of the muffins are plain vanilla flavored and the look basically the same except for the blue muffins I’ll eat tomorrow for breakfast. No, they aren’t blue because of any blueberry flavoring. It’s food dye, and it serves the purpose of helping the study team identify “GI Transit Time” (GI = gastrointestinal as in the GI tract or “gut”). Catching on yet? After eating the blue muffins, I’ll be able to let the study team know through the app when I “see them again,” and this data tells them how fast certain food components move through my digestive system and out of my body.

The PREDICT “Muffin Master is nutrition scientist Dr. Sarah Berry from King’s College London. Together with test baker Kate, Dr. Berry created the magical muffin recipe for PREDICT so the study team would have palatable standardized food with known amounts of food components like fat and carbohydrate. 

For the most part, the muffins remove natural variability found in foods because of something called the “whole food matrix.” This is useful for the study, as their aim is to identify how people uniquely respond to high-fat meals, high-carb meals, and anything in between. But how does this data translate to the “real world” where people don’t eat standardized science muffins? I’ll be interested in learning more about the answer to this question as I read more about PREDICT and the science behind the study’s methodology. While I doubt the answer is simply “Well ScienceKara, the data translates seamlessly, and there’s nothing to worry about,” I also think my own answer right now would be: Well ScienceKara, we have to start somewhere.

Read the ZOE blog about the muffins.

Listen to the Gastropod podcast episode where the two hosts talk about their experience in PREDICT 1. Hosts interview:

  • PREDICT study scientist, Professor of Genetic Epidemiology at King’s College London, director of the Twins UK study, and ZOE co-founder Tim Spector, PhD
  • Timothy Caulfield, BSc, LLM, LLB, a Canadian professor of law at the University of Alberta, the Research Director of its Health Law Institute, and current Canada Research Chair in Health Law and Policy. Caulfield provides an interesting opinion on the efforts to personalize nutrition. He makes the point that with this type of technology ZOE is building with data from PREDICT, populations that arguably need a lifestyle change to improve health and prevent disease the most are not likely to get involved (i.e. low socioeconomic status).

Stay tuned for more posts about the science behind the study tomorrow.

PREDICT 2 Study Experience: Day 1 Reflection

This post is the third of a series on my participation in the PREDICT 2 study. Click here to read the first post of the series. Click here to read the second post.

On day 1, I completed three blood spot tests (before breakfast, after breakfast, and after lunch) and ate two standardized meals with fasting in between. 

Blood spot test 1: 8:10am

Individual items included in a blood spot test kit.

A few minutes after 8:00 this morning, I prepared to do my first round of blood spot tests for the day. The PREDICT 2 study team had a video for me to watch as well as written instructions for completing the blood spot test (As my formal training was in technical writing, I definitely appreciate well-written instructions and organization. Also, I have to admit I have found three small typos in the study’s hard copy written materials – whoops.). 

The blood spot test was quick, easy, and mostly painless. I washed my hands thoroughly and rinsed my left hand with warm water for about 90 seconds. This is to improve blood flow to my fingertips. Standing up at my desk, I used a small tool included in the study packaging to prick my finger and push drops of blood onto four absorbent circles on the blood spot card. To register the completed test in the study app, I took a picture of the blood spot card, scanned the card’s barcode, and recorded the exact time I began pushing the first blood drop onto the card. 

Then, I slide the card into a protective sleeve and into an aluminum bag, which would eventually hold all three of my blood spot cards for the day. This aluminum bag contains a desiccant to dry out the blood samples, and I will keep it in the refrigerator until I mail the samples back to the study team on day 4.  

Day 1 breakfast: 8:21am

My breakfast for day 1 was two semi-sweet, dense muffins just a bit larger than golf balls and what looked like about one cup of chocolate milk (milk I received from the study pack mixed with brown chocolate powder, also from the study pack). I had to finish the meal in less than 15 minutes and record the exact time I began eating. After I had eaten a few bites of the muffin, I was able to have coffee or tea – plain – and I chose coffee. 

I haven’t had chocolate milk in years, and the drink did not taste bad! It wasn’t a lot of liquid, and paired with my regular morning coffee, it wasn’t a burden to finish at all. The muffins were not light and fluffy like some muffins can be, but they were also not bad-tasting. I may not go to the store and pay for the very same muffins, but for study food, I feel like I can’t complain. 

After completing the breakfast, I logged all of the food by scanning the packaging and manually recording the coffee I drank. Then, I sat down at my desk, set a timer for two hours, and fasted until my next blood spot test. 

Day 1 lunch: 12:21pm

After doing my mid-morning blood spot test (10:21am), I fasted for another 2 hours and had breakfast a little after noon. Lunch was three muffins that tasted and looked identical to the breakfast muffins. No chocolate milk for lunch, though – just a glass of water (still, not sparkling) and eight fluid ounces of coffee (which I logged in the app). 

To my surprise, I was satiated after my breakfast and was only just starting to feel hungry when lunchtime came around. I’ll be interested to see how I feel as I fast over the next two hours before my last blood spot test of the day.

Day 1 Reflections

Throughout the day, I received notifications from the study app asking me to rank my hunger and alertness by moving a marker along a line, with the left being not hungry or not alert at all and the right being very hungry or very alert. Like I mentioned, I expected to be more hungry following my breakfast and lunch meals, but I felt full from the lunch for most of the afternoon. It wasn’t until about 5:00pm or so that I started wanting my next meal. For perspective, on a normal day, depending on what I eat for breakfast and lunch, sometimes I’m hungry as early as 3:00pm. If I’m hungry in the afternoon I’m much more likely to have a snack when I get home from work, and most often that snack is something unhealthy like chips or another processed food.

Answering the alertness question was more difficult. I had trouble falling asleep last night and thus woke up this morning feeling less rested than I normally would be after a good night’s rest. I’m not sure if the meals I was given were supposed to make me more or less alert, but I’m hoping that my low-quality sleep won’t affect my results from day 1. 

Overall, I think day 1 went pretty well. The meals were tolerable taste-wise and I haven’t felt overly hungry or uncomfortable in between. My biggest complaint is probably of the blood spot tests. My first test went fine; I filled each circle with blood easily. For both the post-breakfast and post-lunch tests though, I had to use a second, spare finger prick device each time to create a second cut on my finger to produce enough blood to fill the four squares. I’m unsure about why my blood flow was so poor for the second two tests, but I’ve got adhesive bandages on my left middle and index fingers, I can type just fine, and all will heal quickly! There is one day in between the two finger spot test days (a purposeful choice by the study team?), so I’ll be ready to do more blood spot tests when day 3 rolls around.

Tomorrow I will be posting a day 2 reflection, but in the days after, I won’t do a study reflection every day as many of the days will be the same after day 3. I’ll continue to post every day about the series, though, including a discussion of the science behind PREDICT 1, PREDICT’s sponsors and collaborators, a reflection on PREDICT 2’s recruitment process, and some other DNA and microbiome tests I’m doing in comparison to / reflection upon my experience with the PREDICT 2 study.

PREDICT 2 Study Experience: Set-up Day + The Study Pack

This post is the second of a series on my participation in the PREDICT 2 study. Click here to read the first post of the series.

Today was set-up day. I started off with a coworker (and friend, @Meghan) helping me install the glucose monitor on my left (non-dominant) arm. It takes 16 hours or so to calibrate, and I’ll keep it on for the duration of the study. Later I also “activated” my activity tracker (“activate” being I took it out of the package and put it on my left wrist like a watch).

Later in the morning I did potentially the strangest thing I’ll ever do at work. I’ll spare you the details; I’ll just say that the stool sample had to be collected during the set-up day so I did what had to be done. 

Before lunch I gave my saliva sample, which was quite literally spitting into a tube. I carried all of my samples home, and I’ll mail them back later this week. I had a call with a nutritionist from the study on my way home from work, and she confirmed that I had done all of my to-do items for the day. She talked me through the first couple of days of the study, asked if I had any questions, and let me know that if I ever needed anything, someone from the study would be able to chat with me through the app. That’s definitely a benefit to having collaborators from both the U.S. and the U.K. – different time zones so someone is always available for the study participants.

For tomorrow, “day 1,” I’ve got my standardized breakfast defrosting in the refrigerator and my standardized lunch ready to go. It looks like it’s muffins and chocolate shakes for me tomorrow! Of course, I can eat whatever I want for dinner, and I’ll just need to do some careful logging of the ingredients and serving sizes so the study team knows what nutrients I’m consuming. 

Standardized meals for day 1 of the study. I’ll scan the barcodes with the app to log each meal.

The study pack

I received my study pack in the mail on the Friday afternoon before I would begin the set-up day on a Monday. The box had several smaller boxes inside with labels and storage instructions. There was a long, thin box labeled “muffins” that I stored in the freezer, and I put a smaller box of other shelf-stable food items in the refrigerator. Non-food items included:

  • Food scale for weighing food ingredients for my meals that are not standardized by the study team
  • Cup and shaker for preparing shakes
  • Photo card for taking pictures of all the non-standardized meals that I eat
  • Study guide with information about the study
  • Tape measure for doing body measurements
  • Wearable activity tracker
  • Glucose monitor, reader, and adhesive patches
  • Return boxes for samples and devices
  • Sample collection kits for at-home collection and in-lab collection

My experience so far has been that the study team is extremely organized and prepared. In a white paper prepared by the ZOE team (“Overview of The ZOE Scientific Project for Researchers & Clinicians”) in June 2019, the authors mentioned that the PREDICT 2 study includes upgrades from PREDICT 1 that “improve data collection” and “reduce participant burden.” I’m pondering if those upgrades include the study pack experience because everything is extremely organized, and I can see how an unorganized box would be particularly disconcerting for the study participant, especially at the beginning of the study. 

That’s all I have for today, the set-up day. Tomorrow I’ll be posting a day 1 reflection as well as discussing the study that started it all: PREDICT 1.

Forging the Future of Personalized Nutrition: My Experience, My Contribution

Introducing the PREDICT 2 Study

I am less than 48 hours away. 48 hours away from beginning my participation in a study that will potentially provide insight into how my body metabolizes food. Not how people my age/gender/race tend to metabolize food, but how my body uniquely metabolizes food.

This study is riding the wave of personalized nutrition that’s been surging through the scientific community for the last decade. Because as scientists learn more and more about nutrition and the digestive system, the clearer it becomes that metabolism is a deeply personal experience.

Imagine this. Instead of following the latest popular diet (i.e. gluten-free, paleo, keto, intermittent fasting), people have the opportunity to take a test to identify their unique responses to different foods. For example, a gluten-free diet may work for some (Celiac disease, gluten intolerance), while for others there might be something else in wheat products that’s causing them discomfort. In theory, this test could identify what “something else” is. 


The study is called PREDICT 2, and it involves following a schedule of eating pre-prepared meals and providing blood, saliva, and stool samples for analysis. “PREDICT” stands for “Personalized Responses to Dietary Composition Trial,” and I’ll talk more about the study, its sponsor (a commercial company, Zoe Global Limited), and the collaborators involved (Stanford University, Massachusetts General Hospital, King’s College London, and Tufts University) in more detail later this week. I’ll also be blogging about PREDICT 1, a two-week study that measured physiological responses to specific foods. Researchers showcased preliminary results from PREDICT 1 at the American Society of Nutrition conference early in summer 2019.

12 Days

The study will take place over 12 days: one set-up day, 10 study days, and one follow-up visit to a clinic where I’ll have blood samples taken. I received my study pack in the mail two days ago, and it contained all of the items I’ll need throughout the study, including the standardized meals. 

On some days, I’ll have standardized breakfast and lunch meals that I’ll eat. Depending on the day and the samples I need to provide, I’ll need to wait 2-4 hours after each standardized meal before eating again. I can drink water, coffee, and tea during these times, but they recommend that I drink about the same amount of caffeine that I normally do every day.

I’ll log all of my non-standardized eating and drinking activity in a mobile app I downloaded on my phone. If you’ve ever used My Fitness Pal to track calories and nutrients you’ve consumed through food, the app as very similar to that.


On the set-up day, I’ll use materials I received in the study pack to collect a stool sample. This sample allows scientists to analyze the diversity of my microbiome, which is key to understanding my body’s unique response to certain nutrients in food. I’ll also collect a saliva sample on the set-up day, which provides the study team with samples of my DNA. The study team describes that DNA samples enable them to “identify certain genetic characteristics that have been previously associated with [my] responses to foods that [they] will measure during the study.” 

On some days, I’ll provide blood samples at specific times before and after meals. From these samples, study scientists can learn about how my blood fat levels change throughout the day and before and after specific meals. The study team describes blood fat levels as a “key metabolic indicator and one of the two main sources of energy in your body,” with blood sugar, or blood glucose, being the other main source of energy.

To measure blood glucose, I’ll be wearing a blood glucose monitor* throughout the study. I activate the monitor myself on the set-up day, about 16 hours before the first study day. The monitor takes this time to calibrate.

*This glucose monitor is FDA-approved for use in the management of diabetes but not for evaluating blood sugar levels in non-diabetic contexts, including this study

In addition to the glucose monitor, I will also be wearing an activity tracker on my nondominant wrist throughout the study to measure my physical activity and sleep levels. This tracker is very similar to FitBit technology.

After the 10 study days, I will visit a Quest Diagnostics Patient Service Center to provide a blood sample. The study team requires that I do this within one week of the tenth study day and before 11:00am so the samples can be shipped and processed at a central lab the following day. 

All of the sample analysis relies on knowing exactly what I consumed to elicit certain physiological responses. Thus, it’s crucial that I follow the food and drink schedule as closely as possible, eating and drinking standardized meals and drinks when scheduled and carefully recording my food and drink consumption at all other times.


I am not getting paid to do this study, but I am pretty pumped about getting my results back. The study team is hoping for 1,000 participants, so potentially you could be involved too! A few months after the study, I will be contacted by the study team to discuss my results. If I want them to, they can also share the results with my primary care physician. Essentially, I’m hoping for information on my personal physiological response to components of different foods and how those responses compare to common responses to food. 

While the insights gleaned from this study will likely help build an algorithm for predicting individual responses to food, the informed consent form states: “the predictive value of this research is not yet proven and it is unknown whether you will benefit from the information”

Over the next couple of days, I’ll be posting regular updates about the study as well as providing background information about the people and science involved. I’m really excited (as nerdy as it sounds) to participate in a study that’s doing awesome things for science by learning how different people respond to different foods and what factors account for those differences. I hope people will ask questions about my experiences and be inspired to participate in meaningful studies like this in the future!

Note: While the PREDICT 2 study team encourages participants to share information about the study, the statements made in this blog post and future blog posts are based on my own research and sets of opinions.