Consensus in dietary recommendations: eat less, more veggies and fruits, less highly processed foods, less sugars. No consensus on fats, starch, alcohol, supplements, artificial sweeteners, or counting calories #isnpr2019
Why is there so little consensus? Getting good dietary data is difficult. Industry money often influences studies. Experts in the field rarely change their minds. @timspector #ISNPR2019
Microbiome: 100 trillion bacteria, archaea, and even fungi. 1000s of metabolic products. NO TWO PEOPLE ARE THE SAME. #inspr2019
Humans share 99.9% of DNA but any two people will only share about 20-30% of microbes with each other. Every area of the body (ie eyes vs elbows vs mouth) is also individual to everyone and different from each other. #isnpr2029
In a study of identical twins who were overfed in Canada, some put on 5kg, others 13kg. Obviously genetics will not explain this difference. @timspector #ISNPR2019
The DIETFITS study compared low carb vs low fat diets over 12 months. Average results in either groups was 5.5kg loss, but some people lost 25kg and others gained 10kg with very similar numbers between the two groups. #ISNPR2019
Dr. @timspector thinks that microbe diversity is one of the most important measures of health. In this slide he shows all diseases that are known to be worse with lower microbial diversity. #isnpr2019 
Many meds including most of those we commonly use both affect and are affected by the microbiome, including inhalers. Effectiveness of chemotherapy in melanoma is affected by the microbiome.
Probiotics: some work, we don’t know which ones in which person. New thinking is more about the metabolites (what the microbes produce) which are still different in people but not as different as microbiota. psychologytoday.com/gb/blog/evolut… #isnpr2019
For example, many of the microbes we consume in yogurt aren’t found in the gut, but people who consume yogurt have different metabolites than people who don’t and have lower visceral fat. #isnpr2019
Y’all are going to dig this science coming up. @timspector has been using continuous glucose monitoring on himself and in the PREDICT1 study for 1000 people (including a twin cohort) + 100 validation subjects. #isnpr2019
PREDICT1 uses CGM, triglyceride measures, insulin levels, microbiome and metabolome to measure individual response to foods. Food intake was measured using an app with photos and logging (no crappy food frequency questionnaire) with 132,000 meals logged. #isnpr2029
Dr. Spector used the CGM on himself and found a few things...his favorite breakfast of museli and orange juice strongly affected his blood sugar in a negative way. #isnpr2019
He also hadn’t believed that “low blood sugar” or up and down spikes caused lousy feelings in normal people (ie not truly “hypoglycemic” people) until he did a muffin test and tracked for himself. Are only muffins and felt bad and cognitively dull several times in the day.
What were some of the results of PREDICT1? There are 8-fold differences in insulin, glucose, and triglyceride responses between individuals eating the EXACT same meals. These differences are repeatable on the individual level, especially glucose &TG, insulin less so. #isnpr2019
Twins in PREDICT1 showed glucose responses to same meal were 30% correlated, triglycerides only 5% (compared to height in identical twins which is almost 100% correlated) #isnpr2019
(Just a note, of the 132,000 meals and responses measured, the original/baseline test meals for folks were all vegetarian, but the follow up test meals could be picked by the folks depending on their personal preferences) #isnpr2019
PREDICT1 Macronutirents (carbs, fats, protein) explain only a *small* proportion of individual response to foods. Microbiome and circadian timing of meals led to huge differences in responses (cc @CaloriesProper). #isnpr2019
What do we learn? 1 size fits all nutritional guidelines will NOT work. Using algorithms from the data of PREDICT1 you can program appropriate diets for people with about 75% success. (This was done in 100 validation subjects at MGH.) #ISNPR2019 @timspector
PREDICT1 is now being repeated in 1000 Americans. Also, dropout rates in the initial studies were only about 2%...the interactive app was felt to be important in this low number.
The second speaker today is Marjolein Visser PhD who ran the Food Mood project in Europe, a 5 year prospective observational study and a nutritional/supplement randomized trial designed to prevent new onset of major depressive disorder in a high risk population.
Her group ran a meta-analysis of all the *prospective* observational trials of food and mood and depression trying to standardize the data for diets used and depression confounders and measurements. #ISNPR2019
They found what other observational trials have found, approx 30% reduced onset of depression in people who more closely followed Mediterranean diet patterns (in Europe this was the Dutch, not the Italians...med diet pattern in studies is based on Italian eating 50 years ago)
Dr. Visser also did a 12 month RCT of +/- supplements +/- Dietary CBT on 1000 adults with BMI of 25-40 and with prodromal depression symptoms (PHQ9 5+) who were not currently or in the last 6 months depressed and not on antidepressants. #ISNPR2019
They screened 6000 to get the 1000 in the study. Here’s the design:
In the follow up there was a lower than expected incidence of new onset depression, 10% (expected in this population would be 20%). Overall the results were a wash. (F-BA is food-behavioral activation therapy which is using CBT paradigm to improve diet)
The supplement was 1400 mg omega3 (3:1 EPA:DHA), calcium, selenium, folic acid, and vitamin D. #isnpr2019
Higher adherence to the actual therapy sessions (15 individual and 6 group) resulted in lower anxiety, depression but greater adherence to healthy dietary pattern was not related to depression at the end. Taking supplements was slightly associated with *worse* outcomes.
(I personally think the supplement they used in this trial is bananas and I don’t know how they came up with it. The only thing that makes sense is the o3)
