Episode 29 – Steven Hamley zooms in & out on dietary patterns affecting our health

Show notes:

  • Steven Hamley’s website http://www.stevenhamley.com.au/ and Twitter @stevenhamley
  • I ask Steven why he started his blog and he explains that it all started when coming across the Paleo concept in 2010, sparking his interest in health and disease given the utility of this framework compared to traditional modes of viewing these matters.
  • We talk about the following points mentioned in the AHA’s 2017 Presidential Advisory paper
    • In summary, randomized controlled trials that lowered intake of dietary saturated fat and replaced it with polyunsaturated vegetable oil reduced CVD by ≈30%, similar to the reduction achieved by statin treatment
      • Steven notes their conclusions are drawn from 4 core trials from the ‘60s and ‘70s that we’re particularly “clean and controlled”
      • I play devil’s advocate, asking Steven if these studies can adequately ‘control’ for the multifactorial nature of these dietary interventions in these study. He says there was no attempt to do it in these intervention studies and neither would it be appropriate (more appropriate for observational studies)
    • …replacement of saturated fat with mostly refined carbohydrates and sugars is not associated with lower rates of CVD and did not reduce CVD in clinical trials
      • In many observational studies, total carbohydrate and saturated fat have similar kind of risk, but in randomized controlled trials though this doesn’t hold true. Steven uses the Women’s Health Initiative as an example of where saturated fat doesn’t seem to be (as) bad
      • We discuss AHA’s ‘heroes’ are (1) the only non-essential macronutrient [carbs] and (2) PUFAs, the kind of fats we encountered in minute quantities during our evolution and still aren’t proven to be truly essential (~ 0.12% to >0.5% of calories, using Chris Masterjohn’s estimates)
    • low-density lipoprotein cholesterol, a cause of atherosclerosis
      • I ask Steven if the AHA needs this statement to be true to support their recommendations. He says that hypothetically, no, but in practice not everything that reduces LDLc will reduce heart disease risk. He says the association of LDLc and heart disease has been blown out proportion. Looking at triglycerides-to-HDLc and other risk factors are much more important.
  • We then move onto a paper asking testing the statement “patterns of insulin concentration during the oral glucose tolerance test (OGTT) predict type 2 diabetes?” (2013 Diabetes Care paper)
    • Steven explains that an OGTT involves drinking 75g so to then measure insulin and glucose in the blood. Importantly, he emphasizes that just because your glucose levels are normal, this doesn’t mean your insulin response is healthy. In fact, when Steven’s group performed OGTTs in other studies, finding that 90% of people with normal glucose tolerance had abnormal insulin responses.
    • People taking OGTTs with insulin can be categorized into 1 of 5 patterns (see below) and the patterns in this study are based on the peak time of the insulin concentration (similar but not identical to Kraft Patterns). If your insulin peaks at 30 to 60min, you’re not very likely to have type 2 diabetes in the future. But if it peaks later, you’re quite likely to have it. Later peaks commonly come with larger insulin responses (or poorly functioning beta-cells).
    • So can an OGTT with insulin assay help you predict your risk of T2D? Yes! If you are show a Pattern 4, you’re x12.55 more likely to have T2D (95%
      CI 4.79 – 32.89) and if you show a Pattern 5 you’re x8.34 more likely (95% CI 2.38 – 29.27)
    • For more on how to use such an OGTT test, please read up on Ivor Cummin’s take on it. Kraft Patterns take into account total insulin concentrations and other measures, not simply insulin peak timing.
    • Steven explains that insulin sensitivity is how responsive a particular tissue is to insulin. Insulin secretion is how much insulin your pancreatic beta-cells secrete.
  • Steven’s group published a paper in 2017 in the American Diabetes Association’s journal exploring what happens to your ongoing production of endogenous glucose after eating a certain amount of glucose. Steven blogged about it here
    • Different doses of glucose produced a dose dependent insulin response but near identical glucose responses. This is consistent with previous research.
    • Whether you ingested 25, 50 or 75g of glucose, the amount of endogenous glucose production (EGP) was quasi unchanged
    • The rate of appearance of glucose in the blood and the rate of its disappearance were dose dependent, such that higher glucose doses resulted in a greater rate of glucose absorption and glucose disposal – a responsive machine!
    • I asked Steven about how we can square his results with results Gabor and I have discussed in previous podcast episodes (7 and 8) regarding metabolic responses with starch rather than pure glucose solutions
      • I mention that, the breaking down of starch by enzymes (in vitro α-hydrolysis with α-amylase) of differently processed starches correlate strongly (r2 = 0.95, p < 0.0001) with the glycemic and insulin responses in human subjects ingesting those starches
      • Steven mentions this view is quite compatible with his results (and I’d tend to agree!). Incretins may be key molecules to explain the different metabolic responses to different starches with the same glucose load
  • We discuss Steven’s honour’s project that was published in one of Nature’s sub-journals in 2017
    • The experiment with mice lasted 65 days and they were put on 3 diets; (1) chow fed mice (2) HFD fed mice and (3) HFD fed mice switched back to a chow diet in the last 9 days of the 65 day experiment. The mice could eat as much as they wanted (ad lib)
  • Steven explains that glucose intolerance and insulin resistance is present in 3 days with the HFD fed mice
  • Steven also specifies that the Chow diet is nearly like cardboard, very unpalatable, whilst the HFD is much more palatable, akin to a sweet biscuit
    • He mentions the type of mice used here are genetically manipulated to get fat from a HFD
  • Getting the mice off the HFD was “associated with partial normalisation of weight, liver fat, liver AMP, and glycolytic intermediates”
  • This study helps reveal mechanisms whereby insulin resistance and glucose tolerance is improved
    • The liver is very responsive and the type and amount of fats present in it, as well as how these fats end up in the liver, is crucial to metabolic health
    • Energy balance certainly matters, but so does the macronutrient content


Leave a reply:

Your email address will not be published.