Monday, November 25, 2013

Coffee Rules, Green Tea Sucks!? Study Links Green Tea to Insulin Resistance and Coffee to Insulin Sensitivity. Statist. Outlier, Long- vs. Short Term Effect or Heavy Metal Toxicity?

I can't tell if bathing in coffee is healthier than in tea, but my gut tells me that it is.
Coffee may rob your sleep (see "Sleepness Nights: Are Pre-Workouts and A Huge Cup of Coffee Messing With Your Sleep?" | more), but if you look at the currently available epidemiological data, the average coffee drinker is still a pretty healthy chap. Whether this is due or in spite of the stimulating effects of caffeine is not quite certain, but a closer look at its molecular composition reveals that the average cup of real coffee has at least as many potent antioxidants, as green tea. From a molecular perspective coffee is thus by no means inferior to the catechin-laden purported health-elixir green tea. With its glucose-6-phosphatase inhibitory (van Dam. 2006) and insulin sensitizing effects (Shearer. 2003) of its chlorogenic acid and quinide, coffee has, just like green tea, tons of animal data to support its potent anti-diabetic effect.

What's healthier tea or coffee?

If you asked one hundred people, what they'd choose, coffee or green tea, if personal preferences weren't an issue and their only criteria were "beneficial health effects", I am 100% convinced that the vast majority would go with green tea. That's actually quite natural, the media does after all hype every epidemiological study that supports the notion that green tea drinkers are the healthiest chaps on earth. Studies like the one researchers from the National Center for Global Health and Medicine in Tokyo are about to publish in the venerable scientific journal Metabolism, on the other hand, are rarely present on any of the major science news outlets.
Did you know that... An inverse association between coffee consumption and HOMA-IR has been observed in the US, Sweden, the Netherlands and multi-ethnic Asian populations, the study at hand is yet the first to confirm the existence of health benefits in a Japanese population.
This is where the SuppVersity comes into play. Your place to learn that Pham et al. observed an "unexpected", statistically significant association between green tea consumption and elevated HOMA-IR levels - in other words: The regular consumption of green tea was associated with an impaired glucose metabolism / lower baseline insulin sensitivity.
Figure 1: Relative differences to coffee / green tea "no consumers" in subjects who drink different amounts of coffee and green tea every day; p-values are given in boxes only for significant changes (Pham. 2013).
These results were obtained using data from cross-sectional epidemiological surveys that had been conducted among employees of three workplaces, two municipal offices in Kyushu and one
manufacturing company, in Kanto in 2009 and 2012. Pregnant women and subjects reporting a history of stroke or cardiovascular disease (n = 25), cancer (n = 27), diabetes (n = 52) and chronic kidney disease (n = 9), as well as those who were current usersof anti-diabetic drugs (n = 1) or under medical care for hepatitis (n = 4) were excluded to minimize the influence of confounding factors and be able to make a statement about the associations between coffee and tea consumption and glucose management in the average healthy Japanese worker.
Table 1: Characteristics of the study subjects by coffee consumption; the yellow mark indicates stat. sign. inter-group differences (Pham. 2013)
"Information about dietary intake during the preceding month was obtained using a validated brief self-administered diet history questionnaire (BDHQ). Dietary intakes for 58 food and beverage items including coffee and green tea, energy, and selected nutrients were estimated using an ad hoc computer algorithm for the BDHQ, with reference to the standard tables of food composition in Japan.

The response options for coffee or green tea consumption were never, < 1cup/week (wk), 1 cup/wk, 2-3 cups/wk, 4-6 cups/wk, 1 cup/day (d), 2-3 cups/d, or ≥4 cups/d." (Pham. 2013)
It is probably not surprising that the comparatively detailed dietary analysis revealed significant correlations between the consumptions of coffee and green tea (Spearman’s r= 0.83 and 0.77 for coffee in men and women, respectively; Spearman’s r= 0.68 and 0.64 for green tea in men and women, respectively) - with a high enough subject base you can however subtract the individual influence of the two out.

Is there any reason green tea could compromise our insulin sensitivity

I have to admit that there is little direct experimental evidence that would explain the observations th researchers made, but if you remind yourselves of the first serving of the "Supplements to Improve and Restore Insulin Sensitivity" Series (overview), you will remember that the anti-diabetes effects of green tea / green tea supplements are - if they occur at all - indirect ones.

Don't forget that caps and pills are not worth a penny without you committing to the all the lifestyle changes I outlined in episode one of the "Supplements to Improve and Restore Insulin Sensitivity" series.
Unlike the bioactive substances in coffee, which have a direct beneficial effect on skeletal muscle glucose uptake, those of the often-hailed green tea catechins are primarily side effects of improvements in lipid metabolism and reduced chronic inflammation.

Assuming that the average study participant did not have significantly elevated blood lipids and wasn't chronically inflamed either, this would explain the absence of improvements in insulin sensitivity. What it doesn't explain, though, is the fact that the insulin sensitivity decreased - irrespective of the accumulating evidence which suggests that even diabetics don't benefit from green tea supplementation (latest example from Nov, 8  Wang. 2013)

Usually this is, when you take a look at the discussion in a scientific paper, read, digest and get an idea of what could be a reasonable explanation for unexpected phenomena like this. The conclusion of the paper at hand, does yet only summarize the existing scientific evidence that the vast majority of "previous studies have shown no association between habitual green tea consumption and fasting glucose" (Pham. 2013):
  • a Japanese study among 1542 men reportedmarginally-significant higher odds ratios for IR with higher habitual consumption of green tea (T Otake April 2013, personal communication of the authors of the study at hand)
  • Rebello et al. reported no association between green tea consumption and HOMA-IR among a multi-ethnic Asian population
  • a meta-analysis including 6 clinical trials found no evidence to support an effect of green tea catechins on HOMA-IR
As Pham et al. points out conclusive evidence for or against the benefits of green tea consumption from long-term + large-scale trials is missing. Most clinical trials including those in in the previously cited  meta analyses had relatively short-term intervention (4-24 wks) and modest sample sizes (23-88 subjects). Without further research it is thus more or less impossible if genetic, lifestyle or other confounding variables or a hitherto overlooked negative long-term effect of green tea are responsible for the results. 
Table 2: Levels of toxic trace elements in tea infusion; data compiled by Tanmoy Karak, R.M. Bhagat (2010)
Aside from biological factors, the disappointing results could also be the consequence of the quality of the tea. In Japan, the upper tolerable levels for lead (Pb) are for example 4x higher, than in Europe (20mg/kg vs. 5mg/kg). With the ever-increasing Pb concentrations in tea leaves (Jin. 2005), lead and other toxic elements (see data in Table 2) could be turning the health elixir into an increasingly toxic cocktail.
"Coffee, Tea, Cacao, Caffeinated Sodas & Breast Cancer: 5+ Cups/Day?! Study & Meta-Analysis Show, It May Take More Coffee Than Previously Thought to Ward Off Breast Cancer" | more
Treat with caution: I strongly recommend not to overestimate the results of the study at hand - neither the negative ones about green tea, nor the positive ones about coffee. A BMI-stratified analysis did after all reveal that the inverse association (not causation) between coffee consumption and HOMA-IR was really significant only in the overweight/obese, but not in normal-weight, individuals.

In view of the absence of a reasonable explanation for the pro-diabetic effects of chronic green tea consumption (it may in fact be heavy metals; see Table 2), I would not suggest you switch from green tea to coffee, if you don't like it. If you are a coffee-boy or -girl, anyway, the study at hand confirms that reductions in diabetes risk are not a good reason to give up on your beloved brown brew ;-)

  • Jin CW, He YF, Zhang K, Zhou GD, Shi JL, Zheng SJ. Lead contamination in tea leaves and non-edaphic factors affecting it. Chemosphere. 2005 Nov;61(5):726-32.
  • Wang X, Tian J, Jiang J, Li L, Ying X, Tian H, Nie M. Effects of green tea or green tea extract on insulin sensitivity and glycaemic control in populations at risk of type 2 diabetes mellitus: a systematic review and meta-analysis of randomised controlled trials. J Hum Nutr Diet. 2013 Nov 8.
  • Karak T, Bhagat RM. Trace elements in tea leaves, made tea and tea infusion: A review. Food Research International. 2010; 43(9):2234–2252.