Salivary amylase of course is the first step in starch digestion, but the amount between different people can vary by quite a bit due to things like stressors (you can manipulate concentrations with a beta blocker!) or genotype. With regard to the latter, there was a really fascinating study published in 2007 that found a correlation between the number of salivary amylase gene (AMY1) copy numbers and enzyme levels and the starch content of the population’s diet in which the individuals were from. PZ Myers has a nice post on it here. In 2010, a paper was published that showed AMY1 copy number again predicts salivary amylase concentration and activity level, and influences oral perception of starch. This is a highly suggestive example of evolution in action in response to the dramatic increase in starch in many populations since the agricultural revolution.

Until recently, according to the authors of the new paper, it wasn’t understood if salivary amylase added much to digestion: food passes through the mouth rather quickly, and we have pancreatic amylase anyway for starch digestion.  And, AMY2 (pancreatic amylase) doesn’t show the copy number variability that AMY1 does, suggesting AMY1 may be of greater significance to starch metabolism. In addition, studies have found that swallowing whole starchy foods results in lower blood glucose concentrations than when chewed, and that starch itself protects salivary amylase from the acidity of the stomach.

But what does this do to blood glucose concentrations after ingestion in people who have more salivary amylase compared to those who have less? This is what they studied. They hypothesized (as common sense would suggest) that those with more salivary amylase would have a higher postprandial glucose response, since they would break down starch faster and more efficiently. But that is not what they found!

Methods

The study consisted of 10 “high amylase” (enzyme concentrations per minute calculated by salivary flow rate that were 1 standard deviation higher than group (n=48) mean) and 9 “low amylase” (1 standard deviation lower) participants. Side note: they have the subjects chew on a square of parafilm for 90 seconds before collecting the saliva sample. With a google search I see this is commonly used for this which makes me wonder 1) who originally stuck some in his/her mouth and thought “hey lets use this!” and 2) why I now feel obliged to chew on some too. On 2 occasions, the subjects consumed (after fasting overnight) 50 grams of a corn starch hydrolysate solution or 50 grams of a glucose solution (as the control; salivary amylase obviously does not act on glucose!). These were consumed over 20 minutes, during which their rate was monitored, and they were instructed to swish every sip around their mouth for ~5 seconds prior to swallowing. Blood samples were collected intermittently over 2 hours for plasma glucose measurements and to measure AMY1 copy numbers. Lastly, they assessed the diets of each subjects with a food frequency questionnaire to see if there was a relation to self-selected starch intake (there was not). The authors write that “5 individuals were removed from the analysis based on the exclusion criteria described in the “Methods,” but I could find no such criteria (hmm?). This left 7 subjects in each group.

Results

They found a positive correlation (r=0.90, p<0.0001) between amylase concentration and AMY1 gene copies number. Here is a table with group characteristics:

As you can see salivary flow rate, salivary amylase concentration and activity were all significantly higher in the high amylase group, along with the AMY1 copy number.

Here is the most interesting result: the high amylase group had lower plasma glucose responses after consuming the starch solution than the low amylase group at 45, 60, and 70 minutes (p<0.01, p<0.001, and p<0.01, respectively), as well as a lower AUC (89 +/- 21 vs 244 +/- 55 mmol/L over 120 minutes, p<0.05) and lower peak glucose (9.56 +/- 0.43 vs 7.57 +/- 0.35 mmol/L, p<0.01). Resting blood glucose was not different between the groups.

Here is plasma glucose after the consumption of starch solutions (darker line is the low amylase group):

To convert from mmol/L to mg/dL (what we more commonly understand glucose concentrations in), multiply by 18.02. At the hour time-point (the greatest difference), that gives us about (visually estimating from the graph) 150 mg/dL in the low amylase group compared to about 110 mg/dL in the  high amylase group. A ~40 mg/dL difference at this point is quite large.

Interestingly, plasma insulin concentrations were not different between the groups at any of the time-points when analyzed over the whole period, but the high amylase group had higher insulin at the 9 minute point when analyzed separately, as well as a higher AUC from 0-9 minutes. They found a moderate correlation (r = 0.70, p<0.01) between this AUC and oral amylase produced per minute. I am a bit skeptical about these because these are during the time when the subjects are consuming the solution, and the rate of consumption though monitored couldn’t have been exactly equal in each, and apparently none of this was blinded. But here is the graph:

There were no differences between the groups for the control glucose solution for plasma glucose or insulin (at any time-point, AUC, or peak, or within the first 9 minutes).

Finally, they analyzed for differences within each group as well and found that the low amylase group had a larger postprandial glucose AUC following the starch solution compared to the glucose solution, and a higher glycemic index for the starch solution than the high amylase group (111 +/- 7 vs 94 +/-3, p<0.05) which is interesting to consider.

Conclusions

Why would people with more salivary amylase have a lower glucose response to starch consumption? The authors suggest that the higher plasma insulin concentration within the first minutes during consumption may mediate this, even though there were no overall differences in insulin responses to starch or glucose between the groups. They make a reasonable case why this could be, even though I am not sure this data alone is very convincing.

The rise of insulin within the first 15 minutes of consumption is “preabsorptive,” (or the cephalic phase) because no glucose has been absorbed yet. Insulin rises here in part because of a conditioned/anticipatory (Pavlovian) response to the flavor or smell of food. Research has demonstrated that this initial response is important for normal glucose tolerance in animals and humans, even though it is only a small part of total insulin release. According to the data here, the low amylase group did not show this preabsorptive insulin release to starch consumption but they did to the glucose solution, while the high amylase group showed for both. This may suggest that salivary amylase actually stimulates this insulin phase in some way. The authors suggest that perhaps low concentrations (too low to perceive) of glucose and/or maltose from some starch digestion in the mouth triggers T1RS-T1R3 (sweet) taste receptors or glucose transporters in taste receptor cells. Or, short-chain oligosaccharides produced might bind to the “putative polysaccharide receptor.” Or, “hormones or incretins are peripherally released by lingual taste cells into the blood stream in response, stimulating insulin release from the pancreas during the PIR [preabsorptive] period,” for which they provide no references (how is that not challenged in peer-review?).

Clearly much work obviously remains to test these theories.

The authors suggest that the differences in blood glucose may underlie some of the differential development of insulin resistance and diabetes between populations. It is possible- there is a strong observational link between postprandial glycemia and many chronic diseases. This of course needs much more study to verify. Is it possible that AMY1 gene copy number could help in risk assessment? It is an intriguing hypothesis.

I would really like to see this replicated in high and low amylase groups with real foods instead of swishing solutions which may exaggerate the differences (though actually probably better represents the reality in which most in industrialized countries consume them!), and with more subjects to reduce the effect of intra-individual plasma glucose measurement variation (and other random variation) but fascinating nonetheless!

Reference

Mandel, A. L., & Breslin, P. A. S. (2012). High Endogenous Salivary Amylase Activity Is Associated with Improved Glycemic Homeostasis following Starch Ingestion in Adults. The Journal of Nutrition. doi:10.3945/jn.111.156984

Design

The Health Professionals Follow-up is a prospective cohort of males started in 1986, and at baseline and every 4 years participants filled out a (validated, fairly close with food journals) food frequency questionnaire that asked about usual intake of sugar-sweetened and artificially-sweetened beverages. This analysis contained 42,883 participants, and a subset of these (18,225) gave a blood sample between 1993 and 1995 (relevant data from these are total, HDL, and LDL cholesterol, triglycerides, HbA1c, CRP, IL-6, TNF-1 & 2, ICAM-1, VCAM-1, adiponectin, and leptin). Participants/family reported non-fatal and fatal heart attacks (incident coronary heart disease) by questionnaire and these were confirmed with medical records.

For the analyses, time-dependent proportional hazard modeling was used with cumulative beverage and diet intakes updated every 4 years. A second analysis compared with only baseline diet data. Keep that in mind as compared to the study that found a positive association with cardiovascular events only collected a baseline food frequency questionnaire. Beverage consumption was divided into quartiles and modeling was adjusted for smoking, physical activity, alcohol intake, multivitamin use, family history of coronary heart disease, pre-enrollment weight gain, weight loss, low-calorie diet adherence, total energy intake, and BMI. Importantly, the also controlled for diet pattern with the Healthy Eating Index which scores by intake of fruit, vegetables, nuts/soy, cereal fiber, PUFA/sat fat ratio, white/red meat ratio, alcohol intake, multivitamin use, and trans fat intake.

Results

Again prior to statistical adjustments, relationships of beverage consumption with lifestyle variables is interesting; people who consumed sugar-sweetened drinks, smoked more, had lower physical activity, lower overall diet quality (HEI), but a decreased family history of coronary heart disease. They also had higher weight gain prior to enrollment and decreased weight loss, and lower adherence to a low-calorie diet. Consumers of artificially sweetened beverages tended to smoke less, have a higher physical activity, but a greater coronary heart disease family history. Although overall diet quality was higher and they were more active, they tended to have greater incidence of high triglycerides, cholesterol, and blood pressure- small differences but still there. Strange.

3683 incidence coronary heart disease cases over 22 years (790,852 person years) occurred. Prior to adjusting for covariates, comparing the quartile of most consumption for both sugar-sweetened drinks and artificially sweetened drinks, they found a RR of 1.21 for sugary drinks (95% CI, 1.10-1.33 p for trend < 0.01) and a RR of 1.04 for artificial sweeteners (0.96-1.15 which is not significant). Another good lesson of the complexities comes from the results after adjustment. Adjusting for smoking, physical activity, alcohol intake, multivitamins, and family history slightly reduced the risk for sugar beverages, but increased the risk for artificial beverages into significance (RR = 1.10, see table below for CI and p values). Further adjusting for pre-enrollment weight change lowered this back into non-significance, then adjusting for dietary factors raised it to significance again, and adjusting one last time for BMI removed it! Final adjustment for past diabetes, high lipids, and high blood pressure didn’t much effect. So the increased risk of heart attacks with sugar beverages in the upper quartile (median 6.5 servings per week) was about 20% compared to not drinking any, while there was no risk with artificially sweetened beverages. Also note the lack of significance at other quartiles, even for sugar sweetened beverages with a median of 2 drinks per week. And, converting the RR to absolute risk after all adjustments in quartile 4 for sugar beverages, this is only going from a .47% risk per year to .55% risk pear year of heart attacks in this population if my math is right. Not so scary!

The results were similar when intake was treated as continuous (except for carbonated non-colas which was consumed less & doesn’t really make sense here). Importantly, artificially sweetened colas did not increase risk of coronary heart disease. So 1 serving per day of a sugar-sweetened beverage increases coronary heart disease risk by about 20%, but we cannot say that artificially sweetened beverages increase risk.

They did the analyses again only baseline beverage intake associations were similar when coronary heart disease incidences in the first 4 years were removed (not done in the other study), as well as when diet was updated every 8 years- good to know for future research I imagine.

Among the previously mentioned blood markers, 1 sugar-sweetened beverage serving was associated with a significantly increase in triglycerides, CRP, IL-6, TNFalpha1 & 2 (all inflammatory factors which corroborate other research with sugar), lower HDL, Lp(a), and leptin, whereas artificial beverages did not reach significance in any.

Though there are some limitations to these types of designs, we have increasing confidence that the original positive associations between artificially sweetened beverages and cardiometabolic diseases were the result of insufficient controlling for confounding variables.

Reference

Lawrence de Koning, Vasanti S. Malik, Mark D. Kellogg, Eric B. Rimm, Walter C. Willett, & Frank B. Hu (2012). Sweetened Beverage Consumption, Incident Coronary Heart Disease and Biomarkers of Risk in Men Circulation : 10.1161/​CIRCULATIONAHA.111.067017

Luckily we didn’t have to wait much longer for a stronger research design. A new paper by Kiyah Duffey and colleagues sought if food patterns do change the diet beverage-cardiovascular risk link, as there was little statistical adjusting for potential diet-related confounders in the 3 epidemiologic studies that found positive associations between diet soda and metabolic syndrome and diabetes (see previous post). They also point to the paper by de Koning, Malik, Rimm, Willett, and Hu that found no effect in the Health Professionals Follow-Up Study cohort of artificially sweetened beverages on diabetes risk in men after adjusting for a number of confounders- more than the previous studies but still overall diet pattern was not considered.

The new study used data from the 20 year CARDIA prospective study, of 4161 subjects. Diet information was obtained at baseline by a diet history questionnaire followed by a diet history over the previous month. Foods and beverages were classified by an algorithm into 43 food groups for a cluster analysis by diet patterns. The CARDIA study has follow-up exams/diet data 6 times after baseline, but the authors state that cluster analysis with 7 and 20 years after baseline showed stable diet patterns so they could use just baseline data. Diet beverage consumers were classified as “consumers,” and non-drinkers as “nonconsumers.” Diet patterns were divided into “prudent,” or higher in fruit, fish, and whole grains, and “Western“, or higher in fast foods, refined grains, and sugar-sweetened soda. Metabolic syndrome components (i.e. waist circumference, fasting glucose, blood pressure, serum triglycerides, and HDL) were assessed at each of the 6 exams over the course of the study.

The following self-reported data was also collected at baseline: race, sex, age, smoking status, highest education, physical activity score, and family structure. These were included in statistical adjustments. So there are still some inherent limitations in self-reported data, only using baseline data (though they show this is relatively stable over time), sample size, etc but such is the difficulty of nutrition research on tight budgets. This study also did not stratify by dose as Gardener et al. did.

Here is an example of why it is so difficult to disentangle what effect diet beverages really have:

• people who didn’t consume diet beverages consumed more “healthy” nuts, seeds, vegetables, and milk, but also more “less healthy” high-fat refined grains and sugar-sweetened soda
• people who didn’t consume diet beverages who consumed a Western diet pattern got more calories from snacks and fast food than people who did consume diet drinks and ate that pattern
• people who didn’t consume diet beverages who ate a prudent diet pattern got more calories from fruit and low-fat refined grains.
• people who didn’t consume diet beverages in both diet patterns consumed more total calories (interesting find!)

Results

Unsurprisingly, the prudent diet group had a significantly lower risk of metabolic syndrome and several individual components than the Western group (note that it is not all diet responsible for this, as this group also tends to be more educated, active, and have lower BMIs, etc).

Without adjusting for diet pattern, those who didn’t consume diet beverages (nonconsumers) had a lower risk of metabolic syndrome but only 1 of the individual components was significant. Here is the table when diet pattern and beverage consumption was considered together:

Although those consuming the prudent diet pattern that didn’t drink diet beverages had significantly lower metabolic syndrome incidence compared to the Western diet pattern/consumers of diet drinks, it is notable that Western/nonconsumers of diet drinks wasn’t statistically different. In addition, there is little consistency in the results for individual components of metabolic syndrome: sometimes risk is lower if you are not consuming diet drinks, but sometimes it is lower if you are consuming them, and any significance was only reached for the prudent diet pattern. Together, it seems likely that there is no real relationship, though confident conclusions of course cannot be made either way for the various results. These results make little sense in the light of a lack of mechanistic theories, and based on the research we have so far I think those making suggestions to drop diet beverage consumption because of supposed risks is highly inappropriate.

Reference

Duffey, K., Steffen, L., Van Horn, L., Jacobs, D., & Popkin, B. (2012). Dietary patterns matter: diet beverages and cardiometabolic risks in the longitudinal Coronary Artery Risk Development in Young Adults (CARDIA) Study American Journal of Clinical Nutrition DOI: 10.3945/ajcn.111.026682

• cool work, nice post on it: Evolving Health: One Tomato at a Time: Feeding the World with Controlled Environment Agriculture
• a potential mechanism on why PUFAs are consistently found positive to cardiovascular health PUFAs acutely affect triacylglycerol-derived skeletal muscle fatty acid uptake and increase postprandial insulin sensitivity
• a meta-analysis finds a 20% crop yield gap between organic and conventional agriculture: GMO Pundit
• Good post by Mike Gibney on risk perceptions gibneyonfood: Snakes, astronauts and consumer risk perception
• theobromine is why chocolate is toxic to some pets, interesting article: The Curious (Toxic) Chemistry of Chocolate | Speakeasy Science
• great post filled with references Chiropractic: An Indefensible Profession – Skeptical Health : Skeptical Health
• I like this summary by Yoni Freedhoff and Scott Kahan (I’ve blogged on both Cochrane reviews) 2011 – Review: Interventions to reduce dietary salt do not reduce mortality or morbidity

• Larry Parnell, Jose Ordovas et al have a new paper on SIRT1 & CLOCK variants on resistance to weight loss: SIRT1 and CLOCK 3111T>C combined genotype … [Int J Obes (Lond). 2012] – PubMed – NCBI
• policy worked: Blood Levels of Trans Fats Are Declining in Americans – NYTimes.com
• space-nutrition? Vision Changes after Spaceflight Are Related to Alterations in Folate– and Vitamin B-12–Dependent One-Carbon Metabolism
• very interesting work but a lot still needed to get to an appropriate confidence in humans: Fasting might boost chem’s cancer-busting properties
• important editorial on the (lack of) solid evidence for some practices: Nutrition for Critically Ill Patients – — JAMA
• interesting pilot study on seasonal micro-nutrition: Periconceptional maternal micronutrient supplementation leads to widespread changes in the epigenome: a study of a unique resource in the Gambia
• Yoni commented on the (diet) soda/vascular event paper that I blogged on: Weighty Matters: Why that Diet Soda/Stroke Paper is Worthless and a Failure of Peer Review
• Excellent post by Scott Gavura: Science-Based Medicine » IgG Food Intolerance Tests: What does the science say?

• hope to do a long post on the PCRM eventually, but this was annoying. NPR consulted an embassy for an alternative perspective to the PCRM claims about cheese on their billboards instead of the science. Poor journalism in my opinion Billboards Slather On The Guilt With Anti-Cheese Campaign : The Salt : NPR
• PLoS ONE: Mimicry of Food Intake: The Dynamic Interplay between Eating Companions
• Interesting about treating to cardiovascular risk (multiple risk factors) instead of to target LDL levels Hayward and Krumholz: Open Letter to the Adult Treatment Panel IV of the National Institutes of Health « Marilyn Mann’s Blog

• people are irrational, part eleventy million There Are No Fetal Cells In Soda, But Fear Is A Real Threat To Biotech – Forbes
• Huge Study Finds Risk Factors Do In Fact Predict Risk
• more evidence in humans Brown Fat Burns Ordinary Fat, Study Finds
• PLoS ONE: Modulation of miRNA Expression by Dietary Polyphenols in apoE Deficient Mice: A New Mechanism of the Action of Polyphenols
• the 1500mg/day sodium goal set for a subpopulation of Americans is not compatible with the 2010 dietary guidelines for nutrient-adequate diets, concludes this study A conflict between nutritionally adequate diet and meeting the 2010 dietary guidelines for sodium [Am J Prev Med. 2012] – PubMed – NCBI

The prospective study consisted of 2,564 subjects who went through interviews, physical examinations, and some measurements between 1993 and 2001 (average follow-up was 9.8 years). A food frequency questionnaire estimated their consumption of regular or diet soda consumption as well as other dietary factors. They were screened annually after this- but not for soda intake, which was only collected at baseline and correlated with future events, which is a major weakness of studies like these. Vascular events included stroke, heart attacks, or vascular death.

They performed multivariate analyses to adjust for a number of possible confounders, but it is always possible there are habits associated with diet soda consumption that aren’t controlled for. For example, before adjustment, the authors found that frequent diet soda consumption was associated with former smoking, hypertension, elevated blood sugar, lower HDL, elevated triglycerides, increased waist circumference, BMI, peripheral vascular disease, previous cardiac disease, and the metabolic syndrome. So right there we know that 1 or few studies on this topic is not going to be enough for strong conclusions because of so many potential confounders. Importantly, different study designs and potential mechanisms (of which are lacking, which also should weaken conclusions of a link) should build an evidence-base that collectively suggests that there might be a risk of diet soda and vascular disease. This one study is fine by itself- but the conclusions being drawn in the press are not.

They found no association between regular sugary soda consumption and risk of vascular events at any dose of consumption, which is surprising given previous research that has demonstrated consistent associations, and should raise the question about why fact went relatively unnoticed. In fact, light regular soda consumption was associated with a reduced risk. Diet soda was positive for vascular events at all but 1 dose (the lightest consumers) in all of the statistical models they did. Of note, however, the number of subjects and events in the group consuming the most diet drinks was quite small (163 subjects, 51 events).

As the authors point out, their study is the first to specifically examine the relationship between vascular events and diet soda consumption. But one with coronary heart disease has been studied in the nurses health cohort, with neutral results after similar statistical adjustments. In other words, there was no relationship between diet soda and heart disease. The authors note:

“Beyond the inherent differences between our NOMAS cohort and the NHS in terms of age, sex, race-ethnic composition, and sample size, reasons for the discrepant results for the association between diet soft drinks and cardiovascular disease are not immediately obvious, underscoring the need for further study in other cohorts.”

On the other hand, several other studies have found positive associations between diet soda consumption and risk of metabolic syndrome in the Framingham cohort, the MESA cohort (also increased diabetes risk), and the ARIC study. However, they all caution on interpretation- it could simply be that diet soda consumption is a marker of poor diet/lifestyle habits and residual confounding by unknown variables (there may be many) that aren’t statistically controlled for could explain the results. This is why we need more research before drawing conclusions.

An important thing to note is the weak mechanistic explanations. The relationship could be real, but it also could be an artifact; some strong mechanisms could strengthen the theory, but there aren’t any. They cite the fact that research suggests artificial sweeteners increase (or at least don’t decrease) body weight (good review here), but they controlled for that. They suggest the caramel coloring could influence but through glycation but I am skeptical and this should be apparent in the regular group as well (they didn’t analyze for it either). Overall it is not convincing.

Reference

Gardener, H., Rundek, T., Markert, M., Wright, C., Elkind, M., & Sacco, R. (2012). Diet Soft Drink Consumption is Associated with an Increased Risk of Vascular Events in the Northern Manhattan Study Journal of General Internal Medicine DOI: 10.1007/s11606-011-1968-2

• A large position paper: Pesticides and Health, myths vs. realities
• EDIT: See comments about the group who wrote this paper
• Another lab shows beetroot juice (dietary nitrate) benefits on endurance (cycling) performance Nitrate Supplementation’s Impro… [Int J Sport Nutr Exerc Metab. 2012] – PubMed – NCBI
• Emily Willingham published a great piece about the shoddy Atlantic article mentioned last week Genetically modified foods: Ari Laux’s alarmism in the Atlantic – Slate Magazine
• Another study finds microRNA in breast milk Immune-related MicroRNAs are Abundant in Breast Milk Exosomes [Int J Biol Sci. 2012] – PubMed – NCBI (older one here)
• More information on the resveratrol research fraud Red Wine and Lies – Percolator – The Chronicle of Higher Education
• 13.8% of Americans (NHANES sample) had antinuclear antibodies and they are lower in overweight/obese- wonder what that means Prevalence and sociodemographic correlates of antinuclear antibodies in the United States – Satoh – Arthritis & Rheumatism – Wiley Online Library
• Yoni has a rational take on Paula Deen: Paula Deen has diabetes and takes Victoza, so what?

It uses NCBI E-Utilities, javascript, and Google Spreadsheets to automate tracking of paper count in the pubmed database for a number of search terms. Right now I have embedded on the page 4 spreadsheets, which show the following:

1. Number of papers so far in 2012 for 50 random terms compared to 2011 on the same day (uses the current date). With this we can see the change in % difference between the two years each day (it refreshes daily).
2. This one is a publicly editable spreadsheet that is the same setup as (1) but anyone can enter their own search terms.
3. The same idea as (2) but searchable by author.
4. The last uses the same 50 random terms as (1) and calculates the slope and R^2 values of the linear regression of the paper counts over 6 years (2005-2011).

This is very much in the early stages, but hopefully it will become more useful in the future. Let me know if you have any requests/ideas in the comments here or on that page.