Monday, July 24, 2017

Protein-Sugar Interactions: Will a Coke/DietCoke Turn Your Lean Steak into a Cheeseburger - Metabolically Speaking?

Does a Coke (diet or regular) really ruin the metabolic benefits of high protein meals? Decrease fat oxidation? Increase fat storage? Boost your appetite?
In her recently published blog about the study, Dr. Shanon Casperson writes acknowledges the "beneficial effects of protein-rich diets", its beneficial effect on satiety, its ability to decrease both prospective and real-world food intake, and its beneficial effects on human metabolism. With all the things we know about protein, it is yet quite interesting that we don't know "what happens when we drink a sugar-sweetened beverage with our steak dinner?" (see BMC blog).

The former is the research-question of a recently published study in the OpenAccess Journal BMC Nutrition that was conducted by - you guessed it - Casperson et al. (2017).
Learn more about fructose at the SuppVersity

Bad Fructose not so Bad, After All! Learn its Benefits.

Fructose From Fruit is NOT the Problem

US Fructose Consumption Did not increase!

An Apple A Day, Keeps... & More (Guestpost)

Fructose is Not Worse Than Sugar

The Obesogenic Fructose Fat Connection
As the authors point out, there's a plethora of evidence to prove that the substitution of one macronutrient, particularly protein, for another can "markedly affect both sides of the energy balance equation" (Casperson 2017).
Figure 1: There's plenty of data that confirms the metabolic and satiety benefits of replacing carbs w/ protein.
Since there's #1, human studies to confirm that increasing dietary protein while maintaining energy intake produces a greater and more prolonged thermic effect and greater total energy expenditure (Westerterp-Plantenga 2009); and there's #2, reliable evidence from Labayen et al (2004) that dietary protein intake potentially increases fat oxidation by up to 50%, protein has the (unique) ability to affect both sides of the energy balance equation in ways that will make it easier (not easy) for you to lose weight or maintain a healthy body weight. On the other hand, Casperson et al. rightly remind us that decreasing your protein intake will, probably because "protein intake [is tightly] regulated" stimulate "an increase in energy intake in an attempt to maintain a constant absolute intake of dietary protein" (Casperson 2017):
  • A 1.5%E decrease in dietary protein intake increases energy intake from carbohydrates and fats by 14%, perhaps in an attempt to increase protein intake from less protein-rich food sources (Simpson 2005). 
  • In a 4-day inpatient ad libitum crossover feeding trial, a 5%E decrease in dietary protein intake produced a 12% increase in total energy intake (Gosby 2011). The authors calculated that this was equivalent to a 4.5 kJ increase in non-protein foods for every 1 kJ decrease in habitual protein intake. 
Both is in contrast to observations by Austin et al. (2011) who calculated that a 1% increase in dietary protein intake  (note: that's 1% of total energy intake, i.e. ~5g protein assuming you ingest 2000kcal per day) corresponds to a 130 – 226 kJ decrease in daily energy intake dependent upon weight status and macronutrient substitution (Austin 2011).  This does also mean that "the American diet towards greater carbohydrate intake and reduced dietary protein, may explain the increase in total energy intake over the last 50 years" (Casperson 2011)... but that's a topic for a different article. After all, the question is...

Will having that (diet)coke turn your steak into a burger...? As far as its metabolic effects go?

Casperson et al. (2017) hypothesized that "compared to non-nutritive-sweetened beverage (NNSB) consumption, consuming a [sugar-sweetened beverage] (SSB) with a meal will increase appetite and diet-induced thermogenesis (DIT) independent of dietary protein". They also believed that the "consumption of a[n] SSB will reduce postprandial fat oxidation and that this effect will be greater when consumed with a usual (15%E) protein meal compared to a higher (30%E) protein meal" (Casperson 2017).
Table 1: Overview of the meal composition. For the individual tests, all meals had the same macronutrient content, i.e. 83, 23, and 20g in the low protein condition and 49, 45 and 20g in the high protein condition (Casperson 2017).
To invalidate this hypothesis (for the geeks: that's how Popper would have phrased it ;-), they conducted a closely controlled study that involved a 12-by-10-foot metabolic chamber in which the participants, 27 (initially 34) normal-weight healthy adults, spent 24 hours on separate occasions:
  • low protein condition: 15% of protein in all their meals, with either sugar-sweetened or artificially sweetened drinks accompanying their breakfast and lunch, respectively
  • high protein condition: 30% of protein in all their meals, with either sugar-sweetened or artificially sweetened drinks accompanying their breakfast and lunch, respectively
  • the subjects were randomly assigned to do either the high or low protein day first
  • the assignment to sugar- vs. non-nutritively sweetened drinks was counterbalanced across participants (i.e. if someone got NNSB with breakfast he/she got SSB w/ lunch).
  • for all testing procedures, the volunteers ate the exact same foods (not in the same ratios, though) throughout the day
  • subjects were tested after a ≥ 12 h overnight fast and had to keep a 3-day food log to account for possible dietary confounders in the days leading up to the test
After each meal, the scientists asked the participant about their hunger and desire to eat certain types of foods. The analysis of the corresponding questionnaires yielded the expected results:
Figure 2: Appetite sensation scores. Subjective appetite sensations of hunger (a), prospective food intake (b), fullness (c), and satiety (d) after meals containing 15%E or 30%E protein with a sugar-sweetened beverage (SSB) or a non-nutritive-sweetened beverage (NNSB) are presented as area under the curve (AUC | Casperson 2017).
As the data in Figure 2 goes to show you, "males reported feeling hungrier and that they could eat more food. Conversely, females reported greater feelings of fullness and satiety" (Casperson 2017).

More importantly and, as previously pointed out, just in line with the scientists' expectations, hunger and satiety AUC were lower, respectively higher after consuming a meal of 30%E protein. There was no significant main effect of beverage nor were there any sex, protein level or beverage type interactions. Items with similar letters are not significantly different.
What's the effect of sucralose? Yes, the lack of a water control is a definite weakness of the study at hand. Since there's no such "control" group, it's impossible to say whether or not drinking water instead of a non-nutritively sweetened beverage would have triggered a different metabolic response. Previous experimental data about non-nutritional sweeteners does yet suggest that there's no such effect or that the effects are practically irrelevant.
The scientists also investigated possible influences of the macronutrient composition on the subjects' desire to eat foods with a specific taste. After all, common wisdom tells us that high protein diets may increase your desire to binge on sweets/high-CHO foods. In the study at hand, the desire to eat something...
  • sweet was yet not affected by sex, protein amount or beverage type, on the other hand...
  • savory (p = 0.0011), salty (p < 0.0001) and fatty (p = 0.0188) was sign. reduced by the high(er) protein meals,  
Interestingly enough, the protein content of the meal also modified the appetitive response in the NNSB condition, with the combination of high protein (30%E) + non-nutritive sweetened beverage (the NNSB used was sucralose | all about sucralose) producing significant increases in the desire to eat something savory and salty in both, men and women. The effect size, however, is small and the practical relevance of this finding thus questionable.
Figure 3: If we focus on the effect of having a sugar- vs. non-nutritively sweetened drink with breakfast and lunch, we have to record an increase in energy expenditure, intake, balance and the contribution of CHOs in the postprandial phase and a concomitant decrease in thermogenesis, fat oxidation and protein oxidation (based on Casperson 2017)
That's in contrast to the observations the scientists made when they analyzed the effects on substrate oxidation and energy metabolism (summary see Figure 3):
  • Substrate oxidation: Main effect of SSB - Fat oxidation (-8%), carbohydrate oxidation (+15%), protein oxidation (-0.6%) -- As expected, consuming a SSB with a meal suppressed fat oxidation compared to NNSB consumption (135 ± 45 g/day and 145 ± 46 g/day, respectively | that's approximately 7%). This translates to 7.2 ±11 g and 12.6 ± 11 g with the addition of a SSB to a meal - depending on whether it was consumed with 15% or 30% protein, respectively (adjusting for lean body mass didn't change this result fundamentally - after adjustment for lean body mass the SSB reduced the fat oxidation by approximately 8%) .

    In contrast to previous studies, there was no significant main effect of protein amount and there was no significant interactions between sex, protein amount, or beverage type on postprandial fat oxidation. In other words: (a) If everything else was identical, simply eating more protein didn't increase the amount of fat the subjects burned after the meal; and (b) while the beverage type made a difference, this difference did not differ significantly with either sex or the amount of protein - meaning: SSBs lower the post-prandial energy expenditure in a man who consumed a high protein diet just as (statistically) significantly as they do in a women on a low protein diet ... that's at least in the short run.

    Fat oxidation was yet not the only thing that was reduced in response to the SSB. In fact, SSB lovers may be pleased to hear that the LBM-adjusted postprandial protein oxidation was likewise reduced in the SSB (0.338 ± 0.112 g/kg LBM/day) vs. NNSB (0.340 ± 0.123 g/kg LBM/day) condition. The relative difference of 0,6% is yet hilariously small enough to forget it right away. The latter cannot be said of the of the effect on postprandial carbohydrate oxidation. Carbohydrate oxidation was greater after SSB (271 ± 76 g/day) compared to NNSB (231 ± 79 g/day) consumption with a meal with the 15% difference mirroring the previously discussed 7-8% difference in fat oxidation (remember FAT = 8-9kcal, CHO = 4kcal).
  • Energy metabolism: Main effect of SSB - Postprandial energy expenditure (EE) +3.4% (80kcal), energy intake per meal +120kcal -- Aside from the expected sex difference (EE_men > EE_women) and the positive association of LBM and EE, the scientists also observed a +3.4% increase in energy expenditure (EE) with SSB vs. NNSB. The absolute difference between 2463 ± 395 kcal/day in the SSB compared to 2383 ± 384 kcal/day in the NNSB trial (80kcal) is yet smaller than the extra 120kcal from the 30g of sugar in the test beverage (66%). The overall effect of SSB vs. NNSB on the energy balance (EB) is thus positive - also because the diet-induced thermogenesis (DIT) decreased by 3% due to the sugary SSB (NNSB: 18% ± 7% vs. SSB 15% ± 5%). The previously established (cf. Westerp-Platenga 2009) positive effect of dietary protein on DIT didn't reach statistical significance, but a trend (p = 0.0690) was observed for the amount of dietary protein contained in the meal to increase DIT.
One of the things I like most about the study at hand is its truly thorough analysis of covariates. In that, Casperson et al. also tested potentially confounding effects of the subjects' habitual carbohydrate intake - without finding significant results. In contrast, a high habitual fat intake, on the other hand significantly and linearly related to the subjects' fat oxidation (F(3102) = 7.60, p < 0.0001, R2 = 0.18), a high protein intake predicted greater rates of protein oxidation (F(3104) = 4.89, p = 0.003, R2 = 0.12) and people with higher energy intakes showed an overall higher energy expenditure (F(3104) = 8.19, p < 0.0001, R2 = 0.19) - none of these correlations were modified by the consumption of NNSB or SSB in the study at hand. This means that, at least acutely, you can't turn a "fat burning machine" into "CHO guzzling couch potato" by feeding him/her SSBs ;-)
If we focus on the comparison between SSBs and NNSBs in the study at hand, the results confirm the previously discussed experimental evidence that refutes the assumption that the epidemiologically observed association between non-nutritive sweeteners and obesity is anything but corollary | more
So, what's the most important message here? For me personally, it's none of the changes in substrate oxidation. Those were expected and - albeit by the means of less thorough experimental means - previously established. For me, it's the discrepancy between the extra-energy (+120kcal) from the SSB and the corresponding increase in energy expenditure (+80kcal) the author also highlights in her previously referenced blog post, as it aligns beautifully with the recently discussed results of experimental studies on artificial sweeteners.

On a related note, it's worth pointing out that the question from the headline can be answered in the affirmative only for regular, sugar-sweetened, not for diet coke.

Ok, there were no cheeseburgers and stakes in the study, but you would probably observe similar trends for both substrate oxidation and energy balance as they were observed in the study at hand - with a steak + coke producing similar trends towards decreased fat oxidation, increased CHO oxidation and energy balance compared to steak + diet coke, as you'd observe them when eating a cheeseburger instead of your lean steak. The claim that the study at hand would, as Medscape titles, prove that "Sugary Drinks With Protein 'Help Store Body Fat'" is complete nonsense (fat storage wasn't even measured, by the way). After all, neither steak + coke, nor cheeseburger are going to make you store any fat if you ain't in a caloric surplus. | Comment on Facebook!
References:
  • Austin, Gregory L., Lorraine G. Ogden, and James O. Hill. "Trends in carbohydrate, fat, and protein intakes and association with energy intake in normal-weight, overweight, and obese individuals: 1971–2006." The American journal of clinical nutrition 93.4 (2011): 836-843.
  • Casperson, Hall and Roemmich. "Postprandial energy metabolism and substrate oxidation in response to the inclusion of a sugar- or non-nutritive sweetened beverage with meals differing in protein content." Nutrition BMC  (2017): 3:49.
  • Gosby, Alison K., et al. "Testing protein leverage in lean humans: a randomised controlled experimental study." PloS one 6.10 (2011): e25929.
  • Simpson, S. J., and D. Raubenheimer. "Obesity: the protein leverage hypothesis." obesity reviews 6.2 (2005): 133-142.
  • Westerterp-Plantenga, M. S., et al. "Dietary protein, weight loss, and weight maintenance." Annual review of nutrition 29 (2009): 21-41.