Tuesday, December 27, 2011

The Potato Manifesto - Part 2/2: The Sweet Potato, Is It More Than Just the "En Vogue Tuber of the Year"?

Image 1: A mixer like this would be one of the best choices to turn your healthy low-GI sweet (or regular) potato into a high GI "nightmare".
If you've read yesterday's first part of the Potato Manifesto, you should by now be aware that the common notion of the pro-diabetic high-glycemic regular potato is another of the numerous black-or-white nutrition myths that do not become right, no matter how many bloggers and forum posters reiterate them. In today's second part of the series I will try to elucidate, whether the sweet potato, of which I would venture to say that she is the "en vogue tuber of the year", is not still the "safer starch alternative". So bear with me while I am using my scientific peeling knife to check whether there is a bitter truth hidden beneath skin of the sweet potatoes ;-)

Come on sweety, show me what's beneath your skin!

Now, if we take a look at the literature, a review of the glycemic index of 33 commonly available foodstuffs from the 1990s lists sweet potatoes with a glycemic index of 49 as #17 (with #1 peanuts and a GI of 11 and #32 cornflakes and a GI of 83 as the "extremes"; Nishimune. 1991). It is "outperformed by spaghetti (GI 47) and closely followed by Buckwheat, Yam (both GI 51) and you guessed it the good old "regular" potato, to which Takahiro Nishirnune and his colleagues assign a GI of 54, which - as you should know from yesterday's installment of the Potato Manifesto is nothing but an average on a scale that ranges from 10 to 110!
Figure 1: Difference in plasma glucose concentrations (expressed as the increase due to the "high GI" normal potato vs. the "low GI" sweet potato) in response to the ingestion of 50g of carbohydrates in form of bush potato or regular potato in 7 Aborigines and 7 Caucasians (data adapted Thorburn. 1986)
A brief note on the fallacy of common arguments for "traditional diets" (=sweet potato diets): I recently started listening to archived episodes of Jimmy Moore's Living La Vida Low Carb show and encountered time and again the 99% nonsensical argumentation that the XYZ thrived on whatever diet and that by just mimicking their diet(s), we would thrive, as well. To me this is like saying: "Look, the Panda bears thrive on a 100% bamboo diet! So, the ice-bears in our zoo should get along pretty well, if we stopped feeding them meat." Now, this is obviously a provocative comparison, but if I look at myself and then look at an Aborigine, the difference may not be as significant as the one between panda and ice-bear and yet, according to the results of a 1986 study by Anne W. Thorburn et al. (Thorburn. 1986) our glucose responses to the ingestion of 50g of carbohydrates from either the sweet-potato'esque bush potato and a "normal" western potato would be completely different (cf. figure 1). While the poor Aborigine would encounter the typical blood sugar ups and downs that are so characteristic for the ingestion of "high GI" carbs, my body would probably not care whether I eat the 100% paleo bush potato or its readily available, cheap cousin. The futility of common "neo-paleolithic" reasoning aside, this example also shows that genetic individualities are another major determinant of the glycemic index.
Three yeas later, in 1994, Thomas Wolever and his colleagues from the University of Toronto tested the glucose response of diabetic patient to 102 complex carbohydrate foods (Wolever. 1994). "Complex", in this case, indicates that they tested combinations of foods like they could appear on someone's dinner table and lo and behold, the "low GI" sweet potato took a close second to the "high GI" boiled white potato, when the both were boiled and served to 16 diabetic patients along with tomatoes. The 1 GI point difference between sweet and white potato (60 vs. 59) is yet well within the statistical margin.

Identical glycemic index, but differing free sugar compositions

Although, I hope that I should by now have convinced you that the glucose response is no convincing argument in favor of the sweet potato. I am even convinced that, if people abused their sweet potatoes in similar ways as their "regular" potatoes, i.e. mashed, fried, pureed, powdered and instantized it, we would soon see the (contemporarily) sexier sister of the regular potato on the list of "foods to avoid if you want to lose wait or just live a long and healthy life". Nevertheless, taste and name of the sweet potato, leave no doubt that there must be a sugary difference between her and her white brethren.
Figure 2: Free sugar composition (in mg/g) of 16 "regular" and 3 sweet potatoes cultivars (comparison based on data from Zhu. 2011 and Dincer. 2011)
And in fact, as my juxtaposition of data from a 2010 study on the relation between amino acid and sugar content of commercially available regular potatoes and its effect on the formation of acrylamite during heating (Zhu. 2011) and data from the previously mentioned (cf. Part 1) study by Cuneyt Dincer and his (or her) colleagues from the Akdeniz University in Ankara, Turkey (Dincer. 2011) shows (cf. figure 2), there is a non-negligable difference in the amount and composition of "free sugars" (vs. sugars in the form of starch). Although I am not 100% about the exact nature of the exotic Chinese potatoes, it is quite obvious that, despite large varieties in the absolute and relative sugar content of "regular" potatoes, the "classic" potato has a lower absolute free sugar content and contains more fructose and glucose than her sweet sister, 90% of the free sugar content of which is plain table sugar (=sucrose = glucose + fructose).
Figure 3: Mean free sugar composition (in mg/g) of fresh "regular" and sweet potatoes (comparison based on data from Zhu. 2011 and Dincer. 2011)
This "sugary" pattern does yet broaden, when you expose the sweet potato to heat in the form of either boiling or baking (cf. figure 4). In that, the cultivar-dependent slight reductions in glucose and fructose carry little weight compared to the sudden appearance of significant amounts of maltose, which had been beyond the detection before the sweet potatoes were boiled or baked (cf. figure 4).
Figure 4: Free sugar content (mg/g dry weight) of fresh and cooked sweet potatoes (data adapted from Dincer. 2011)
From a chemical perspective, the latter is not really surprising and its formation in response to heat treatment has been reported as early as 1923, when H.C. Gore published a paper in the Journal of Industrial and Engineering Chemistry (Gore. 1923), in which he makes the important observation that the formation of sugars from starch takes place within the first minutes of heating:
The  raw  potatoes  contained  0.34  per cent of  reducing  sugar calculated as maltose, the  cubes  cooked for  5  minutes in boiling water contained 8.32 per cent, and the cubes cooked for 1  hour,  7.82 per  cent.  Thus,  no sugar formation  occurred  at the  boiling point.
In view of this immediate heat-induced increase in the high GI free sugars, maltose (maltose has a ~10% higher GI than glucose), it should not surprise you that a study that was conducted by Jamaican scientists (Bahado-Singh. 2011), only a few months ago, found statistically significant increases in glycemic index and the area under the incremental glucose curve after boiling, frying, baking or roasting 10 sweet potato cultivars, which are commonly consumed in Jamaica (cf. figure 5):
Figure 5: Glycemic indices and glucose AUC of 10 common Jamaican sweet potato cultivars after boiling, frying, baking and roasting (data adapted from Bahado-Singh. 2011)
If we disregard the differences between the 10 cultivars and focus on the overall pattern, it is quite obvious that on the "less to worst thing you can do with your sweet potatoes if you are concerned about GI"-scale boiling is at the most benign end of the continuum, while baking and roasting compete for the position at the other end - a pattern, those of you who have already read Part 1 of the Potato Manifesto should be familiar with: Contrary to common believe new boiled (regular / sweet) potatoes are in fact a low GI food. In our kitchen (and even more so in the industrial processing machinery of the food industry), where the harmless tubers are mashed, pureed, dried, instantized, fried, baked, roasted etc., both, regular, as well as sweet potatoes do yet acquire at least some resemblance to the "pro-diabetic frankenfood", as which they have gotten labeled within the blogosphere, all along.
Figure 5: Comparison of regular and sweet potato glycemic indices after boiling, frying, baking and roasting.
If we now take a final look at the comparison of the average GI (figure 6 is based on data from both installments of the Potato Manifesto) of regular and sweet potatoes after boiling, frying, baking and roasting, it seems as if the main differences were that on average the regular potato has a slightly higher baseline GI (I do not need to remind you that according to Soh. 1999 boiled Desiree potatoes have a GI of 10!, do I?) and tends to be more resistant to heat-induced deteriorations in her starch / free sugar composition.

So, are boiled sweet and baked and roasted regular potatoes the way to go?

In view of the fact that probably 99% of the average and maybe 70% of the health conscious consumers don't even know the name of the potato cultivar they are using, let alone their age, storage temperature, the amount of phosphate in the soil on which it was grown, and all the other countless variables that have an impact on the "basal" GI of a potato, I honestly doubt that switching from one of the waxier, new regular potatoes (low basal GI) to a random (I mean, how many sweet potato cultivars are available at your grocery store?) sweet potato cultivar, would make you healthier, help you lose weight or offer any other GI-related benefits.
Image 2: Tapioca, another purpoted "health food" with a surprisingly high glycemic index GI of 84
Just as an aside, the Wolever study also examined the glucose response to another of the funky foods that are resurfacing on the Internet as "health foods", these days: Tapioca - also known as cassava, manioc, aipim, bitter-cassava, boba, mandioca, macaxeira, manioca, tapioca plant, yuca. With a GI of 84 it easily outperforms, corn chips (GI 76) and even waffles (GI 79).

Note: Wherever this was necessary I converted GIs that were given with white bread as a reference to the glucose as a reference.
Whether you could benefit from the (again, on average and in this case specifically referring to the white variety of the regular potato) higher beta-carotene and vitamin C content of sweet potatoes would depend on the rest of your diet. If the latter is "99% paleo" ;-), the likelihood that you would not get enough of these antioxidants is close to zero and it should thusly not really make a difference whether you go for the "paleo-" or the "neolithic" potato, as long as you don't (over-)process her into another of the innumerable frankenfoods the average member of the neolithic convenience society is so fond of.