1 July 2006

Food for Thought

Slow Carb Not Low Carb
In May, ‘News Brief’ reported on a woman who was hospitalised for life-threatening ketoacidosis after following the Atkins diet. We then listed the reasons why we advocated a low GI diet rather than a low carbohydrate diet. Some of our readers admonished us for our stance. Here are their comments and our responses.

  1. Opposers of low carb regimes tend to stick to criticism that are really directed to the so called ‘induction phase’ which lasts only two weeks. After that, dieters on Atkins can and should consume a great variety of green, leafy nutritious vegetables, among many others, like mushrooms, eggplant, peppers, broccoli, cauliflower etc. Atkins never encourages you to eat zero carbs. 20 grams for the first 2 weeks, and you raise those levels after that.
    Yes, that’s true, criticism is often aimed at the induction phase but it’s also true that the second phase is restricted in carbohydrates (around 50 grams a day) too, and adherents are encouraged to return to the induction phase (20 grams a day) if weight loss slows. Furthermore, because Atkins recognised that the diet was not nutritionally balanced, a vitamin and mineral supplement program is compulsory.

    The study by Yancy et al (Annals of Internal Medicine 2004: 140; 769) compared an Atkins diet with a prudent (low fat but high GI) diet in 120 overweight volunteers. Those following the Atkins diet lost twice as much weight in 24 weeks but read the small print. Specifically, adverse effects occurred more frequently in the low-carbohydrate diet group than in the low-fat diet group, including constipation (68% vs. 35%; P < p =" 0.03)," p =" 0.02)," p =" 0.01)," p =" 0.006)." style="font-style: italic;">Annals of Internal Medicine 2004; 140: 778), two persons on the low carbohydrate diet died, and a third was hospitalised. No such adverse events were recorded in those following the high carbohydrate diet.

  2. There might have been some underlying health reasons for the ketoacidosis you mention.
    The case of a woman who was hospitalised for life-threatening ketoacidosis was written up in The Lancet (Chen TY, Smith W, Rosenstock JL, Lessnau KD The Lancet – Vol. 367, Issue 9514, 18 March 2006, Page 958). The precipitating factor, whatever her individual vulnerability, was strict adherence to a low carbohydrate diet. Fortunately, most people don’t or can’t adhere strictly to the Atkins diet.

  3. No diet is really recommended for ill or pregnant people.
    Pregnant women have to eat and the diet they eat affects their baby’s development. We use the word ‘diet’ in the sense of ‘eating plan’ (as defined by most dictionaries), not restricted energy intake. We agree that it’s not a good idea to restrict energy intake during pregnancy (one reason being the adverse effect of ketones on fetal development). But it’s vitally important that women eat a healthy diet throughout their pregnancy and throughout their reproductive years – because not every pregnancy is planned. A low GI diet fits the bill perfectly, a low carbohydrate diet doesn’t (and you appear to agree). If a diet’s not good for a developing fetus, why would it be good for anyone else?

  4. It is not true that consuming bad fats is unavoidable. You can and should avoid them and stay within Atkins.
    If carbohydrates occupy only 10% of your energy intake (i.e. you eat about 50 grams of carbohydrates per day), then the other 90% of energy must come from a mix of protein and fat. The upper limit on protein intake by humans is around 40% of their calories (kilojoules) because of limits on the liver’s capacity to produce urea. By a process of deduction, that means more than 50% (and more often 60%) of energy comes from fat. Even if you ate a perfectly healthy balanced diet with a P:M:S: ratio of 1:1:1 (polyunsaturated, monounsaturated, saturated fat), you’d be consuming about 20% of your calories as saturated fat. The recommendation is less than 10%.

  5. I love carbs, I could be happy with a GI diet, and I will resort to it as soon as I reach my goal weight, but losing weight with GI diets is complicated because as soon as you surpass the level of carbs you can consume, you stop losing weight and start gaining.
    That was the claim Atkins made but he had no scientific evidence to back that. To our knowledge, there’s still no evidence. Indeed, recent studies suggest that people find it very hard to stick to a diet with so little carbohydrate (too much discipline is needed) and eventually re-gain the weight they lost. Is there any point in losing weight and then re-gaining it? Wouldn’t it be better to align food habits with something that’s not only healthy, helps you lose weight and keep it off for good?

  6. As soon as I can start incorporating more carbs into my diet, they will be of the low glycemic load kind.
    That’s good. Why not cut to the chase early?

  7. There should really be no quarrel between low glycemic and Atkins because they are fundamentally the same. The diets you should be strongly opposing are the low caloric and the low fat diets.
    No, that’s incorrect. Atkins wants to ditch carbs. Low GI diets can be moderately or even very high in carbs, but of the low GI kind. If Atkins had known what we know now, he would have seen ways to lower insulin levels without cutting the carbs. He recognised that high insulin levels interfered with weight control, but he did not have sufficient knowledge to work out the best way to lower insulin. Remember he had no training in nutrition.

  8. The truth is that Atkins works. Once you get to your equilibrium level, then you can do low GL and stay in a narrow weight band.
    Yes, we agree that Atkins works in the short term. But not the long term. So what’s the point in following such a diet?
James Krieger, M.S., M.S. 20/20 Lifestyles Research Associate, PRO Club (http://www.proclub.com) and Editor, Journal of Pure Power (http://www.jopp.us) makes the point that: ‘You lose fat, water, and muscle’ is true of ANY diet. My recent meta-regression, published in the February 2006 issue of AJCN, shows quite clearly that fat loss is significantly higher on ketogenic diets, even after control for energy intake. Fat-free mass loss was also greater, but not by a large amount.

krieger
James Krieger

‘There is no apparent metabolic advantage associated with ketosis during dieting’ report researchers from the Department of Nutrition, Arizona State University, Inflammation Research Foundation, Marblehead and Conscious Cuisine, Scottsdale in the May issue of the American Journal of Clinical Nutrition. In summarising their findings, Prof. Carol Johnson says: ‘In the current study, the ketogenic low-carb diet did not offer any significant metabolic advantage over the nonketogenic low-carb diet. Both diets were effective at reducing total body mass and insulin resistance, but, because blood ketones were directly related to LDL-cholesterol concentrations and because inflammatory risk was elevated with adherence to the ketogenic diet, severe restrictions in dietary carbohydrate are not warranted. Furthermore, the nonketogenic low-carb diet was associated with feelings of high energy and a more favorable mood profile than was the ketogenic low-carb diet.’ They recommend anyone wanting to follow a low-carb diet to choose low-fat meats and dairy products, and eat 8–9 servings of fruit and vegetables and 100–125 grams of carbohydrate a day.

7 comments:

Anonymous said...

Hi. you seem to have missed my question in your last issue so here it is again.

Searching in your GI database amongst rice types the following seems to give the best value for GI/GL:

Bangladeshi rice, variety BR16, pressure parboiled (27% amylose)

Where in Sydney can I find this product?

Thanks for your help.

Anonymous said...

Hi
I also have the same question as Jane but in Melbourne. Where in Melbourne can I find Bangladeshi rice variety BR16?

Sat

James Krieger said...

My comments are in bold

There is no apparent metabolic advantage associated with ketosis during dieting’ report researchers from the Department of Nutrition, Arizona State University, Inflammation Research Foundation, Marblehead and Conscious Cuisine, Scottsdale in the May issue of the American Journal of Clinical Nutrition. In summarising their findings, Prof. Carol Johnson says: ‘In the current study, the ketogenic low-carb diet did not offer any significant metabolic advantage over the nonketogenic low-carb diet. Both diets were effective at reducing total body mass and insulin resistance, "

I would like to note that the final sample size in this study was only 19, meaning there were only 10 subjects in one group and 9 in the other. As weight loss is highly variable among subjects, larger sample sizes are necessary to detect differences, particularly in a study that lasted only 6 weeks.

To illustrate my point, my meta-regression detected a 1.74 kg greater loss of body mass in low-carb diets. Let's see if Johnston et al. had the statistical power to detect such a difference. Let's use the smallest standard deviation from Johnston et al. for body mass. The low-carb group had a SE of 5.7. With a sample size of 9, that makes a SD of 17.1.

In this study, to detect 1.74 kg difference in loss of body mass, at 80% power and a SD of 17.1, you would need a sample size of over 1,500 subjects!!!!!!

You have to be very careful of Type II errors when drawing conclusions from such small studies.

I would like to note that at least 3 studies, with much larger sample sizes, have shown "metabolic advantages" to low-carb diets. I refer you to Brehm et al (2003), Yancy et al (2004), and Brehm et al (2005).

This is also where the usefulness of meta-analysis comes in. Meta-analysis overcomes the limitation of small sample sizes by aggregating results from a large number of studies. My meta-analysis combined the results of 87 studies, 165 intervention groups, and over 1,500 subjects, giving it the statistical power to detect a "metabolic advantage" that a low-carb diet provides. These findings were quite robust to sensitivity analysis as well.

Given all of these factors, I would argue that the lack of differences in this study are most likely due to Type II error.


"but, because blood ketones were directly related to LDL-cholesterol concentrations "

There are problems with this. First, Pearson correlations were used; Pearson correlations are highly sensitive to violations of assumptions such as normality (normality is less likely with such a small sample). Thus, there's a good chance that some correlations could be spurious. The authors did not report whether their data met the assumptions of parametric statistics. Spearman-rank correlations are generally more appropriate in such situations.

Second, correlation does not equal causation. It should be noted that there was no diet by time interaction in this study, indicating the diets had similar effects on LDL. Again, though, we are limited in detecting a true difference due to the small sample.

If one is going to present this correlation as evidence against a low-carb diet, then, to be fair, you should also present the negative correlation between blood ketones and insulin resistance. In other words, people in this study who had higher blood ketones had less insulin resistance.

It should also be noted that the low-carb diet certainly did not have a detrimental effect on LDL. There was no change in LDL over the 6 weeks in the low-carb diet group.


"and because inflammatory risk was elevated with adherence to the ketogenic diet, "

I would like to note that only the AA:EPA ratio was affected by the ketogenic diet. CRP (another marker of inflammatory risk) was not affected. In fact, while there were no statistically significant changes in CRP in either group, the mean value of CRP tended to go down in the ketogenic diet while it tended to go up in the other diet. Now, I am not saying that a ketogenic diet has more favorable effects on CRP than a non-ketogenic diet. However, I am saying that a more complete picture of the results of this study need to be presented.

I would also like to note that other studies have shown inflammatory markers to decrease on ketogenic diets. I refer you to Wood et al (2006) and Sharman & Volek (2004).


"severe restrictions in dietary carbohydrate are not warranted. "

I don't think such a conclusion should be drawn from this study.

"Furthermore, the nonketogenic low-carb diet was associated with feelings of high energy and a more favorable mood profile than was the ketogenic low-carb diet.’ "

This conclusion is questionable. The significant group-by-time interaction is misleading. If you look at each week, only in week 5 was vigor higher in the nonketogenic group. I would also like to point out that POMS data tend not to be normally distributed. My previous criticism on the use of parametric statistics on non-normally distributed data stands here as well.

I again feel that the GI Newsletter is not giving a complete & fair representation of the data on low-carbohydrate diets. As I have mentioned, when one examines the details of this study (and of other research), one comes away with a somewhat different conclusion than what was implied in the paragraph about it. While I do not advocate low-carbohydrate diets for all people, there is sufficient scientific data to indicate that they can be a useful treatment option for [i]some[i] obese individuals, and should not be completely ruled out.


James Krieger, M.S., M.S., ACSM-HFI
20/20 Lifestyles Research Associate
PRO Sports Club
http://www.proclub.com
Editor, Journal of Pure Power
http://www.jopp.us

Melissa S. Green said...

I am obviously not up on all the science re: low carb or re: GI, but on a personal level I'll say that a combination of those approaches work best for me: low GI carbs, moderate intake (100-125 grams/day).

I've cautiously come to agree that sometimes, for some people, low-carb diets are helpful. Even more specifically: high carb diets, even when the carbs are low GI, are disastrous for Type 2 diabetics, at least based on anectodal evidence from the people I'm talking with.

I talk a lot with people who are diabetic who have benefitted extraordinarily from following the advice of Dr. Richard Bernstein (Dr. Bernstein's Diabetes Solution), which differs in some notable ways from Atkins' low carb approach. The problem I have with Bernstein is that he generalizes from his own experience as a Type 1 diabetic & what works for him to an insistence that everyone must be as tightly controlled in their carb intake. He flatly states that his readers should ignore GI, a stance with which I very much disagree.

I'm dealing with prediabetes (& hence diabetes prevention), but I also have a history with depression. Right now I'm eating about 100 to 125 grams of carb per day, which I regard not as "low carb" but as "moderate carb." I find that if I go over that, my blood glucose goes a little too high, even though I'm eating low GI; if I go under that, I have problems with depression. My intake has some room for change depending on my level of physical activity.

From conversations with other members of the discussion list lowglycemiceating@yahoogroups.com, I'd say there's a lot of anecdotal evidence to support moderate carb for people with Type 2 diabetes or insulin resistance, even when eating low GI. Many members report having to watch portion sizes (i.e., glycemic load) as much as GI, & some absolutely avoid any kind of rice or pasta -- even basmati rice or whole grain pasta.

Anonymous said...

Re Bangladeshi rice variety BR16: The GI test results for this product were published in the European Journal of Clinical Nutrition 2000; 54: 380-5. The GI Group is currently checking to see if this variety of rice is widely available commercially and under what brand names and where. We will post the results of our research as soon as they come to hand.

Anonymous said...

Thanks - I shall be very interested as will Grace and Sat I'm sure. If the GI/GL given in your database for this type of rice is accurate then it will be an excellent choice, particularly for diabetics.

I shall watch this space.

Jane

Anonymous said...

Re rice: Some new varieties of rice are being tested at present and we should have results to publish in September or October GI News.