Gabriel Guzman
07-12-2006, 01:42 PM
For some time now, I've been wanting to write about the article "Effects of Low-Carbohydrate vs Low-Fat Diets on Weight Loss and Cardiovascular Risk Factors - A Meta-analysis of Randomized Controlled Trials" (Arch Intern Med. 2006;166:285-293;I attached the pdf to this post).
The words 'meta-analysis' and 'randomized controlled trials' were what interested me to read the article. First because meta-analysis are sometimes anything but objective and it's hard to find a meta-analysis that doesn't advance a particular interest; second, because one would expect that beting a meta analysis of randomized controlled trials (of anything), the authors would do their homework and truly explore the available literature on the subject they write about. Okay... there was a third reason... I know the work of two of the co-authors and I know one of them personally so I'm always interested in following their research.
To my disappointment, even though it's a 9-pager with hard statistics to follow for the beginner, the number of references is no larger than my first published paper... A meta-analysis is supposed to cover a hefty number of references, which would reflect how deep the author went into the literature and how 'meta' is his analysis. Anyway, after giving them the benefit of the doubt, then you start reading what everybody in the field writes how the prevalence of obesity has increased during the past decades and how numerous diets have been proposed to promote weight loss but there is hardly any data from well-controlled trials about the most effective dietary approaches to achieve meaninful and long-term weight loss in overweight individuals (Incidentally...who else would need long-term weight loss if not overweight individuals?)
So, in the meta-analysis, the authors compared the effects of low-carbohydrate diets without energy restriction vs energy-restricted low-fat diets, on weight loss, blood pressure, and lipid values in randomized controlled trials with diet interventions of at least 6 months.
How did they decide which studies to compare? They used the Cochrane Collaboration Search Strategy, let's abbreviate that as CCSS (you can read more about it here (http://www.cochrane.org/docs/descrip.htm)). The Cochrane Collaboration is used to identify relevant studies for systematic reviews and one is supposed to get truly the relevant studies on the subject one is reviewing. In this case the 'relevance' is given by the comparison of low-carbohydrate/non energy restriction to low-fat/energy restrction diets. Similar to what you do when you 'google' a topic or search on MEDLINE. There are other search strategies and engines that are used in science that are not limited to material found in medical jounrals or databases. Anyway, the point is that the Cochrane Collaboration produces and disseminates systematic reviews of healthcare interventions and promotes the search for evidence in the form of clinical trials and other studies of interventions. By using this strategy, the authors make sure that only relevant studies will come up after selecting for pertinent parameters such as 'humans' (i.e. don't include studies in rodents or ther animals); duration (at least 6 months), etc. Besides the CCSS, the authors also relied on a number of other data bases, including MEDLINE, and GLOBAL HEALTH. We could say that they did their homework in searching for the relevant studies that could be appropriately compared.
In short, using their search strategy, the authors made the first cut and came up with 166 relevant publications suitable for retrieval. Then they apply a more stringent set of parameters to exclude studies on basis of title and abstract. By doing that, they exluded 138 our of the 166 they started with. Of the 28 remaining studies, they excluded 22 because of various reasons (follow up too short, cross over design, energy restriction, low-fat vs conventional, very low-fat vs low-fat, and other parameters that didn't fit into the relevance of the topic).
So they eneded up with 6 studies that could be compared without problem. However, one of the trials coalesced into other trials so in the very end, only 5 studies survived the scrutiny and the meta-analysis is then based on those five studies. A great example on how to choose your material to compare but I'm not sure it qualifies as a meta-analysis. Those words, however, are always welcome in a title to increase publication chances.
Before we discuss the final results, there is on more detail that could be good or bad, depending on your point of view. The first three and the last authors are affiliated to the Institute for Clinical Epidemiology, University Hospital Basel, in Basel Swittzerland. The fourth author is affiliated with the Division of Endocrinology and Metabolism of the same University Hospital. Why is this important? Because they may or not have experience with actual metabolic research (except the last author), even though they may know everything there is to know about epidemiology. The first two authors independently assessed the eligibility and quality of the trials included in the comparison. That's good, especially if they have no training in metabolic diseases, because they're basing their assessment in cold facts, and hopefully without bias. Nonetheless, here is where one rises the antena because epidemiology very rarely shows a causal relationship, so is important to keep that in mind when we read the conclusions of the analysis. Authors Yancy and Brehm are known in the literature of nutritional and metabolic studies comparing carbohydrate-controlled diets with standard nutritional approaches.
The 5 trials included in the analysis were compared based on mean difference in weight loss from base-line to 6 and 12 months of follow up between the two groups (main end point) and attrition rates on diets and mean differences in percentage change of body weight, systolic and diastolic blood pressure, blood lipid levels [HDL, LDL and TGs], fasting glucose, and insulin levels and quality of life as secondary end points. Question... would you choose to compare studies based on weight loss as your main objective or improvement in other clinical parameters such as blood glucose, insulin, blood pressure, etc?
Which trials were compared in this analysis?
Brehm BJ, Seeley RJ, Daniels SR, D’Alessio DA. A randomized trial comparing a very low carbohydrate diet and a calorie-restricted low fat diet on body weight and cardiovascular risk factors in healthy women. J Clin Endocrinol Metab. 2003; 88:1617-1623.
Foster GD, Wyatt HR, Hill JO, et al. A randomized trial of a low-carbohydrate diet for obesity. N Engl J Med. 2003;348:2082-2090.
Samaha FF, Iqbal N, Seshadri P, et al. A lowcarbohydrate as compared with a low-fat diet in severe obesity. N Engl J Med. 2003;348:2074-2081.
Yancy WS Jr, Olsen MK, Guyton JR, Bakst RP, Westman EC. A low-carbohydrate, ketogenic diet versus a low-fat diet to treat obesity and hyperlipidemia: a randomized, controlled trial. Ann Intern Med. 2004;140:769-777.
Dansinger ML, Gleason JA, Griffith JL, Selker HP, Schaefer EJ. Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: a randomized trial. JAMA. 2005;293:43-53.
(I can e-mail you these studies if you're interested in reading them)
So, what did they find after comparing the 5 published articles that fulfilled their selection criteria (my bold type)? Although I'm inserting a 'quote' mark, I've removed some of the abundant statistical values (which you can read in the paper if curious), to make the reading a bit clearer.
Weight loss
After 6 months, individuals assigned to low-carbohydrate diets had lost more weight than individuals randomized to low-fat diets[/B] (weighted mean difference, −3.3 kg; −5.3 to −1.4 kg
Attrition (completion rates on diets)
After 6 months, individuals randomized to low-carbohydrate diets were more likely to complete the trial than were individuals randomized to low-fat diets (156 [70%] of 222 individuals randomized to low-carbohydrate diets vs 129 [57%] of 225 individuals randomized to low-fat diets). After 12 months, this difference was no longer significant (84 [62%] of 135 individuals vs 72 [54%] of 134 individuals)
Any idea as to why there was no difference after 12 months? (hint: how truly 'low-carbohydrate' was the diet named as such, and how did the authors in the studies made sure people actually followed it?
Blood pressure
There was a trend toward lower systolic and diastolic blood pressure in individuals randomized to low-carbohydrate diets after 6 months. However, this trend was no longer detectable after 12 months.
Lipid values
Summary estimates of weighted mean differences at 6 months were not in favor of low-carbohydrate diets for total cholesterol values (8.9 mg/dL [0.23 mmol/L]), nor for LDL-C values (5.4 mg/dL [0.14 mmol/L]). At 12 months, the summary estimates of the weighted mean change for total cholesterol level (10.1 mg/dL [0.26 mmol/L]) and LDL-C level (7.7 mg/dL [0.20 mmol/L]) were basically unchanged when compared with the 6-month data.
Wasn't it at least curious to the authors that the values didn't get worse with time?
Contrarily, summary estimates of weighted mean differences [B]after 6 months in HDL-C and triglyceride values were in favor of low-carbohydrate diets (for HDL-C, 4.6 mg/dL [0.12 mmol/L] and for [B]triglyceride, −22.1 mg/dL [−0.25 mmol/L]). At 12 months, summary estimates of weighted mean differences for HDL-C and triglycerides, respectively, were as follows: 3.1 mg/dL (0.08 mmol/L) and −31.0 mg/dL (−0.35 mmol/L).
Remember that they were measuring 'change' so a positive number means 'increased' and a negative number means 'decreased' In other words, HDL didn't g down while triglycerides remained down after 12 months. So... there is no worsening trend in Total cholesterol or LDL and there was an improvement of HDL and Triglycerides... A discussion of these findings is warranted somewhere in the discussion section.
Glucose and Insulin values
This was a very tricky part to compare because of methodological differences between trials. Still, there is an important finding when the analysis focuses on a subgroup of diabetics:
In the subgroup of patients with diabetes in this trial (the Samaha study), hemoglobin A1c values changed more favorably in individuals on the low-carbohydrate diet than on the low-fat diet at 12 months (−0.7%±1.0% vs −0.1%±1.6%; P=.02; after adjustment for weight loss).
After these results, it's interesting and actually amazing that the authors start their discussion (comment) with:
There was no clear benefit of either diet when their effects on cardiovascular risk factors were examined. Changes in blood pressure were not different between the 2 groups. Whereas total and LDL-C levels decreased more in individuals randomized to low-fat diets, HDL-C and triglyceride values changed more favorably in individuals randomized to low-carbohydrate diets.
So let's recapitulate on the changes:
Those on the low-carb diets w/o energy restriction achieved more weight loss in 6 months than those in a low-fat, energy restricted diet. After a year, the change was insignificant.
Those on the low-carb diets without w/o energy restriction tended to stay on the diet longer than those on the low fat diet. After a year, there was no difference.
Those on the low-carb diets without w/o energy restriction tended to improve their blood pressure, even though there was no difference after 12 monhts. But it didn't worsen, dit it?
Those on the low-carb diets without w/o energy restriction , despite not improving their total and LDL cholesterol (in the eyes of the researchers), they did improve their HDL and TGs, changed that remained after 12 months.
Where is the support to the claim that 'a low-carbohydrate diet' didn't have a clear benefit on cardiovascular risk?
The mega-flaw of this comparison lies on ignoring how the dynamics of cholesterol works (more on that in a minute). Other flaws are inherent to the studies included in the analysis. First of all, none of them offer a clear-cut 'low-carbohydrate approach'. The studies start with certain amounts of carbs but then the participants were allowed to increase their carb intake after two weeks. Some of those studies don't even have proper counseling during the whole duration of the studies so the participants were left pretty much on their own and we all know that without accountability, compliance cannot be 100%. More over, the trials included in the analysis don't understand 'low-carb' in the same way. As expected, 'low-carb' is immediately associated with Atkins or South Beach, which only shows poor research into what is actually proposed by what they call 'popular low-carb diets'.
The authors of the meta-flawed analysis chose to comment on the strenghts and limitations of their work, not based on the actual and fundamental flaws of the trials included, from the physiological point of view that is, but instead they chose to pick on the quality of the studies.
The quality of the included trials was moderate. Whereas most trials used concealed treatment allocation, losses to follow-up were quite substantial. No trial reported blinded outcome assessment. Therefore, we were not able to conduct sensitivity analyses comparing trials with blinded outcome assessment vs trials without it, as originally planned. The absence of blinded outcome assessment is a flaw that potentially limits the validity of individual trials.
This is a good criticism on how to actually design studies that can be compared later with studies that explore different nutritional approaches but doesn't address the fact that in the trials included in the analysis, the low-carbohydrate diets scored better in the main and secondary end points the authors selected to compare.
Dropout rates were substantial. After 1 year of follow-up, between 31% and 48% of individuals randomized to low-carbohydrate diets, and between 37% and 50% of individuals randomized to low-fat diets, had dropped out of the trials.
This only shows that without proper counseling during the entire study, half of the participants will drop out for whatever reason and it has nothing to do with one diet or the other. Nonetheless, the trend of dropout was higher in the low-fat groups.
We conclude that low-carbohydrate diets appear to be at least as effective as low-fat diets in inducing weight loss for a duration of up to 1 year. Low-carbohydrate diets are associated with unfavorable changes in total cholesterol and LDL-C levels, but favorable changes in triglyceride and probably HDL-C values. In the absence of evidence that lowcarbohydrate diets reduce cardiovascular morbidity and mortality, such diets currently cannot be recommended for prevention of cardiovascular disease.
So what is the actual meta-flaw? The authors are thinking of cardiovascular risk as reflected by values of LDL. Basically, they're concluding that since the low-carbohydrate diets didn't reduce LDL values, then they didn't bring up any benefits. In the author's minds, there may be a paradox of some sort because eventhough total and LDL cholesterol levels didn't 'improve' (i.e. go down), HDL went up and TGs went down and remained down despite not improving their LDL (which of course is bad in the researcher's mind). Should they have stopped for a moment to reflect on this particular finding, they would summon their mighty search strategies just to see what's out there regarding the role of fat in 'the cholesterol issue', and the actual role of LDL particles in the whole dynamics of cholesterol homeostasis. That would also have led them to find that there is a relationship between triglycerides and HDL and they would have find out that in terms of cardiovascular risk, the relationship between TGs and HDL is far more indicative than the actual level of LDL. That would have opened a whole different can of worms for them because then their whole interpretation of the trials included in their analysis had to be done under a different light.
With the amount of publish information that establishes that it is the LDL profile, rather than the amount of LDL what matters in terms of assigning risk, it's almost unacceptable that they wouldn't at least offer a comment on that in their discussion. More over, none of the 5 trials included in the meta-analysis actually measured LDL profiles; LDL was calculated from the Friedewald formula, which has been shown to be flawed as well. This methodological shortcoming would be enough to exclude the trials from any comparison that looks at cardiovascular risk.
What about the changes in HDL and TGs? That alone should direct any discussion to the fact that cardiovascular risk improved in those who increased their HDL and lowered their TGs; those on a low-carbohydrate diet.
If anything, this meta-analysis (of 5 studies, mind you) shows exactly what aspects to consider when interventional studies are designed to test nutritional approaches. I learned from this analysis that attrition is something that can hamper a good study, thus counseling is of utmost important for the whole duration of the study. It is also clear that not everybody knows what exactly is 'low-carbohydrate' and even less know how to implement that in the long-term. If the quality of the studies is 'moderate' as the authors point out, then the conclusions cannot be so categoric. In other word, what gives anybody the authority to say 'this didn't work... although the source is more or less of good quality...'
At best, one can only conclude that even though the quality of the studies was moderate, the trend favors a low-carbohydrate diet to improve cardiovascular risk. Therefore, what we need is not more 'long-term' studies that use a low-carbohdyrate diet but better designed ones, that measure parameters that are truly informative (i.e. LDL profiles instead of LDL calculation from formulae), parameters that truly reflect cardiovascular risk.
I did learn a lot about how not to design a study but I don't think the authors learned much about the actual benefits of carbohydrate control and its benefits on cardiovascular risk.
My only question is if Yancy and Brehm actually took part in the writing and analysis in this work or if they were invited as co-authors because their trials were included. That little Adam, may be another story...
The words 'meta-analysis' and 'randomized controlled trials' were what interested me to read the article. First because meta-analysis are sometimes anything but objective and it's hard to find a meta-analysis that doesn't advance a particular interest; second, because one would expect that beting a meta analysis of randomized controlled trials (of anything), the authors would do their homework and truly explore the available literature on the subject they write about. Okay... there was a third reason... I know the work of two of the co-authors and I know one of them personally so I'm always interested in following their research.
To my disappointment, even though it's a 9-pager with hard statistics to follow for the beginner, the number of references is no larger than my first published paper... A meta-analysis is supposed to cover a hefty number of references, which would reflect how deep the author went into the literature and how 'meta' is his analysis. Anyway, after giving them the benefit of the doubt, then you start reading what everybody in the field writes how the prevalence of obesity has increased during the past decades and how numerous diets have been proposed to promote weight loss but there is hardly any data from well-controlled trials about the most effective dietary approaches to achieve meaninful and long-term weight loss in overweight individuals (Incidentally...who else would need long-term weight loss if not overweight individuals?)
So, in the meta-analysis, the authors compared the effects of low-carbohydrate diets without energy restriction vs energy-restricted low-fat diets, on weight loss, blood pressure, and lipid values in randomized controlled trials with diet interventions of at least 6 months.
How did they decide which studies to compare? They used the Cochrane Collaboration Search Strategy, let's abbreviate that as CCSS (you can read more about it here (http://www.cochrane.org/docs/descrip.htm)). The Cochrane Collaboration is used to identify relevant studies for systematic reviews and one is supposed to get truly the relevant studies on the subject one is reviewing. In this case the 'relevance' is given by the comparison of low-carbohydrate/non energy restriction to low-fat/energy restrction diets. Similar to what you do when you 'google' a topic or search on MEDLINE. There are other search strategies and engines that are used in science that are not limited to material found in medical jounrals or databases. Anyway, the point is that the Cochrane Collaboration produces and disseminates systematic reviews of healthcare interventions and promotes the search for evidence in the form of clinical trials and other studies of interventions. By using this strategy, the authors make sure that only relevant studies will come up after selecting for pertinent parameters such as 'humans' (i.e. don't include studies in rodents or ther animals); duration (at least 6 months), etc. Besides the CCSS, the authors also relied on a number of other data bases, including MEDLINE, and GLOBAL HEALTH. We could say that they did their homework in searching for the relevant studies that could be appropriately compared.
In short, using their search strategy, the authors made the first cut and came up with 166 relevant publications suitable for retrieval. Then they apply a more stringent set of parameters to exclude studies on basis of title and abstract. By doing that, they exluded 138 our of the 166 they started with. Of the 28 remaining studies, they excluded 22 because of various reasons (follow up too short, cross over design, energy restriction, low-fat vs conventional, very low-fat vs low-fat, and other parameters that didn't fit into the relevance of the topic).
So they eneded up with 6 studies that could be compared without problem. However, one of the trials coalesced into other trials so in the very end, only 5 studies survived the scrutiny and the meta-analysis is then based on those five studies. A great example on how to choose your material to compare but I'm not sure it qualifies as a meta-analysis. Those words, however, are always welcome in a title to increase publication chances.
Before we discuss the final results, there is on more detail that could be good or bad, depending on your point of view. The first three and the last authors are affiliated to the Institute for Clinical Epidemiology, University Hospital Basel, in Basel Swittzerland. The fourth author is affiliated with the Division of Endocrinology and Metabolism of the same University Hospital. Why is this important? Because they may or not have experience with actual metabolic research (except the last author), even though they may know everything there is to know about epidemiology. The first two authors independently assessed the eligibility and quality of the trials included in the comparison. That's good, especially if they have no training in metabolic diseases, because they're basing their assessment in cold facts, and hopefully without bias. Nonetheless, here is where one rises the antena because epidemiology very rarely shows a causal relationship, so is important to keep that in mind when we read the conclusions of the analysis. Authors Yancy and Brehm are known in the literature of nutritional and metabolic studies comparing carbohydrate-controlled diets with standard nutritional approaches.
The 5 trials included in the analysis were compared based on mean difference in weight loss from base-line to 6 and 12 months of follow up between the two groups (main end point) and attrition rates on diets and mean differences in percentage change of body weight, systolic and diastolic blood pressure, blood lipid levels [HDL, LDL and TGs], fasting glucose, and insulin levels and quality of life as secondary end points. Question... would you choose to compare studies based on weight loss as your main objective or improvement in other clinical parameters such as blood glucose, insulin, blood pressure, etc?
Which trials were compared in this analysis?
Brehm BJ, Seeley RJ, Daniels SR, D’Alessio DA. A randomized trial comparing a very low carbohydrate diet and a calorie-restricted low fat diet on body weight and cardiovascular risk factors in healthy women. J Clin Endocrinol Metab. 2003; 88:1617-1623.
Foster GD, Wyatt HR, Hill JO, et al. A randomized trial of a low-carbohydrate diet for obesity. N Engl J Med. 2003;348:2082-2090.
Samaha FF, Iqbal N, Seshadri P, et al. A lowcarbohydrate as compared with a low-fat diet in severe obesity. N Engl J Med. 2003;348:2074-2081.
Yancy WS Jr, Olsen MK, Guyton JR, Bakst RP, Westman EC. A low-carbohydrate, ketogenic diet versus a low-fat diet to treat obesity and hyperlipidemia: a randomized, controlled trial. Ann Intern Med. 2004;140:769-777.
Dansinger ML, Gleason JA, Griffith JL, Selker HP, Schaefer EJ. Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: a randomized trial. JAMA. 2005;293:43-53.
(I can e-mail you these studies if you're interested in reading them)
So, what did they find after comparing the 5 published articles that fulfilled their selection criteria (my bold type)? Although I'm inserting a 'quote' mark, I've removed some of the abundant statistical values (which you can read in the paper if curious), to make the reading a bit clearer.
Weight loss
After 6 months, individuals assigned to low-carbohydrate diets had lost more weight than individuals randomized to low-fat diets[/B] (weighted mean difference, −3.3 kg; −5.3 to −1.4 kg
Attrition (completion rates on diets)
After 6 months, individuals randomized to low-carbohydrate diets were more likely to complete the trial than were individuals randomized to low-fat diets (156 [70%] of 222 individuals randomized to low-carbohydrate diets vs 129 [57%] of 225 individuals randomized to low-fat diets). After 12 months, this difference was no longer significant (84 [62%] of 135 individuals vs 72 [54%] of 134 individuals)
Any idea as to why there was no difference after 12 months? (hint: how truly 'low-carbohydrate' was the diet named as such, and how did the authors in the studies made sure people actually followed it?
Blood pressure
There was a trend toward lower systolic and diastolic blood pressure in individuals randomized to low-carbohydrate diets after 6 months. However, this trend was no longer detectable after 12 months.
Lipid values
Summary estimates of weighted mean differences at 6 months were not in favor of low-carbohydrate diets for total cholesterol values (8.9 mg/dL [0.23 mmol/L]), nor for LDL-C values (5.4 mg/dL [0.14 mmol/L]). At 12 months, the summary estimates of the weighted mean change for total cholesterol level (10.1 mg/dL [0.26 mmol/L]) and LDL-C level (7.7 mg/dL [0.20 mmol/L]) were basically unchanged when compared with the 6-month data.
Wasn't it at least curious to the authors that the values didn't get worse with time?
Contrarily, summary estimates of weighted mean differences [B]after 6 months in HDL-C and triglyceride values were in favor of low-carbohydrate diets (for HDL-C, 4.6 mg/dL [0.12 mmol/L] and for [B]triglyceride, −22.1 mg/dL [−0.25 mmol/L]). At 12 months, summary estimates of weighted mean differences for HDL-C and triglycerides, respectively, were as follows: 3.1 mg/dL (0.08 mmol/L) and −31.0 mg/dL (−0.35 mmol/L).
Remember that they were measuring 'change' so a positive number means 'increased' and a negative number means 'decreased' In other words, HDL didn't g down while triglycerides remained down after 12 months. So... there is no worsening trend in Total cholesterol or LDL and there was an improvement of HDL and Triglycerides... A discussion of these findings is warranted somewhere in the discussion section.
Glucose and Insulin values
This was a very tricky part to compare because of methodological differences between trials. Still, there is an important finding when the analysis focuses on a subgroup of diabetics:
In the subgroup of patients with diabetes in this trial (the Samaha study), hemoglobin A1c values changed more favorably in individuals on the low-carbohydrate diet than on the low-fat diet at 12 months (−0.7%±1.0% vs −0.1%±1.6%; P=.02; after adjustment for weight loss).
After these results, it's interesting and actually amazing that the authors start their discussion (comment) with:
There was no clear benefit of either diet when their effects on cardiovascular risk factors were examined. Changes in blood pressure were not different between the 2 groups. Whereas total and LDL-C levels decreased more in individuals randomized to low-fat diets, HDL-C and triglyceride values changed more favorably in individuals randomized to low-carbohydrate diets.
So let's recapitulate on the changes:
Those on the low-carb diets w/o energy restriction achieved more weight loss in 6 months than those in a low-fat, energy restricted diet. After a year, the change was insignificant.
Those on the low-carb diets without w/o energy restriction tended to stay on the diet longer than those on the low fat diet. After a year, there was no difference.
Those on the low-carb diets without w/o energy restriction tended to improve their blood pressure, even though there was no difference after 12 monhts. But it didn't worsen, dit it?
Those on the low-carb diets without w/o energy restriction , despite not improving their total and LDL cholesterol (in the eyes of the researchers), they did improve their HDL and TGs, changed that remained after 12 months.
Where is the support to the claim that 'a low-carbohydrate diet' didn't have a clear benefit on cardiovascular risk?
The mega-flaw of this comparison lies on ignoring how the dynamics of cholesterol works (more on that in a minute). Other flaws are inherent to the studies included in the analysis. First of all, none of them offer a clear-cut 'low-carbohydrate approach'. The studies start with certain amounts of carbs but then the participants were allowed to increase their carb intake after two weeks. Some of those studies don't even have proper counseling during the whole duration of the studies so the participants were left pretty much on their own and we all know that without accountability, compliance cannot be 100%. More over, the trials included in the analysis don't understand 'low-carb' in the same way. As expected, 'low-carb' is immediately associated with Atkins or South Beach, which only shows poor research into what is actually proposed by what they call 'popular low-carb diets'.
The authors of the meta-flawed analysis chose to comment on the strenghts and limitations of their work, not based on the actual and fundamental flaws of the trials included, from the physiological point of view that is, but instead they chose to pick on the quality of the studies.
The quality of the included trials was moderate. Whereas most trials used concealed treatment allocation, losses to follow-up were quite substantial. No trial reported blinded outcome assessment. Therefore, we were not able to conduct sensitivity analyses comparing trials with blinded outcome assessment vs trials without it, as originally planned. The absence of blinded outcome assessment is a flaw that potentially limits the validity of individual trials.
This is a good criticism on how to actually design studies that can be compared later with studies that explore different nutritional approaches but doesn't address the fact that in the trials included in the analysis, the low-carbohydrate diets scored better in the main and secondary end points the authors selected to compare.
Dropout rates were substantial. After 1 year of follow-up, between 31% and 48% of individuals randomized to low-carbohydrate diets, and between 37% and 50% of individuals randomized to low-fat diets, had dropped out of the trials.
This only shows that without proper counseling during the entire study, half of the participants will drop out for whatever reason and it has nothing to do with one diet or the other. Nonetheless, the trend of dropout was higher in the low-fat groups.
We conclude that low-carbohydrate diets appear to be at least as effective as low-fat diets in inducing weight loss for a duration of up to 1 year. Low-carbohydrate diets are associated with unfavorable changes in total cholesterol and LDL-C levels, but favorable changes in triglyceride and probably HDL-C values. In the absence of evidence that lowcarbohydrate diets reduce cardiovascular morbidity and mortality, such diets currently cannot be recommended for prevention of cardiovascular disease.
So what is the actual meta-flaw? The authors are thinking of cardiovascular risk as reflected by values of LDL. Basically, they're concluding that since the low-carbohydrate diets didn't reduce LDL values, then they didn't bring up any benefits. In the author's minds, there may be a paradox of some sort because eventhough total and LDL cholesterol levels didn't 'improve' (i.e. go down), HDL went up and TGs went down and remained down despite not improving their LDL (which of course is bad in the researcher's mind). Should they have stopped for a moment to reflect on this particular finding, they would summon their mighty search strategies just to see what's out there regarding the role of fat in 'the cholesterol issue', and the actual role of LDL particles in the whole dynamics of cholesterol homeostasis. That would also have led them to find that there is a relationship between triglycerides and HDL and they would have find out that in terms of cardiovascular risk, the relationship between TGs and HDL is far more indicative than the actual level of LDL. That would have opened a whole different can of worms for them because then their whole interpretation of the trials included in their analysis had to be done under a different light.
With the amount of publish information that establishes that it is the LDL profile, rather than the amount of LDL what matters in terms of assigning risk, it's almost unacceptable that they wouldn't at least offer a comment on that in their discussion. More over, none of the 5 trials included in the meta-analysis actually measured LDL profiles; LDL was calculated from the Friedewald formula, which has been shown to be flawed as well. This methodological shortcoming would be enough to exclude the trials from any comparison that looks at cardiovascular risk.
What about the changes in HDL and TGs? That alone should direct any discussion to the fact that cardiovascular risk improved in those who increased their HDL and lowered their TGs; those on a low-carbohydrate diet.
If anything, this meta-analysis (of 5 studies, mind you) shows exactly what aspects to consider when interventional studies are designed to test nutritional approaches. I learned from this analysis that attrition is something that can hamper a good study, thus counseling is of utmost important for the whole duration of the study. It is also clear that not everybody knows what exactly is 'low-carbohydrate' and even less know how to implement that in the long-term. If the quality of the studies is 'moderate' as the authors point out, then the conclusions cannot be so categoric. In other word, what gives anybody the authority to say 'this didn't work... although the source is more or less of good quality...'
At best, one can only conclude that even though the quality of the studies was moderate, the trend favors a low-carbohydrate diet to improve cardiovascular risk. Therefore, what we need is not more 'long-term' studies that use a low-carbohdyrate diet but better designed ones, that measure parameters that are truly informative (i.e. LDL profiles instead of LDL calculation from formulae), parameters that truly reflect cardiovascular risk.
I did learn a lot about how not to design a study but I don't think the authors learned much about the actual benefits of carbohydrate control and its benefits on cardiovascular risk.
My only question is if Yancy and Brehm actually took part in the writing and analysis in this work or if they were invited as co-authors because their trials were included. That little Adam, may be another story...