OBESITY
ABSTRACT & COMMENTARY
The Final Word? Low-carb vs Low-fat
By David Kiefer, MD, Editor
Summary Points
- In a racially diverse obese population, people on a low-carbohydrate diet for 12 months were better able to lose weight than those on a low-fat diet.
- The low-carbohydrate diet also lowered cardiovascular risks more than the low-fat diet.
Synopsis: This randomized, parallel group study over 12 months found greater weight and cardiovascular risk benefits in obese adults eating a low-carbohydrate diet vs a low-fat diet.
Source: Bazzano LA, et al. Effects of low-carbohydrate and low-fat diets. Ann Intern Med 2014;161:309-318.
The researchers of this randomized, parallel-group trial aimed to address gaps in the literature about so-called "low-carb" (low-carbohydrate) diets, particularly focusing on cardiovascular effects and the effects in a racially diverse population.
The recruitment was from the general population, and, unlike most past research, in people with no comorbidities. Exclusions were cardiovascular disease, type 2 diabetes, kidney disease, surgery, recent weight loss (> 6.8 kg), or the use of prescription weight-loss medications. The 148 participants (51% African American, 88% female, mean age of 46.8 years) were randomized to either a low-carbohydrate group (a diet containing < 40 g per day of digestible, non-fiber carbohydrate) or a low-fat group (a diet of < 30% of total energy intake containing total fat [< 7% saturated fat] and 55% carbohydrate). There were no calorie or energy goals stipulated. Also, participants were asked to not change their exercise amount during the intervention. Each group received the same information about fiber intake (recommended at 25 g daily) and dietary fats (emphasizing monounsaturated fats and avoidance of trans saturated fats).
Each participant received a detailed handbook about food suggestions and recipes for their specific diet. In addition, participants had weekly individual counseling sessions with a dietitian for the first month, then small group sessions every other week for 5 months, and monthly for the last 6 months. One replacement meal (bar or shake) was provided daily for each participant, tailored to his/her group assignment.
With respect to data collection, a detailed medical history was obtained from each study participant, and they were asked to do two 24-hour dietary recalls (one weekday and one weekend) at baseline, and 3, 6, and 12 months in order to calculate dietary nutrient intakes. The researchers measured body weight and height, body composition, blood pressure, and collected urine and blood (lipids, C-reactive protein, glucose, creatinine, insulin).
At baseline, participants in the two groups were similar across all of these measurements, except for the parameters listed in Table 1.
Of the 148 people randomized at the beginning of the study (73 in the low-fat diet group and 75 in the low-carbohydrate group), 60 and 59, respectively, were assessed at the end of the 12 months, though all 148 participants were included in an intention-to-treat data analysis. At the end of 12 months, as expected, the low-fat diet group consumed higher amounts of carbohydrates and lower amounts of proteins and fats (P < 0.001). The researchers stated that "Physical activity amounts were similar throughout the study," but it is unclear if this refers to baseline vs 12 months, or between the two diet groups. The total calorie consumption was similar between the two groups over the course of the study.
Compared to baseline, people on the low-carbohydrate diet lost 3.5 kg more weight over 12 months than people on the low-fat diet (P = 0.002). Also, the low-carbohydrate group gained 1.7 kg lean mass (P = 0.003), lost 1.5 kg fat mass (P = 0.011), increased HDL 6.9 mg/dL (P < 0.001), decreased triglyceride 14.2 mg/dL (P = 0.038), decreased total-HDL ratio 0.44 (P = 0.002), and decreased C-reactive protein 15.2 nmol/L (P = 0.024). Overall, the low-carbohydrate diet group lowered their 10-year Framingham risk 1.4 points compared to the low-fat group (P < 0.001). There were no changes in waist circumference, total cholesterol, LDL cholesterol, blood pressure, glucose, serum insulin, nor serum creatinine.
Symptoms that developed over the course of the study were similar between the two groups except for more headaches reported in the low-fat diet group (P = 0.03). No serious adverse effects were reported.
COMMENTARY
There are several compelling aspects to this study. Few other studies have analyzed a racially diverse population, in this case 51% African American. Too often, clinicians are left speculating about how to extrapolate research results from a predominantly white study population to their patient demographic. Also, this study included people who were obese but otherwise healthy, allowing the results to be an important commentary on the possible preventive aspect of the dietary change, again a relatively unique aspect to this research thread.
The benefits of the low-carbohydrate diet in this population were clear; numerous significant changes in laboratory tests were seen, including an increased HDL, otherwise a difficult laboratory test to budge. With a movement away from "number chasing" in recent guidelines, it is good to also see that overall risk via the Framingham also showed a concrete benefit.
It could be argued that a 3.5 kg (almost 8 pounds) weight loss over 1 year is negligible, but it should be mentioned that this was in the context of stagnant exercise amounts and no limit on calorie intake. It makes one wonder what could have been accomplished by a shift in those other variables too.
Did the baseline inter-group differences (see Table 1) affect the final results? Ideally, the two groups would have been identical at baseline, but this was not the case and it brings up some questions. The people in the low-carbohydrate diet group were less active, but had better lipid profiles, than the low-fat diet group. If the lipid profile is representative of metabolic (dys)function, then it could be argued that it would have been even more difficult to shift any of the measurements in the low-carbohydrate diet group, starting as it was at baseline better than the low-fat diet group. This, then would strengthen the findings. However, less activity may indicate that the low-carbohydrate diet group had "wiggle room" in affecting insulin sensitivity, making it more likely that dietary change would shift physiology, and, therefore, weight and cardiovascular risks. If future, larger trials can increase numbers and random sampling to even the initial playing field, then these methodological issues will be addressed.
How difficult would it be to implement this dietary change in the "real world"? The low-carbohydrate goal of < 40 g on non-fiber carbohydrate is truly low. Many people consume 200+ g of fiber and non-fiber carbohydrate a day, so the study diet would be a significant change from normal. Adding to this reality check, it helps to view this study’s diets in context of some other well-known dietary guidelines (see Table 2). Also, this study had a high completion rate (about 80%), higher than most other comparable studies, which was almost certainly due, in part, to the many individual and group counseling sessions. Such an intensive effort is usually beyond what clinics can offer their patients. Perhaps the most useful part of these results is to serve as a benchmark for which people can aim, a goal to shoot for, an ideal that gives them hope that their efforts at dietary change do, in fact, matter.
Interestingly, as much as the literature supports a connection between weight and risk for diabetes, no changes were seen here in serum glucose nor insulin, recognizing that this study population was disease free. It would have been useful to follow this cohort longer and analyze for incidence of diabetes; would the weight loss and metabolic changes translate into improvements in diabetes risk as well?
In the context of cardiovascular risk and obesity, this study is yet another strong indication that carbohydrates, rather than fats, should be the focus when it comes to treating or preventing obesity. In general, there seems to be little reason to recommend a low-fat diet in this population, and many (more) reasons now to help our patients to avoid non-fiber carbohydrates.
|
Table 1. Baseline Characteristics Differing Between the Low-carbohydrate and Low-fat Diet Groups |
||
|
Measurement |
Low-carbohydrate Diet Group |
Low-fat Diet Group |
|
Serum cholesterol (md/dL) |
198.8 |
204.3 |
|
Serum triglyceride (md/dL) |
112.6 |
125.5 |
|
Serum creatinine (micromol/L) |
88.4 |
97.2 |
|
Mean physical activity level (MET-hours/week) |
16.3 |
19.6 |
Table 2. Comparison of Daily Nutrient Intakes Between the Groups in this Study with Three Other Well-known Nutritional Guidelines |
|||||
|
This Study: |
This Study: |
USDA 2005 Food Pyramid1 |
MyPlate (based on the 2010 Dietary Guidelines for |
Ornish Spectrum Diet3 |
|
|
Carbohydrate |
< 40 g non-fiber |
55% of daily calories |
3 oz of whole grains daily |
||
|
Fat |
< 30% of daily calories |
20-35% of daily calories |
< 10% of daily calories |
||
|
Protein |
5-5.5 oz |
||||
References
- USDA Food Pyramid. 2005. Available at: http://www.foodpyramid.com/mypyramid. Accessed September 7, 2014.
- USDA ChooseMyPlate.Gov. 2010. Available at: http:// www.choosemyplate.gov. Accessed September 7, 2014.
- The Ornish Spectrum. Available at: http://ornishspectrum.com. Accessed September 7, 2014.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content
Already have an account? Log in