"A low-carbohydrate diet like Atkins is better at cutting blood pressure than weight-loss pills," BBC News reported. It said a study has found that twice as many people on a low-carb diet had lowered blood pressure compared to those taking orlistat. The main aim of this study was to assess weight change from these diets, and it found that both a low-carb diet and a low-fat diet with orlistat had a similar effect. Although it found that the low-carb diet was associated with lowered blood pressure, it also found the low-fat/ orlistat diet was associated with better cholesterol levels. Also, many statistical tests were carried out, increasing the likelihood these findings are due to chance alone. On balance, this is not strong evidence that low-carb diets lower blood pressure or are better in general. Carbohydrates are important, and substituting them with fat and protein for a long period of time is not recommended. It should be noted the study did not include pregnant or breastfeeding women, or people with long-term health conditions. People in these groups should not consider such prescriptive diets. Where did the story come from? This research was carried out by Dr William S. Yancy Jr. and colleagues from the Department of Veterans Affairs Medical Center and Duke University Medical Center, North Carolina. The study was funded by the Department of Veterans Affairs. The paper was published in the peer-reviewed journal Archives of Internal Medicine. The newspapers have headlined this study’s finding that the low-carb diet reduced blood pressure to a greater extent than the low-fat/orlistat combination diet. However, this was not the primary outcome that this study set out to investigate. The high number of statistical tests that were carried out for other outcomes reduces the strength that should be placed on these results. Although the low-carb diet gave improved results for certain outcomes, other outcomes were better with the low-fat/orlistat diet. What kind of research was this? This randomised controlled trial aimed to compare a low-carbohydrate diet with a low-fat diet combined with orlistat tablets. A randomised controlled trial is the best way of comparing the effectiveness of different treatments. Trials should have a reasonable number of participants, who should be followed up and assessed in their designated treatment groups. There should also be pre-defined main outcome(s) that the researchers aim to examine (called the primary outcomes). Though researchers may also look at other outcomes (called secondary outcomes), the results from these analyses should be considered as tentative, especially if many secondary outcomes are assessed. In trials assessing diet, it is also helpful if there is some way of ensuring standardisation of the diet, and assessing people’s compliance to it. What did the research involve? Overweight or obese people were recruited from the Department of Veterans Affairs primary care outpatient clinics in 2005-06. Those eligible for the study either had a BMI of over 30; or had a BMI of 27-30 in addition to having a weight-related medical condition (e.g. type 2 diabetes). People were excluded if they had a variety of medical complications, including liver or kidney impairment, mental health illness or severe high blood pressure. This resulted in 146 people suitable for inclusion, with an average age of 52, average BMI of 39.3; 72% of whom were male, and 32% of whom had type 2 diabetes). These people were randomised to either a low-carb diet (initially less than 20g of carbohydrates daily), or to orlistat treatment (120mg three times daily) combined with a low-fat diet (less than 30% energy from fat, 500-1000kcal/day energy deficit). The participants were given dietary instructions, including pocket guides and handouts at small group meetings at the clinic throughout the 48-week study period. The low-carb diet involved unlimited meat and eggs, cheese and low-carb vegetables, with no restriction to calorie intake. All participants were advised to limit their intake of alcohol and caffeine, drink 6 to 8 glasses of fluid a day, take a daily vitamin tablet, and exercise for 30 minutes at least three times a week. The main outcome of interest was body weight (the primary outcome). But blood pressure, fasting cholesterol and blood sugar were also assessed at each clinic visit. Diet adherence was measured using four-day food records completed by the participants at the study start and at weeks 2, 12, 24, 36, and 48. Urinary ketones (compounds that are increased when consuming an Atkins-style diet) were also assessed as a measure of compliance to the low-carb diet: higher levels indicated greater compliance. All participants were assessed in their designated treatment groups, regardless of whether they completed treatment. The trial was well-designed, although both study participants and assessors were aware of their assigned treatments, which is a limitation (the study was unblinded). What were the basic results? Of the 146 people included, 77% completed the study. At two weeks, 72% of the low-carb groups had high urinary ketones. This declined to 13% at 48 weeks, suggesting that fewer people were complying with the diet by that time. Weight loss, the primary outcome, was similar for both groups. However, there were group differences for the other outcomes. Compared with the orlistat/low-fat diet, the low-carb diet had a better effect on lowering blood pressure and improving blood sugar control. There was a similar reduction in high-density lipoprotein (‘bad’ cholesterol) in both groups, but low-density lipoprotein (‘good’ cholesterol) levels increased in the orlistat/low-fat diet group only. How did the researchers interpret the results? The researchers conclude that their study in a sample of medical outpatients has found that a low-carb diet gives similar improvements in terms of weight loss to a low-fat diet in combination with orlistat tablets. The low-carb diet was found to be more effective for lowering blood pressure. Conclusion This is a generally well-designed study. However, interpretations should be made with care. The study doesn’t provide evidence that the low-carb diet is better than the low-fat/orlistat combination diet. The main outcome that the study was designed to assess was weight change. This was the same between both groups at the end of the study. The press has focused on the finding that the low-carb diet reduced blood pressure to a greater extent. However, although the low-carb diet was associated with improved results for some outcomes, other outcomes were better with the low-fat/orlistat diet. The considerable number of statistical tests that were carried out for other outcomes (that were not the prime focus of this study) reduces the reliability of these results. Although the study assessed all people in their assigned treatment groups, regardless of whether they completed the study, dietary compliance appeared to decline as the study went on, and was low at the 48-week completion. One of the main drawbacks of the study is that the longer-term health effects of the low-carbohydrate diet were not assessed. Carbohydrates form an important food group, and abstinence from them for a long period of time while substituting fat and protein is not recommended. Additionally, it should be noted that pregnant or breastfeeding women, or those with various health complications, were not included in the study and should not consider such prescriptive diets. The ideal way to achieve weight loss and reduce the risk of weight-related health complications is to have an active lifestyle and eat a balanced diet.