Sunday, February 2, 2014

The St Thomas Atherosclerosis Regression Study (STARS)

Studies Associated with the Trial 

Effects on coronary artery disease of lipid-lowering diet, or diet plus cholestyramine, in the St Thomas' Atherosclerosis Regression Study (STARS) (1992) [1]
Coronary atheroma regression trials (1992) [2]
The St. Thomas Atherosclerosis Regression Study (STARS) (1992) [3]
Independent associations between plasma lipoprotein subfraction levels and the course of coronary artery disease in the St. Thomas' Atherosclerosis Regression Study (STARS) (1993) [4]
Nutrient intake and progression of coronary artery disease (1994) [5]
Metabolic Determinants of the Course of Coronary Artery Disease in Men (1994) [6]
von Willebrand factor, a possible indicator of endothelial cell damage, decreases during long-term compliance with a lipid-lowering diet (1995) [7]
Relationships between nutrient intake and progression/regression of coronary atherosclerosis as assessed by serial quantitative angiography (1995) [8] (no access)
Nutritional, metabolic, and genetic determinants of the progression of coronary heart disease. STARS Group (1995) [9]
Dietary fatty acids and progression of coronary artery disease in men (1996) [10] 

Participants and Diets 

90 men with coronary heart disease (CHD) and high cholesterol (6.1-10.0 mmol/l or 236-387 mg/dl) were randomised to one of three groups: 

·         Usual care (control) (U)
·         Dietary intervention (D)
·         Dietary intervention + cholestyramine (a bile acid sequestrant) (DC) 

The dietary intervention was intended to be low fat (27%), low SFA (8-10%), low dietary cholesterol (100mg/1000kcal) and high PUFA (8%), with soluble fibre (mainly pectin) (3-6g/1000kcal).  Overweight patients (BMI > 25) were put on a low calorie diet (1,000-1,200kcal) to achieve a BMI of 25.  The diet + cholestyramine group was also told to take 8g of cholestyramine twice daily with meals [1].  Due to the low calorie diets for weight loss, and consequently the lower energy intake in the diet group, the diet group had modest, but significant weight loss at 3 years (-2.9kg vs. 0.3kg) [10]

 The diet intervention was based off some other papers that were promoting a multifactorial diet.  The first mainly discusses things to reduce LDL-C like PUFA, fibre and soy protein, while the second was the fairly standard whole foods, mostly plants, low animal fat, but also commercial baked goods were severely restricted”.

The diet group in this trial also had 43.3% less trans fat (2.47g vs. 4.36g) and twice as much long chain omega 3 (0.48g vs. 0.24g) [10] 

“The control group received, in common with both intervention groups, cardiological supervision and treatment, repeated counselling against smoking, and antihypertensive treatment if appropriate. All participants were advised about a suitable level of daily exercise. Patients in the control group with a BMI above 25 kg/m2 were advised to lose weight but did not receive formal dietary counselling.” [1] 

At baseline, there were several minor, non-significant differences between the groups [1]


LDL-C decreased in the diet group and more so in the diet + cholestyramine group, there was no effect on HDL-C, and triglycerides decreased in the diet group, but not the diet + cholestyramine group [1]


Measures of atherosclerosis worsened in the control group, but improved in both the diet and the diet + cholestyramine groups [1]

Both the diet and the diet + cholestyramine groups had significantly fewer total cardiac events and significantly fewer angina symptoms than the control group*.  The improvement in angina symptoms was correlated with the change in atherosclerosis (MAWS).  All deaths were from CVD and the three patients who died in the control group had severe cardiovascular disease at baseline [1]

Yet again we have another multifactorial trial, so you can’t tell whether the benefit came from replacing SFA with PUFA, weight loss, increasing pectin, less trans fats, more LCO3, restricting commercial baked goods or potentially more whole plant foods.  As PUFA intake wasn’t substantially different between the groups (7.4% vs. 4.8%), it should certainly not be concluded that such a small increase in PUFA was responsible for the impressive result in STARS
* ApoE4 frequency was significantly higher in the STARS patients than in a healthy UK sample” [9] 

** The researchers looked for correlations, but I haven’t mentioned them and wouldn’t put much weight in them because the intervention was multifactorial (so things like MUFA were associated with atherosclerosis because the intervention was low fat), the correlations can simply be a measure of compliance for another thing (like smoking, exercise and junk food) and things that the intervention didn’t change (like HDL-C) didn’t have significant correlations


  1. What are your thoughts on these

    1. Hi Charles, I'm looking at trials that replace SFA with PUFA, whereas those papers are looking at drug trials or other multifactorial trials (diet + exercise) for regression of atherosclerosis. As such I haven't heard of them (except STARS and Dean Ornish's) and can't really comment.

      People generally don't dispute that certain cholesterol lowering drugs are effective for people with CVD and those with FH, but are instead more concerned about adverse side effects and the conflict of interest with trials funded by pharmaceutical companies

  2. Thank you so much for these article reviews! I hope I am able to find them again whenever I am thinking about a related issue in the future.

    1. Hi Zooko, I'm glad you like them. This is the last trial in the series and next week I'm going to do a summary and conclusions kind of post. You'll be able to access the posts on the trials from there