- Click here for explanation of classification scheme.

To assess the combined effect of plant sterols, soy proteins and viscous fibers on blood lipids in hyperlipidemic subjects.

• Attended the Risk Factor Modification Center, St. Michael's Hospital, Toronto, Canada
• Participated in previous diet studies and experienced in following dietary protocols
• History of raised LDL-cholesterol levels (4.1mmol per L)
• Following NCEP Step 2 diet.

• No history of diabetes, renal or liver disease
• Not taking medication known to influence serum lipids
• Not taking anti-inflammatory drugs.

Recruitment

Recruited from patients attending the Risk Factor Modification Center, St. Michael's Hospital, Toronto, Canada.

Design

Time series.

Intervention

• Dietary protocol
  • Week zero: Run in diet.
    • Subjects followed their usual diet
    • Dietary advice on low saturated fat (less than 7% calories) and low cholesterol (less than 200mg per day) had been reinforced on at least two occasions during the past year and at study entry
  • Weeks one to four: Portfolio diet (combination diet)
    • All foods provided at weekly clinic visits except fresh fruit and most vegetables, which participants purchased at the grocery store
    • Subjects received the same seven-day rotating menu, which was modified to suit individual preferences and provided goals for viscous fiber, soy protein, plant sterol and almond consumption
    • Weight maintaining (based on estimated caloric requirements)
    • Subjects provided with a scale and asked to weigh all food items consumed
    • Combination diet specifics:
      • Macronutrient composition was approximately 22% protein, 50% carbohydrate, 27% fat, 4% SFA, 12% MUFA, 10% PUFA
      • One gram plant sterols per 1,000 calories as enriched margarine
      • 8.2g viscous fibers per 1,000 calories from oats, barley and psyllium
      • 22.7g soy protein per 1,000 calories
      • 2.9g almonds per 1,000 calories
      • Additional protein provided as beans, chickpeas and lentils
      • Additional viscous fiber provided by eggplant and okra
  • Weeks five to six: Run-out diet. Subjects followed their usual diet
• One subject undertook two additional one-month phases separated by two-week washout periods
  • Control diet: Low saturated fat and low cholesterol with the same macronutrient profile as the combination diet minus the sources of viscous fibers and plant sterols, and skim milk products replaced the soy and vegetable protein sources
  • Control diet plus 20mg lovastatin daily to compare the combination diet directly with drug therapy.

Statistical Analysis

•  Analysis of variance was used to determine whether there was a significant F-value in the comparison of weeks two, three and four of the combination diet. In the absence of a significant F-value, the restricted mean of weeks two, three and four was used as the combination treatment value.
• The significance of the differences between the retreatment diet, combination diet and post-treatment diet was assessed by the least squares means test with Tukey adjustment. Model used had the treatment value as the response variable and week and week by sex as main effects and a random term corresponding to subject nested with sex.
• Two-tailed Student's paired T-test was used to asses the significance of the percentage change from pre-treatment
• Hegsted equation was used to predict the changes in serum cholesterol resulting from alteration in dietary fatty acid and cholesterol intakes
• The Framingham 10-year cardiovascular disease risk equation was applied to the data using systolic blood pressure, age, sex and LDL:HDL cholesterol values.

 

Timing of Measurements

 

	Week 0	Week 1	Week 2	Week 3	Week 4	Week  5	Week 6
Seven-day, weighed, diet history	x					x	x
Satiety questionnaire	x	x	x	x	x	x	x
Weekly menu check lists		x	x	x	x
Blood pressure	x	x	x	x	x	x	x
Body weight	x	x	x	x	x	x	x
Fasting blood lipids	x	x	x	x	x	x	x
Apolipoproteins A-1, B; lipoprotein (a)	x	x	x	x	x	x	x
Conjugated dienes in the LDL fraction	x	x	x	x	x	x	x
C-reactive protein	x	x	x	x	x	x	x
Homocysteine	x	x	x	x	x	x	x
Red cell fragility	x	x	x	x	x	x	x

Dependent Variables

• Blood lipids
  
  • Total cholesterol
  • LDL-cholesterol (calculated)
  • HDL-cholesterol
  • Triglycerides
  • Apolipoproteins A-1, B
  • Ratios
    • Total cholesterol: HDL-cholesterol
    • LDL-cholesterol:HDL-cholesterol
    • ApoB:ApoA-1
• Lp(a)
• Homocysteine
• C-reactive protein
• Oxidized LDL (conjugated dienes in the LDL fraction)
• Blood pressure
  • Systolic
  • Diastolic
• Red cell fragility
• Diet satiety: Seven-point bipolar semantic scale where -3 is extremely hungry, zero is neutral and +3 is stage of satiety just prior to discomfort
• Diet acceptability.

Independent Variables

Combination diet: Diets analyzed with a program based on USDA data with additional foods analyzed in the lab for protein, total fat and dietary fiber. Additional dietary fiber values obtained from literature.

 

 

Initial N

N=13 (7 men, 6 post-menopausal women).

Attrition (final N): 

• N=12: Run-in and run-out diet
• N=13: Portfolio diet
• N=11: C-reactive protein assessment
  • One subject received antihistamines
  • One subject received anti-inflammatory drugs
• One subject withdrew after three weeks due to dyspepsia associated with Helicobacter pylori infection requiring antibiotics.

Age

65±3 (mean±SE), range 43 to 84 years.

Anthropometrics

 No subjects smoked or had evidence of diabetes or left ventricular hypertrophy.

CHD Risk Factor	mean±SE	Range
Weight, kg	69.9±3.6
BMI	25.6±0.9	20.6 to 30.7
Cholesterol
Total cholesterol (mmol per L)	6.46±0.21
LDL-cholesterol (mmol per L)	4.22±0.11	3.45 to 6.61
HDL-cholesterol (mmol per L)	1.37±0.11
Triglycerides (mmol per L)	1.92±0.35
Apolipoproteins
Apo A-1 (g per L)	1.70±0.07
Apo B (g per L)	1.32±0.05
Ratios
Total C:HDL-C	5.06±0.41
LDL-C:HDL-C	3.31±0.26
Apo B:Apo A-1	0.80±0.05
LDL conjugated dienes (µmol per L)	47.9±4.2
Homocysteine (µmol per L)	6.9±0.5
Lp(a) (mg per dL)	11.5±3.0
C-reactive protein (mg per L)*	1.81±0.55
Blood pressure (mm Hg)
Systolic	117±4
Diastolic	70±3
Calculated CHD risk (10 year %)	10.3±1.2

 At study entry:

• Raised LDL-cholesterol (more than 4.1 mmol per L): N=5
• Raised triglycerides (greater than 2.30mmol per L): N=1
• Raised cholesterol and triglycerides: N=3
• Low HDL-cholesterol (less than 0.9mmol per L): N=1
• Blood lipids in normal range: N=3.

Location

St. Michael's Hospital, Toronto, Canada.

 

 

	Baseline	Portfolio Diet	P-Value	% Reduction	P-Value
Total cholesterol (mmol per L)	6.46±0.21	5.01±0.20	P<0.05	22.3±2.0	P<0.001
LDL-cholesterol (mmol per L)	4.22±0.11	3.01±0.17	P<0.05	29.0±2.7	P<0.001
HDL-cholesterol (mmol per L)	1.37±0.11	1.34±0.11	NS
Triglycerides	1.92±0.35	1.45±0.18	NS
Apo A-1 (g per L)	1.70±0.07	1.61±0.08	NS
Apo B (g per L)	1.32±0.05	1.01±0.05	P<0.05	24.2±2.0	P<0.001
Total cholesterol:HDL cholesterol ratio	5.06±0.41	4.00±0.30	P<0.05	19.8±2.9	P<0.001
LDL-cholesterol:HDL cholesterol ratio	3.31±0.26	2.45±0.24	P<0.05	26.5±3.4	P<0.001
Apo B:Apo A-1 ratio	0.80±0.05	0.64±0.05	P<0.05	19.7±2.7	P<0.001
Calculated CHD risk (10 year %)	10.3±1.2	7.3±1.1	P<0.05	30.0±4.6	P<0.001
Oxidized LDL (conjugated dienes)	47.9±4.2	31.0±1.9	P<0.05	32.9±12.3	P<0.001

• The predicted reduction in serum cholesterol based on the change in dietary fatty acid and cholesterol intake was 9.3%±1.1% on the combination diet. The difference between the observed and predicted reduction in serum cholesterol was 13.3%±2.5% (P<0.001), attributable to the viscous fiber, soy protein and plant sterols.
• The two-week run-out post portfolio diet LDL-cholesterol values were 3.8±0.20. This represents 10.1%±3.8% reduction compared to baseline values, P=0.022
• No significant differences were seen in Lp(a), homocysteine, blood pressure and red blood cell fragility between run-in diet, portfolio diet and run-out diet
• Data for the subject who completed three phases indicated that the statin and combination phases produced similar reductions in the ratios of total cholesterol:HDL and LDL-cholesterol:HDL-cholesterol
• Subgroup analyses of subjects who initially had either normal lipids, raised LDL-cholesterol alone, raised triglycerides alone, raised LDL-cholesterol and triglyceride combined, or low HDL-cholesterol all demonstrated large reductions in total and LDL-cholesterol and calculated CHD risk in response to the combination diet
• Classification according to raised or normal LDL-cholesterol initially indicated a similar response to the portfolio diet, including calculated CHD risk, which was independent of starting LDL cholesterol level.

Other Findings

• Throughout the study, subjects tended to lose weight: -0.10±0.05kg per week (P=0.127) over the combination diet; and -0.2±0.05kg per week (P=0.001) during the run-out phase
• For the majority of the subjects, compliance in terms of caloric intake was good at 92.5%±2.9%
• Overall rating of the diet was 6.3±0.6 (scale zero to 10), which was significantly above 5.0, the level at which diet was acceptable with minor modifications (P=0.043). Nine  of 13 subjects would be willing to continue the combination diet, possibly with small modifications, as their therapeutic diet.
• All subjects felt that they were eating as much food as they were capable without experiencing discomfort (satiety rating: 2.9±0.2 vs. 1.3±0.2 at week zero, P=0.002; scale -3 to +3).

 

We conclude that a portfolio approach of diversifying the dietary investment in a range of cholesterol-lowering components with different mechanisms of action is effective in reducing lipid risk factors for cardiovascular disease. In combination, viscous fibers from oats and psyllium, vegetable proteins emphasizing soy and almonds and plant sterols reduce LDL-cholesterol to the same extent as the starting dose of the older statins. It is also possible that, as with the statins, these diets will have pleotrophic effects, including anti-inflammatory effects.

Government:	Federal Government of Canada
Industry:	Almond Board of California Commodity Group:

• Authors reported compliance relative to caloric intake but unsure whether this meant compliance with the viscous fibers, plant sterols and soy proteins as well
• Authors failed to account for autocorrelated errors in the  statistical model described. Their model as stated assumes every observation is independent, yet the authors acknowledge that repeated measures were taken from subjects, therefore not all observations were independent. Specifically, some non-independent error structure should have been specified.
• Author afiliations not listed individually. EA Trautwein affiliated with Unilever Health Institute, Unilver R&D; KG Lapsley affiliated with the Almond Board of California. Their roles on the project were not described. The Almond Board supported the project.