Portfolio foods versus lovastatin: what one month can show
A small randomized trial compared a multi-component dietary portfolio with lovastatin and a low-saturated-fat control.

Illustrative food image, not a photograph of the study intervention.
Three groups, not a single-food contest
The Jenkins trial randomly assigned 46 adults with hyperlipidemia to three outpatient interventions for one month. Sixteen received a control diet very low in saturated fat, 14 received the same diet plus 20 mg of lovastatin daily, and 16 received a dietary portfolio. Participants included 25 men and 21 postmenopausal women, with a mean age of 59 years. This was a direct comparison involving a particular medication and a particular dietary combination. It was not a trial of almonds alone, nor a comparison with every statin or every approach to cholesterol treatment.
Jenkins DJ, Kendall CW, Marchie A, Faulkner DA et al.. JAMA, 2003. PMID 12876093 · DOI
The LDL comparison in numbers
Mean LDL cholesterol fell by 8.0% in the control group, 30.9% in the lovastatin group and 28.6% in the portfolio group. The respective standard errors were 2.1, 3.6 and 3.2 percentage points. Both active approaches produced changes significantly different from the control diet, while the investigators reported no significant efficacy difference between portfolio and lovastatin. That last finding needs careful wording: a small study failing to detect a difference does not establish that the treatments are interchangeable. The result describes the observed LDL response over one month, not equivalent protection over years.
Jenkins DJ, Kendall CW, Marchie A, Faulkner DA et al.. JAMA, 2003. PMID 12876093 · DOI
What the portfolio actually contained
The portfolio combined plant sterols, soy protein, viscous fibers and almonds rather than singling out one ingredient. Amounts were specified per 1000 kcal: 1.0 g plant sterols, 21.4 g soy protein, 9.8 g viscous fibers and 14 g almonds. Those details matter when interpreting the impressive LDL change. A meal containing one of these foods is not automatically the intervention tested here. Nor does this three-group design reveal how much of the observed response each component contributed separately. The evidence supports the tested combination, with its specified composition, rather than a numerical promise for an individual recipe.
Jenkins DJ, Kendall CW, Marchie A, Faulkner DA et al.. JAMA, 2003. PMID 12876093 · DOI
Blood markers are not clinical events
The investigators measured fasting lipids and C-reactive protein, along with blood pressure and body weight, at baseline and at two and four weeks. C-reactive protein reductions were 10.0% in controls, 33.3% with lovastatin and 28.2% with the portfolio; the control change was not statistically significant. These are laboratory outcomes in a short trial of 46 people. The abstract does not report a comparison of heart attacks, strokes or survival. Its findings should therefore not be presented as proof that a dietary portfolio can replace prescribed medication, even though the short-term LDL comparison is notable.
Jenkins DJ, Kendall CW, Marchie A, Faulkner DA et al.. JAMA, 2003. PMID 12876093 · DOI