Blueberries improved vascular measures, but not the primary insulin endpoint
A six-month metabolic-syndrome trial produced selective rather than across-the-board improvements.

Illustrative food image, not a photograph of the study intervention.
Begin with the primary question
This double-blind trial enrolled 115 people with metabolic syndrome and compared blueberry intakes equivalent to half a cup or one cup daily with matched placebo for six months. Participants averaged 63 years of age, and 68% were men. Insulin resistance was the primary endpoint, assessed using a homeostasis model and examined with a clamp method in a subset of 20 participants. Vascular function and other cardiometabolic measures were also assessed. Stating that hierarchy matters: an encouraging secondary vascular finding should not silently replace the primary metabolic question in the story.
Curtis PJ, van der Velpen V, Berends L, Jennings A, et al.. The American journal of clinical nutrition, 2019. PMID 31136659 · DOI
The larger dose changed selected vascular measures
The one-cup-equivalent group showed improved flow-mediated dilation, with a reported increase of 1.45 percentage points and a 95% confidence interval from 0.83 to 2.1. The augmentation index, another vascular measure, also improved. These results concern physiological measurements during the intervention, not heart attacks or strokes counted over subsequent years. The abstract discusses predicted cardiovascular risk reductions, but those predictions should not be retold as observed clinical events. The directly supported claim is a sustained change in selected vascular markers within this six-month randomized comparison in metabolic syndrome.
Curtis PJ, van der Velpen V, Berends L, Jennings A, et al.. The American journal of clinical nutrition, 2019. PMID 31136659 · DOI
The null findings belong beside the benefit
Insulin resistance did not improve. Blood pressure, pulse wave velocity, nitric oxide measurements, and overall plasma thiol status were also unaffected. The half-cup-equivalent assignment had no effect on the reported biomarkers. That collection of null findings rules out a blanket headline that blueberries improved every cardiometabolic endpoint. It also means this study cannot be cited as evidence that the tested blueberry intake corrected insulin resistance. A fair account can retain the vascular result while giving equal visibility to the unchanged primary endpoint and the less successful lower-dose comparison.
Curtis PJ, van der Velpen V, Berends L, Jennings A, et al.. The American journal of clinical nutrition, 2019. PMID 31136659 · DOI
A subgroup does not describe everyone
Improvements in HDL-related measures were reported among the 71 participants who were not using statins. Those findings need their subgroup label rather than being generalized to every participant. The trial tested specified blueberry equivalents against placebo, not the linked oats recipe or every berry-containing meal. Its vascular measurements do not establish prevention of cardiovascular disease, and its insulin result does not establish diabetes treatment. The practical editorial boundary is straightforward: preserve the dose, the population, the endpoint hierarchy, and the subgroup restrictions instead of turning several selective signals into a universal food claim.
Curtis PJ, van der Velpen V, Berends L, Jennings A, et al.. The American journal of clinical nutrition, 2019. PMID 31136659 · DOI