Two caveats travel with everything below. That review’s lead author disclosed past and present support from the US Highbush Blueberry Council, the Cranberry Institute, and the California Strawberry Commission. And most of the twenty human trials it tabulates gave other berries or mixed juice blends, so this page keeps blueberry-alone results separate from berry-general ones rather than reading the review’s summary sentences onto blueberries.
The one blueberry-alone trial with a lipid endpoint is a three-week randomized controlled study in 20 smokers, who ate 250 g of blueberries a day against a control group on a usual diet restricted in fruit, vegetables, and vitamin supplements. Lipid hydroperoxides fell in the blueberry group relative to control at three weeks (P < 0.001). That single trial is the whole of the blueberry-specific human cardiovascular evidence curated here.
The berry-general picture in the same review is wider but does not transfer. Across its twenty trials, using chokeberry, cranberry, strawberry, bilberry, black currant, and mixed products, improvements appeared in LDL oxidation, lipid peroxidation, total plasma antioxidant capacity, dyslipidemia, and glucose metabolism. Only two of those trials lowered systolic blood pressure, and both used non-blueberry products: a cranberry juice cocktail, and a mixed-berry regimen that also inhibited platelet function and raised HDL cholesterol. The review states that none of the clinical studies showed a significant effect on biomarkers of inflammation, apart from reduced adhesion molecules with cranberry juice.
Where blueberries appear beyond that single trial, the model is not human. Blueberry and cranberry anthocyanins reduced TNF-alpha-induced upregulation of inflammatory mediators in human microvascular endothelial cells, and purified blueberry and strawberry anthocyanins in drinking water prevented dyslipidemia and obesity in mice fed a high-fat diet for 90 days.
The observational evidence for blueberries specifically is null or unstable. In the Iowa Women’s Health Study (34,489 postmenopausal women, 16 years), blueberry intake tracked lower coronary heart disease mortality in an age- and energy-adjusted model, but the significance did not persist after adjustment for other confounders. In the Women’s Health Study (38,176 women, about 11 years), no significant association was reported between blueberry intake and either cardiovascular disease or C-reactive protein.
What this page does not claim: that blueberries reduce LDL oxidation, dyslipidemia, or blood pressure in humans. Those endpoints moved in trials of other berries and berry blends, and no curated source measured them under blueberries alone, so the claim here stops at lipid peroxidation in smokers.
Three small trials in the review gave a blueberry product on its own and measured antioxidant status. In two single-blind crossover studies of a high-fat meal, one in 8 middle-aged men and one in 5, adding 100 g of freeze-dried wild blueberry powder raised serum antioxidant status (P < 0.05). In a three-week randomized crossover, 22 male smokers drinking 330 mL a day of a fruit drink made with 30% clarified blueberry juice concentrate had higher vitamin C, carotenoids, and plasma antioxidant capacity (P < 0.05).
A fourth trial found nothing. Among 9 healthy women given 500 mL of juice, cranberry juice raised antioxidant capacity, vitamin C, and phenols, while there were no effects with blueberry juice. Two further positive trials used blends, one of aronia, blueberry, and boysenberry juice and one a multi-fruit juice blend, so their reductions in plasma TBARS, lymphocyte oxidative DNA damage, and lipid peroxidation cannot be attributed to blueberries.
These are 5- to 22-person studies, mostly postprandial, measuring plasma antioxidant markers rather than health outcomes, and the review’s own bioavailability caveat applies to all of them.
What this page does not claim: that blueberries are exceptional in anthocyanin content, or that neutralizing free radicals is a demonstrated clinical effect. The review’s own table ranks them fifth among the whole berries listed, and what the trials measured is a short-term shift in plasma antioxidant markers.
The only blueberry blood-sugar result in either curated source is an animal one: male C57Bl/6j mice on a high-fat diet supplemented with blueberry powder for eight weeks showed attenuated inflammatory gene expression and protection against insulin resistance and hyperglycemia, compared with unsupplemented mice.
The human glucose findings in the review belong to other berries. Three months of 200 mL a day of chokeberry juice lowered fasting glucose, HbA1c, and lipids in 62 patients with type 1 or type 2 diabetes (P < 0.001), while two cranberry trials in people with type 2 diabetes changed neither glucose nor HbA1c. The proposed carbohydrate-digestion mechanism is also non-blueberry and preclinical: chokeberry fruit extract reduced intestinal disaccharidase activity in a rat model.
What this page does not claim: a glycemic index or glycemic load for blueberries, or any human blood-sugar effect. Neither source reports a glycemic measurement for any berry, so the low-glycemic-load line previously carried here has been removed rather than kept as background, and the claim now states the mouse result at its own scope.
The curated colorectal-cancer source is a narrative review of anthocyanins, not a systematic one: it reports no search strategy or inclusion criteria. Its blueberry content is entirely in cell lines. Blueberry extract inhibited proliferation of Caco-2 cells with an IC50 of 0.78 mM and inhibited proliferation while promoting apoptosis in HCT116 and HT-29 cells; among the fractions compared, the blueberry anthocyanin extract showed the greatest antiproliferation effect, and an extract of Chinese blueberry (Vaccinium uliginosum L.) inhibited proliferation in DLD-1 and COLO205 cells. The review reports no colorectal carcinogenesis animal study using blueberry; its only in vivo blueberry result is a microbiome one, in which blueberry anthocyanin extract increased Bifidobacterium, Lactobacillus, Roseburia, Faecalibaculum, and Parabacteroides in mice.
The mechanisms the review catalogues across all anthocyanin sources — ROS scavenging, cell-cycle arrest, caspase-3 activation and a raised Bax/Bcl-2 ratio, PI3K/AKT and NF-kappaB suppression, MMP-2 and MMP-9 effects on invasion, DNMT1 and DNMT3B demethylation, and microbiota shifts — come from cell lines and rodent carcinogenesis models, not from people.
Human data are epidemiological or biomarker-level. A meta-analysis put anthocyanin intake against colorectal cancer risk at a pooled relative risk of 0.78 (95% CI 0.64 to 0.95); an Italian case-control study reported a hazard ratio of 0.78 (0.61 to 0.99) and a Chinese case-control study an adjusted odds ratio of 0.80 (0.64 to 1.00), while other cohort and case-control studies found no association at all. Those exposures are total anthocyanin intake, not blueberries. The only interventional human data are biomarker measurements in colorectal cancer patients given black raspberry powder — decreased methylation of SFRP2, PAX6a, and WIF1, regulation of beta-catenin and E-cadherin, and decreased GM-CSF and IL-8 — plus microbiota changes with anthocyanin-rich apple products. The authors rest their conclusion on epidemiology, cell lines, and animal studies, and name low bioavailability, with first-pass estimates around 1%, as a large pitfall in the field.
What this page does not claim: that blueberries prevent or treat cancer in humans. No curated source reports a cancer outcome in people from any anthocyanin source, let alone from blueberries, and the cell-line effects above were produced at millimolar concentrations. The pterostilbene material previously carried here, including its bioavailability advantage over resveratrol, its SIRT1 and NF-kappaB signaling, and HDAC inhibition, has been removed rather than kept as background: pterostilbene appears nowhere in either curated source, and neither does HDAC.
Cognition and brain health. Neither curated source addresses cognition. The cardiovascular review’s scope is cardiovascular and metabolic and it mentions neither the blood-brain barrier nor BDNF; the second source is confined to colorectal cancer. The statements previously carried here — that blueberries have the strongest human neuroprotective evidence among berries, that anthocyanins cross the blood-brain barrier and accumulate in the hippocampus and prefrontal cortex, that they modulate BDNF and suppress neuroinflammation, that cohorts associate blueberry intake with slower cognitive decline, and that 12 weeks of daily supplementation improves memory and processing speed in older adults — have been withdrawn rather than restated behind an “emerging evidence” hedge, and the cognitive claim has been removed. That is a sourcing gap, not a refutation: it means no evidence for it is curated here.
Practical guidance. Neither source recommends a serving size, compares fresh with frozen blueberries, or compares wild lowbush with cultivated highbush varieties; “wild blueberry” appears only as a description of the freeze-dried powder used in two trials, and “highbush” only in the funding disclosure. The gram amounts on this page are trial doses — 250 g of whole blueberries a day, 100 g of freeze-dried powder with a meal — not intake advice. The one preparation statement either source makes is the cardiovascular review’s own recommendation that, because pressing, pasteurization, and drying reduce polyphenol and vitamin content, fresh or frozen whole berries may be better than juices or extracts, while encapsulated supplements may suit the management of specific metabolic conditions. The half-cup serving figure, the fresh-equals-frozen equivalence, the wild-versus-cultivated density comparison, and the low-calorie, low-glycemic serving suggestions have been removed.