Biscuits: what the grain data says and what it cannot

Across seventeen prospective cohorts and more than 875 000 people, the pooled analysis did not find a link between refined grain intake and cardiovascular disease, stroke or heart failure. Read the numbers rather than the headline. For cardiovascular disease the hazard ratio was 1.08 with a confidence interval of 0.99 to 1.18, which touches one, and the cohorts disagreed with each other at an I squared of 70 percent. A link not shown is a different sentence from no link, and on a page about sweet biscuits that difference is most of the page. The same null held whether the analysis counted staple grain foods only or added cakes, cookies and doughnuts. A cohort records what people say they eat, so it answers a question about diets rather than about biscuits. Look at what was actually fed in the trials on this page and the same gap opens again. A whole-grain breakfast biscuit with a glycaemic index of 54. A biscuit fortified with plant sterols. A biscuit designed to carry micronutrients to undernourished children. Only one trial here tested the kind of sweet biscuit you buy in a shop, and it did so by swapping them for almonds, calorie for calorie. The biscuits lost on blood glucose and on how much people ate afterwards.

The numbers, per 100 g

How much is actually in it

• Pooled across 125 studies, plant sterols in bread, biscuits and cereals lowered LDL less than the same sterols in butter, margarine and spreads. — adults given phytosterol-fortified foods (n = 125 studies) Evidence A source
  Phytosterols lowered LDL cholesterol by 0.55 mmol/L on average, with a larger fall at higher daily dose; the food format bread, biscuits and cereals produced a smaller reduction, by 0.14 mmol/L, than the predominant format of butter, margarine and spreads, while treatment duration, intake pattern, number of daily intakes and statin use made no significant difference

What a real portion looks like

• Matched calorie for calorie against almonds, sweet biscuits raised blood glucose more and left people eating more for the rest of the day. — adults, 25 men and 75 women, in New Zealand (n = 100 with available data of 106 randomised) Evidence B source
  Mean area under the blood glucose curve was lower for almonds than for biscuits by 53 mmol/L.min (95% CI 45 to 61, p<0.001); appetite ratings barely differed and lunch intake did not differ, but total energy intake after the snack was 638 kJ lower on the almond day

What your body absorbs

• A whole-grain biscuit measured as a low glycaemic index food, and adding peanut butter to it pushed the index lower still. — nineteen healthy men, nine Chinese and ten Indian, mean age 24.7 years, mean BMI 21.7 (n = 19) Evidence B source
  Whole-grain biscuit gave a glycaemic index of 54, and the same biscuit filled with peanut butter gave 44, both in the low range, against 85 for rice idli and 76 for rice dosa; glycaemic index correlated negatively with the protein, fat and fibre content of the food

What cooking and storage change

• Baking does not only take things out of flour, it puts things in, and acrylamide is the main one. — measured samples of wheat-based foods including bread, cookies and pasta, not people (n = 28 studies) Evidence C source
  Acrylamide, 5-hydroxymethylfurfural and polycyclic aromatic hydrocarbons form during processing; acrylamide carried the highest pooled concentration at 136.29 µg/kg, followed by HMF at 70.59 µg/kg and PAHs at 0.11 µg/kg, with 69.7% of studies measuring acrylamide in bread, cookies and pasta

What people believe that the data does not support

• Across seventeen cohorts and nearly nine hundred thousand people, the pooling did not find a link between refined grain foods, the category biscuits sit in, and heart disease. — adults in prospective cohorts (n = 17 cohorts, more than 875 000 participants) Evidence C source
  Refined grain intake was not associated with cardiovascular disease (HR 1.08, 95% CI 0.99 to 1.18), stroke (HR 1.06, 0.92 to 1.23) or heart failure (HR 0.95, 0.77 to 1.16), and the same null result held whether analyses covered staple grain foods only or included indulgent grain foods such as cakes, cookies and doughnuts

Who this works differently for

• In places where children are undernourished, the biscuit is used deliberately as a way to carry nutrients, and the evidence behind that use is thinner than the practice. — children at risk of or affected by stunting (n = 11 studies) Evidence C source
  The included studies support biscuits as a carrier for essential nutrients, while the authors conclude that more well designed and standardised studies are needed before guidelines for formulating and implementing biscuit-based interventions can be set

What is still unknown

• For whole regions of the world nobody has measured what is actually in the biscuits on the shelf. — commercial wheat-based food samples worldwide (n = 28 studies) Evidence C source
  Acrylamide is the most researched processing contaminant, and for some regions of the world no studies of commercial samples were found at all; the authors add that levels change with process parameters and with the ingredients used

The bottom line

Biscuits are food you eat because you want them, and the portion is the decision. The cohort data did not find a link with heart disease, and a link not found is a thinner thing than an all-clear. Two details are worth taking away. The first is that what is in the biscuit changes what it does to your blood sugar. Glycaemic index fell as protein, fat and fibre rose, which is why a whole-grain biscuit measured 54 and the same biscuit filled with peanut butter measured 44, both in the low range. The second is about anything added to a biscuit for effect. Pooling 125 studies of plant sterols, the food formats bread, biscuits and cereals came out 0.14 mmol/L behind butter, margarine and spreads on LDL cholesterol. Two warnings travel with that figure. It is a comparison between food formats inside the pool rather than a trial that randomised a biscuit against a spread. And the confidence interval printed beside it in the paper, minus 8.71 to minus 2.16, sits on a different scale from the estimate and does not contain it, which we cannot explain from what the abstract says. Take the direction and leave the decimal alone. A biscuit looks like a harder vehicle than a spread. Baking also puts things in. Acrylamide, hydroxymethylfurfural and polycyclic aromatic hydrocarbons all form during processing of wheat foods, pooled at 136, 71 and 0.11 µg/kg. Those pooled figures describe a category rather than your packet, because levels shift with the process and the ingredients, and for some regions of the world nobody has measured commercial samples at all. Where biscuits are used on purpose to carry nutrients to children at risk of stunting, the reviewers of eleven studies say the practice has run ahead of the evidence and ask for standardised trials before anyone writes guidelines.

Sources

1. cky-r1-05 · systematic review and meta-analysis
   The food format "Bread, biscuits, cereals", conditioned a lower decrease of 0.14 mmol/L (95%CI -8.71 to -2.16) in LDL-C levels, compared to the predominant food format group of "butter, margarine, spreads".
   Nutr Metab Cardiovasc Dis, 2023 · checked 2026-09-22
2. cky-r1-01 · acute randomised crossover trial
   The mean area under the blood glucose response curve was statistically and practically significantly lower for almonds than biscuits (mean (95% CI) difference: 53 mmol/L.min (45, 61), p < 0.001).
   Int J Environ Res Public Health, 2021 · checked 2026-09-22
3. cky-r1-02 · randomised crossover trial
   Western breakfast foods were all of low GI (whole-grain biscuit=54 (sem 5); whole-grain biscuit filled with peanut butter=44 (sem 3); whole-grain oat muesli=55 (sem 4); whole-grain oat protein granola=51 (sem 4); whole-grain protein cereal=49 (sem 3))
   Br J Nutr, 2018 · checked 2026-09-22
4. cky-r1-03 · systematic review and meta-analysis of occurrence data
   The highest concentration was for acrylamide (136.29 µg·kg-1) followed by HMF (70.59 µg·kg-1) and PAHs (0.11 µg·kg-1).
   Crit Rev Food Sci Nutr, 2023 · checked 2026-09-22
5. cky-r1-04 · meta-analysis of prospective cohort studies
   Meta-analyses of hazard ratios (HR) and 95% confidence intervals (CI) obtained from 17 prospective cohort studies (>875,000 participants) indicated that refined grain intake was not associated with risk of CVD (HR = 1.08, 95% CI, 0.99-1.18, I2 = 70%; 9 cohorts), stroke (HR = 1.06, 95% CI 0.92-1.23, I2 = 70%; 9 cohorts), or heart failure (HR = 0.95, 95% CI 0.77-1.16, I2 = 10%; 5 cohorts).
   Trends Cardiovasc Med, 2024 · checked 2026-09-22
6. cky-r1-06 · systematic review
   The studies revealed critical findings regarding the utilization of biscuits as a carrier in providing essential nutrients. Nevertheless, more well-designed and standardized studies are required to establish clear guidelines for the formulation and implementation of biscuitbased interventions.
   Afr J Reprod Health, 2024 · checked 2026-09-22
7. cky-r1-07 · systematic review and bibliometric analysis
   Acrylamide is the main processing contaminant researched, and no studies on the subject have been found in commercial samples in some regions of the world.
   Crit Rev Food Sci Nutr, 2023 · checked 2026-09-22