More than five cups of coffee a day harms bones, a study says. Its own test failed to confirm that threshold

Over the past two weeks, Slovak and Czech media have published about a dozen articles about coffee, and some of them cited a specific threshold: more than five cups a day harms bones. That is exactly what the authors of the Australian study they rely on write, even though the overall link between coffee and bone density did not show up at all in their results, and their own test failed to confirm that threshold. The oversimplification, then, does not begin in newsrooms but in the summary of the paper itself.

What the Australian paper actually says

The study was authored by Ryan Yan Liu and Enwu Liu of Flinders University and was published on November 23 last year in the journal Nutrients. It draws on the American Study of Osteoporotic Fractures, which has been following 9,704 women aged 65 and over since the 1980s. After exclusions, 24,638 observations remained from four follow-up visits, and the analytic sample was tracked for an average of 8.1 years. Nutrients ↗

No overall link was found between coffee consumption and bone density. The five-cup threshold comes from a supplementary curve analysis, which visually suggested that consumption above five cups a day might be linked to lower density. However, the tests did not show a significant nonlinear relationship for either coffee or tea — the p-values for nonlinearity were all above 0.05. So this is not a confirmed threshold of harm, but a trend that can be seen on the graph and that the statistics did not confirm. Nutrients ↗

This caveat, however, did not make it into the summary. In both the abstract and the conclusion, the authors write that consumption above five cups a day may be harmful to bone density, even though they no longer mention their own test there. The same paper thus reads differently in two places — cautious in the results, resolute in the summary, which is what most people read. Nutrients ↗

The finding about alcohol, which spread through Slovak articles as a warning to women who drink alcohol, should likewise be treated with caution. It concerns lifetime alcohol intake, and the authors themselves write that the subgroup findings are likely underpowered, may be due to chance, and serve primarily to generate hypotheses. Nutrients ↗

The other half of the same paper concerns tea and is the best-supported part. Women who drank tea had higher total hip bone density at the last check-up, specifically 0.718 versus 0.715 grams per square centimeter. The difference is thus three thousandths, and the authors acknowledge that it falls below the threshold of significance for decisions about an individual patient, while also arguing that from an epidemiological and public-health perspective even such a small shift can be meaningful. The sample consisted almost exclusively of white American women, and there were many missing values in the data. Nutrients ↗

The press release repeated it

Flinders University's press release from December 11 last year picked up exactly that from the conclusion. It states that drinking more than five cups a day was associated with lower bone density, suggesting that excessive intake may be harmful. There is nothing in it about the nonlinearity test. Flinders University ↗

Slovak coverage arrived nine months later and adopted the same shape. On September 8, TV Markíza wrote that scientists had uncovered the dark side of excessive coffee drinking. TV Markíza ↗ Three days earlier, Pravda chose a more cautious verb — according to it, too much coffee may harm bones — and added in the text that one to three coffees a day are apparently not a cause for concern. Pravda ↗ In neither case, however, does the reader learn that the five-cup threshold in the study is merely an observation from a graph.

The Finnish study measures something different

The second paper written about in the same period was published on July 16 in the academic journal European Journal of Nutrition and comes from the University of Oulu. In it, researchers processed data on 2,264 people aged 46, who have been followed since birth by the Northern Finland Birth Cohort 1966. European Journal of Nutrition ↗

Participants were divided into four groups by number of cups: zero, one to two, three to four, and five or more. The higher-consumption groups had, on average, less total and visceral fat and more skeletal muscle mass, even though their body mass index was comparable. In men, this was accompanied by a more favorable glucose and insulin profile. Univerzita v Oulu ↗

Hormones attracted the most attention, and that is precisely where the result is least straightforward. Men with higher consumption had higher total and bioavailable testosterone; for total testosterone, the difference works out to 0.29 nanomoles per liter per daily cup. At the same time, they had higher SHBG — the protein that binds sex hormones in the blood — and lower free testosterone and free androgen index. In women, no link with total testosterone appeared; only higher SHBG and lower free testosterone remained. Živě ↗

""Coffee is consumed by millions of people every day, yet we still know surprisingly little about how it relates to our metabolism and hormones."" said lead author Luca Verroest. According to him, a hormonal signature stood out from the results that did not disappear even after accounting for BMI and lifestyle factors. Univerzita v Oulu ↗

What matters is that the five-cup threshold here means something completely different. It is not a point beyond which something worsens, but the highest comparison category. So the same number in the two papers does not refer to the same thing, and neither of them demonstrates a causal relationship. The Finnish authors state this explicitly — the cross-sectional design rules out a causal conclusion, and reverse causality cannot be excluded. Hormones were measured once, participants self-reported their consumption, and although they recorded the type of coffee for 96 percent of them, they did not collect data on bean variety, roast level, preparation details, or additives such as milk and sugar. Finland is, moreover, the country with the highest coffee consumption in the world, approximately 11.8 kilograms per person per year according to the European Coffee Report for 2022 and 2023. European Journal of Nutrition ↗

What this means for the cup in front of you

Neither of those papers can, on its own, be translated into a recommendation, but practical advice did come from doctors during the same period. Gastroenterologist Ladislav Kužela puts the safe daily limit for healthy adults at approximately 400 milligrams of caffeine from all sources, which is roughly three to five cups, though the exact conversion depends on size and preparation. According to him, how a person feels also matters. "If you feel your heart pounding or don't feel well after another coffee, you should stop," he says. He notes that caffeine's effect can still be felt in the body six hours later, so a late coffee should be considered with sleep in mind. In the case of worsened reflux, he says it is worth first checking what the coffee was drunk with, as a heavy meal alongside it is often the culprit. SME ↗

Urologist Ján Švihra debunks probably the most widespread myth. According to him, a normal amount of coffee does not significantly dehydrate the body — caffeine has a mild diuretic effect, but the water contained in coffee generally more than compensates for that loss, and regular drinking also builds up tolerance. Coffee can count toward fluid intake, though not on a one-to-one basis. Startitup ↗

Attention shifts to the additives

A similar shift from the beverage itself to what surrounds it applies to cancer worries as well. The American outlet Parade approached two oncologists, and both talk about what we add to our coffee. Clinical oncologist Asma Dilawari points to non-dairy coffee creamers, which often contain hydrogenated oils, added sugar, or sweeteners, though she does not call them a major health problem in themselves. Oncologist and hematologist Jitesh Joshi adds that strong direct evidence that creamers, additives, or artificial sweeteners themselves cause cancer is lacking. According to him, what matters is the long-term diet, not a single bottle of creamer. Startitup ↗

Timing relative to meals falls into the same category. Coffee can impair the absorption of non-heme iron, meaning the kind from plant sources and fortified foods, and it is also mentioned in connection with calcium and zinc. Research has not yet established a uniform time gap, so it is more sensible not to drink it right along with such a meal than to stick to a specific number of minutes. Dobré noviny ↗

Where you keep it at home

Meanwhile, something entirely mundane affects the taste. According to Luke Dalton of the company Contact Coffee Co, one of the biggest mistakes is storing beans right next to a kettle, a toaster, or the coffee maker itself. It looks like the best spot, but according to him, the heat and steam from these appliances make coffee lose its freshness much faster. For the same reason, a cabinet above the stove is just as bad a spot as a shelf next to an electric kettle. T3 ↗

Illustrative photo: espresso on a kitchen counter. Author Daniel Case, Wikimedia Commons, license CC BY-SA 4.0.