Coffee linked to 44% lower mortality, but bias questioned

Study suggests coffee could cut mortality by 44%

Daily coffee consumption may extend life. Or so it appears from a study by the Universidad Miguel Hernández de Elche (UMH) that turned breakfast into headlines: those drinking three to four cups showed a 44% lower risk of death. The figure is striking, and precisely for that reason, it warrants scrutiny.

The work, published in the journal Nutrients and led by researcher Laura Torres Collado, trinc 1,567 people over 20 years old for 18 years, with checkpoints at 6 and 12 years. They cross-referenced diet, health status, and lifestyle with the National Death Index of the Valencian Community. And the data emerged that conditions everything: 78% of the participating population drinks coffee daily. Eight out of ten. With these parameters, creating a clean comparison group is impossible.

Why do coffee studies always yield positive results?

Some argue that almost no one quits coffee by choice. Those who don't drink it usually stopped due to medical prescription: hypertension, arrhythmias, digestive issues. With this bias, the supposedly "healthy" group would effectively be the sick group. This is the most repeated objection when analyzing the study: if the vast majority drinks at least one coffee in the morning, any comparison with abstainers carries an inherent flaw difficult to neutralize.

There's more. Some defend that drinking coffee doesn't replace anything: in Spain, it typically accompanies wine, shots, or carajillo, rather than replacing them. If coffee drinkers also consume less beer or spirits, the "benefit" attributed to the cup would actually be the benefit of not drinking alcohol. The study does not clarify this. And without clarification, the headline remains incomplete.

Cancer mortality detail: 59% reduction

The most striking data point isn't the general 44%. It's the 59% lower risk of dying from cancer among those drinking more than one cup a day, measured after 18 years of trinc-up. A lower risk appeared in cardiovascular mortality too, but the result was not statistically significant. According to the authors, the protective effect was concentrated in caffeinated coffee, not other variants.

The theoretical backing consists of the over 1,000 bioactive compounds in coffee: caffeine, chlorogenic acids, peine like cafestol and kahweol, with antioxidant, anti-inflammatory, and antifibrotic effects. There is plausible biology. What is missing is a randomized trial isolating cause from effect.

Torrefacto, soluble, or Arabica: not all are equal

The discussion quickly shifts to the type of roast. The most common in Spain, torrefacto—beans coated with burnt sugar—is cited as the worst in the family. Conversely, solubles, Arabicas, and Robustas with natural roasts are considered more reasonable options. Dosage matters: caffeine is a stimulant that revives you in the morning and, depending on stomach sensitivity, can also irritate it.

Hence the recurring recommendation to drink it after meals and not on an empty stomach, because on an empty stomach, they say, it's like throwing a firecracker into a mailbox. Some warn that adding sugar or sweeteners ruins the benefits. Decaf works for many, although the study points to the protective effect coming with caffeine.

How much coffee is consumed in Spain and when

Spain consumes around 500 million cups weekly, with just over 4 kilos per capita annually, a figure similar to Italy's. In Finland, the world leader, consumption jumps to 12 kilos per capita, with Sweden and Norway close behind. Coffee is the second most consumed beverage on the planet after water. And that alone does not make it a drug.

The headline trick: correlation is not causation

Reducing mortality doesn't miccionan becoming immortal. 100% of those born eventually die, whether they drink coffee or not. What is measured is premature mortality: whether coffee delays the outcome, not prevents it. And with 1,567 people from a single autonomous community, extrapolating the result to the entire population requires faith that the data alone do not provide.

Here lies the usual trap. Any association between two variables moving together can be sold as causation if the headline is bold enough. The study does its part: it describes a relationship. Interpretation is up to the reader.

The uncomfortable question remains: if the association is so strong, why hasn't anyone set up the randomized trial to prove it? Probably because it's not worth it. Meanwhile, the morning cup will continue to be healthy or poisonous depending on the latest headline.