The big medical news story of the past week — at least in the “non-Ebola” category — is a study suggesting that increased consumption of milk is associated with increased rather than decreased risk of fracture, and increased risk of all-cause mortality. Given all the positive press milk and dairy have had recently, this finding is provocative to say the least.

The study, published in the British Medical Journal, examined the associations between dairy intake and health outcomes in two cohorts, one of men and one of women. Overall, more than 100,000 people were included. Researchers followed the women for a period of roughly 20 years, and the men for a period roughly half that.

The researchers collected some information about health and lifestyle, and details of dairy intake from food frequency questionnaires. They analyzed the association between number of glasses of milk consumed per day and health outcomes, in particular fractures and death. In women, more reported milk intake was associated with higher rates of both fracture and death. In men, there was no apparent protective effect — but the adverse association was not significant.

Does this study mean that drinking milk increases the risk of fracture or death? No. It merely provides a reason to be cautious about presuming we can count on the opposite, protective effects.

This was an observational study, meaning there was no intervention. Participants provided information, but did not receive treatment or undergo randomization. Such studies can show associations, but are not designed to prove cause and effect.

A particular limitation of such studies is reverse causation. This is when a study shows an association between X and Y, suggesting that X causes Y, while in reality, Y causes X. In this case, while increased milk consumption could increase fracture and death risk, it is also quite possible that increased fracture risk — based on personal history — prompted more milk intake in an attempt to treat that risk. The authors acknowledge this weakness.

The study noted that while milk intake was associated with increased rates of fracture and mortality in women, that association was not seen with yogurt and cheese. These fermented dairy products actually showed the opposite, protective effects, albeit modest. This may argue against reverse causation, but does not rule it out entirely.

Another potential weakness, also acknowledged by the authors, is confounding. This is when some other factor actually accounts for the outcome, rather than the variable under scrutiny. It may be that people who drink more milk differ in other ways from their counterparts as well. These other differences, rather than milk per se, might actually account for the health outcomes observed.

There are, however, some mechanisms that could account for adverse effects of milk intake. The authors looked particularly at the pro-inflammatory and oxidative effects of milk sugar. These effects could be factors and might explain why yogurt and cheese were not implicated in the same associations.

In terms of fracture risk, it may also be that those who drink more milk have higher overall protein intake than those who drink less. Protein is acidic, and high intake can be associated with reduced bone density as alkaline components of bone are released to buffer the acid. Of note, population studies around the world have long noted the lowest rates of fracture not in those who drink the most milk, but those who are generally physically active, and eat a mostly plant-based diet.

Despite all of the recent attention to saturated fat, most of it misguided, this new paper did not look at dairy fat per se. The findings presumably relate to full-fat dairy as the default choice in Scandinavia, but this is uncertain. The authors do not report the type of milk consumed, just the quantity.

The effects of dairy on health are controversial; I have addressed them here before. This has something to do with the composition of dairy, the diversity of choices within the category, and the effects that varying levels of dairy intake have on varying levels of intake of other choices. Increased milk consumption, for example, may be more likely to propagate health benefits in a culture where milk replaces soft drinks, than in one where milk replaces water.

Where, then, does this study leave us? If you drink milk routinely, there is no clear basis in this study to stop. There is no guidance here either about the relative pros and cons of full-fat or fat-reduced dairy. If you are thinking to start consuming more dairy to reduce fracture risk, or improve overall health, this study may be cause for some doubt. That said, dairy intake in the context of a healthy lifestyle overall is unlikely to confer significant risk, just as dairy intake in the absence of a healthy lifestyle is unlikely to confer any meaningful benefit.

Eat well overall, be active and don’t smoke. These behaviors will reliably reduce the risk of both fracture and premature death. Consume dairy if so inclined based on your personal priorities and preferences — but not because it is certain to improve health outcomes. Avoid dairy as inclined for the same reasons — but not because it poses any certain health threats.

The new study is principally useful for questioning the answers we may have thought we had. It is a bit less useful for answering questions we have still.

More from U.S. News

6 Fruits and Vegetables You Should Eat This Fall

10 Ways to Lower Your Risk of Stroke

16 Health Screenings All Women Need

Milk and Mortality: Questioning the Answers originally appeared on usnews.com