Sugar & sweetened drinks

Sugar is the food most people assume is bad for their joints, and the assumption is running ahead of the data. What exists is one prospective finding with an odd shape, one larger survey that found nothing, and an animal model that cannot separate sugar from the fat it is fed with.

The soft drink analysis

Two thousand one hundred and forty-nine people with radiographic knee osteoarthritis were followed for four years, with medial joint space width measured quantitatively from plain films. Among men, mean joint-space loss rose with baseline soft drink intake: 0.31 mm at none, 0.39 mm at one or fewer a week, 0.34 mm at two to four, and 0.60 mm at five or more, adjusted for body mass index. Among women there was no association at any intake.

Read carefully, that gradient has three features worth noticing. It is not monotonic — two to four drinks a week did better than one or fewer. It is confined to one sex, with no mechanism offered for why. And it was steeper in non-obese men, with only the highest intake separating among obese men, which runs against the obvious explanation that sugary drinks harm knees by adding weight.

The Korean national survey

Five thousand five hundred and three adults in the Fifth Korea National Health and Nutrition Examination Survey had knee radiographs and beverage-frequency data. Soft drinks showed no significant relationship with radiographic knee osteoarthritis in either sex. Neither did green tea, nor milk. The only association anywhere in that survey was coffee, inversely, in women — pointing the opposite way from the genetic evidence covered two entries along.

The pair of findings that should be read together

The milk analysis in this section comes from the same Osteoarthritis Initiative dataset, analysed by an overlapping group of authors, using the same joint-space measurement over the same four years. It found a protective dose-response in women and nothing in men. The soft drink analysis found a harmful dose-response in men and nothing in women.

Neither has been replicated. Two beverage effects from one cohort, each present in exactly one sex and each in the opposite sex from the other, is the pattern a dataset produces when it is asked many questions — and it is a better reason for caution about both than any single criticism of either.

The animal work, at two different joints

Rats on a high-fat, high-sucrose diet develop knee joint damage, and prebiotic fibre prevents it. When the same group looked at the shoulder in the same model, they found both sexes appeared protected at that joint, while male rats developed bone lesions and females did not.

So a metabolic diet does not damage all joints equally, and the difference between a knee and a shoulder in a quadruped is mostly load. That is a useful correction to the idea that sugar attacks cartilage wherever it finds it. It is also a reminder that these diets are fat and sucrose together — no animal study in this literature isolates the sugar.

The route nobody has tested

There is a mechanism sitting right next to this entry that has never been connected to it. Fructose is a considerably more potent glycating sugar than glucose, and cartilage collagen accumulates glycation crosslinks essentially for life because it barely turns over. That is the argument of the dietary AGEs entry, and the bridge — dietary sugar intake measured against cartilage glycation in humans — has not been built by anyone.

Practical notes

There is no trial of cutting sugar with a joint endpoint, so nothing here supports a target. What the evidence in this section does support, clearly and repeatedly, is that body weight matters to a knee, and sugar-sweetened drinks are a well-characterised route to carrying more of it. That argument is made in the body weight entry and it does not need this one.

What would change this entry

A replication of the sex split, in the cohort that produced it. Thirteen years on, the absence of one is starting to be the finding.