Emulated trial in 64,000 Swedish adults credits strict Nordic eating with 4% lower 24-year mortality

ScienceAlert's account does not say whether the 4 percent is a relative or an absolute difference, or how many people in the cohort died over the 24 years ^[4]. Without that baseline, the estimate cannot be turned into deaths avoided per thousand people. The two comparisons do allow one internal check. They sit 2 points apart ^[18]. On the model's own terms, following the 2023 recommendations instead of changing nothing is worth roughly 2 points. The Nordic diet's extra margin over those recommendations is about twice that ^[19].

The design explains the caution. The Danish team took existing health records on more than 64,000 middle-aged and older Swedes and analysed them as if people had been assigned to different diets in a long trial ^[1]. ScienceAlert calls it a statistical shortcut that avoids waiting for a 20-year trial to finish, and is clear that it is not a clinical trial ^[9]. Randomization makes groups alike in everything except the intervention. An emulation built from records has to assume it has measured the differences that matter.

Every arm is also hypothetical: a person kept strictly on one diet for decades ^[8]. For any diet, the real-life bottleneck is whether people keep eating it, and this analysis assumes that problem away. I'd expect people who drift on and off the diet to see less than the modelled figure, if the effect holds at all.

The liver results show how much the choice of endpoint matters. In a recent randomized trial of Swedish adults with prediabetes or type 2 diabetes, the healthy Nordic diet beat the 2012 Nordic recommendations on liver fat, liver enzymes, body weight, glucose and lipids ^[10]. "The healthy Nordic diet gave the best results in the study participants with diabetes; just over 20 percent of their liver fat was reduced, and blood sugar (glucose) control improved over one year," Ulf Risérus of Uppsala University, who led that trial, said in 2025 ^[11]. "More than half of the participants also saw a remission of their fatty liver disease," he said ^[12]. Over decades, against the newer 2023 recommendations, the emulation found no clear reduction in serious liver outcomes ^[6]. The trial used the 2012 advice as its comparator and measured liver fat over a year. The emulation used the 2023 advice and serious outcomes over decades ^[10][6].

The subgroup pattern fits a diet that does more where there is more risk to remove ^[7]. It comes from the same hypothetical framework, and ScienceAlert notes that this kind of nuance will need verifying in future trials ^[8].

The comparisons reported were against the 2023 Nordic recommendations and against no intervention ^[4][5]. The link to the Mediterranean diet rests on shared ingredients. The Nordic version swaps olive oil for rapeseed oil, wheat and rice for oats and rye, and oranges and bananas for apples and pears ^[15]. ScienceAlert notes that more unites the two diets than separates them, and that the Mediterranean diet is also linked to substantially lower risk of disease and death ^[16]. It also has a far larger body of long-term research behind it ^[14]. According to ScienceAlert, Michael Fridén of Aarhus University, the lead author, sees strong grounds for suspecting that traditional Nordic eating, heavy on berries, unsaturated fats and fiber, benefits human health ^[17].

I think the study supports the healthy Nordic diet as a sound pattern for Swedes, with a modest modelled edge over national advice. Treating it as a stand-in for the Mediterranean diet would take a direct comparison between the two, and this analysis ran its arms against Nordic advice and against no change ^[4][5].