Same number of snacks, different snacks: young Finns with coeliac disease had ice cream three times as often and sugary soft drinks half as often

The usual worry about a gluten-free diet is the meals: the bread, the pasta, what gets swapped for what. A group of Finnish researchers looked at the gaps between them instead. They sent a web survey to the 1,257 members of the Finnish Celiac Society aged 16 to 30, and to about 6,500 students at a university of applied sciences and a vocational school in Seinäjoki. The analysis compared 210 people with coeliac disease on a gluten-free diet against 143 with no coeliac disease and no special diet.

The first result is a non-result. Snacking was just as common in both groups: 78% of the coeliac group and 75% of the others had one or two snacks a day, and the difference between the groups was nowhere near significant (P = 0.931). Replacing lunch or dinner with a snack didn't differ significantly either. The coeliac group actually reported slightly more meals, 3.4 a day against 3.0, though the authors say that gap is partly down to who answered, since more of that group were women and fewer were students.

The contents of the snacks are where the groups parted. Counted in times per week, the coeliac group had ice cream 0.9 times against 0.3, and low-fibre bread 1.6 times against 0.8. They had sugary soft drinks 0.6 times against 1.2. All three gaps held when the researchers widened the comparison to include people on other special diets, though ice cream became 0.9 against 0.4 and soft drinks 0.6 against 1.0. High-fibre bread barely differed at 2.6 against 2.5, so on these counts the low-fibre bread doesn't look like a straight swap for the high-fibre kind. Nor was there a general sweet tooth: sweets (1.2 vs 1.1), chocolate (1.0 vs 1.3) and biscuits and snack bars (0.8 vs 0.8) showed no significant difference.

Then there is lunch. Among those who answered the question, 18.3% of the coeliac group ate in a school or workplace canteen every day, against 38.4% of the others, and 23.4% never did, against 11.6%. That difference survived separate adjustments for gender, work or study status and the size of the place they lived. They were also far less likely to get snacks from those canteens (16% vs 43%), and most of the 157 who answered felt the gluten-free snack choice in shops and cafeterias wasn't good enough. There's a Finnish wrinkle here: student lunches are heavily subsidised, and 83% of the comparison group were students against 51% of the coeliac group, so cheap lunches may explain part of the gap as well as the menu.

This is a one-off anonymous survey, and it has real weaknesses. The response rate was about 17% in the coeliac group and 3% in the comparison group, the questionnaire was piloted but not formally validated, and everything is self-reported. It counts how often foods were eaten, not how much, and says nothing about whether people's diets were adequate overall. The groups also differed in ways adjustment may not fully fix, and three snack differences, including soft drinks, disappeared among the men when the sexes were analysed separately. It can't say that coeliac disease or the diet caused any of this. What it does offer is a useful nudge for anyone helping a teenager take over their own gluten-free diet: the conversation about fibre probably shouldn't stop at the dinner table.