Same factory, 265 more calories a day. In 52 Turkish men, those on rotating shifts ate more fat, yet their diets looked no more 'inflammatory' than the day crew's

The usual story about shift work and food goes like this: odd hours lead to worse eating, worse eating feeds inflammation, and inflammation wrecks sleep. Two nutrition researchers from Ankara University and Acibadem University put that chain to the test at a defence factory in Ankara, comparing 26 men who always worked 08:00 to 16:00 with 26 who rotated each week through day, evening and night shifts. Everyone had at least a year on his schedule and no diagnosed chronic disease. The paper appeared in PLOS One on 22 October.

The first link in the chain held up, at least in part. A dietitian took three separate 24-hour food recalls from each man, covering working days and a day off. The rotating group reported about 2,428 kcal a day against 2,163 for the day workers, a gap of 265 kcal. Saturated fat ran about 12 g a day higher (49.4 g against 37.4 g), and the median cholesterol intake was roughly double (577 mg against 266 mg). Fat made up 45.5% of the rotating men's energy and 41.7% of the day men's. They also took in more iron, calcium, folate, vitamin A and phosphorus, which the authors warn may simply reflect eating more food overall.

The second link did not. The Dietary Inflammatory Index, which scores a diet from its nutrients and a few foods such as garlic and tea, was the study's main measure, and the researchers had predicted a higher score for the rotating group. It came out at -0.2 for them and +0.2 for the day workers, a difference well within chance. The authors suggest one reason may be the institutional meals both groups shared at the factory, which could have narrowed the gap. Blood pressure, blood sugar, blood fats, C-reactive protein and homocysteine showed no link with the score either. Sleep scores leaned the expected way (6.0 against 4.7 on a scale where higher is worse), but that gap was not statistically significant.

The third link is where a quick reading could mislead. Across all 52 men, a higher inflammatory score went with worse sleep (r = 0.28, p = 0.046), and the pattern looked stronger among the rotating group. After allowing for age, BMI, smoking, physical activity and work schedule it was no longer statistically significant (p = 0.112). This was one snapshot, in one workplace, of men only, with food and sleep both self-reported, so it cannot say that diet affected sleep or the reverse. The sample was sized to catch a large difference in the index, so a smaller one could have gone unseen, and with so many nutrients compared the authors ask for caution there too.