This study set out to compare the effects of adding about 500 calories a day of either fruit or nut intake to the existing diet of some young lean active healthy Swedish students. The primary objective was to see if the sugar in the fruits had any adverse effect.
As is often the case in free living dietary studies the outcome was not exactly as intended. The fruit eaters cut back on the intake of other foods to such an extent that their (self reported) calorie intake was practically identical to baseline. Over 8 weeks there was no statistically significant increase in body weight.
The nut eaters were less frugal in their eating and did show an increase in body weight of 0.67 kg which was statistically significant (p=0.049) having reported a 244 kcal/day net increase in food intake. In the interests of full disclosure I should point out that the weight increase in the fruit group was numerically very similar, but not statistically significant.
So we have the fruit eaters with the same calories per day but a lot more fructose intake (3x), and the nut eaters with an increased calorie intake but reduced fructose intake by ~50%.
What else happened ? Scanning the results for statistical significance / trends I found
- Fruit eaters tended to move around less than nut eaters according to their accelerometers.
- Nut eaters showed an increase in BMR, although most of this was down to one subject.
- Systolic blood pressure fell 5% in the fruit eaters.
- Fasting insulin rose and blood HbA1c fell in the fruit eaters.
- Triglycerides trended higher with fruit and lower with nuts (NS).
- Fructose intake in the fruit group rose from 9 to 26 grams/day.
- Fructose intake in the nut group fell from 12 to 6 grams/day.
No significant differences between groups were measured in the liver fat and various fat and muscle measurements, I suspect this is because the subjects failed to eat their free fruits and nuts as extra and self-regulated their intake - robbing the researchers of some expected fat deposition.
In the text (but not the data table) it is stated that Levels of lipoprotein(a) (Lp(a)) tended to decrease
in the nut group and to increase in the fruit group and the change in levels between groups was
statistically significant (p = 0.047) which may be an argument in favour of nuts over fruit for health benefits.
Some of the statistics are difficult for me to understand, where near identical changes in means are significant in one group but not the other (despite similar variability). Perhaps this is an artefact of the modest sample size of 15 in each group and of the small changes seen in calorie intake.
The high fruit intake only amounted to 26 grams of fructose, which is barely enough to get even Robert Lustig excited. At 100 calories or <4% of calorie intake it's below even the most hair shirt recommendation on fructose or sugars intake.
The subjects were pretty active at 12,000 steps a day so their carbohydrate oxidation rate would have easily absorbed the added sugars, although we do not know the total change in intake as ee were not given the total intake of monosaccharides, though apparently it correlated well with baseline liver fat in women but not men. The composition of the diet in terms of macronutrients is also not provided.
Another nagging doubt I have is to whether the reported fructose figure is the free fructose in the diet or includes the fructose element of sucrose in the baseline diet and the added fruit.
I think we can conclude that if there is a harmful effect of adding ~500 grams of fruit a day (25 grams/day of fructose intake in total) then it does not appear in a young healthy ideal weight active population who adapt by cutting their intake of other (unspecified) foods.