Tuesday, May 31, 2016

Paleolithic Diet Trials: Jonsson et al (Part 2)


This paper was published recently and adds some more data to the Jonsson Paleo diet trial, which I have previously blogged about here

To summarise, the Jonsson trial was a randomised crossover trial (3 months each) with 13 patients that had type 2 diabetes, where a Paleo diet was compared conventional dietary advice for diabetics [1]


Paleolithic diet
Diabetes diet
Higher in each diet
Fruits, vegetables, meat, eggs
Potatoes, cereals without rice, milk/milk products
Energy (kcal)
1581
1878
Protein: fat: carbohydrate
24:39:32
20:34:42

The Paleo group lost more weight and some metabolic markers were significantly more improved compared to the diabetes diet (HDL-C, triglycerides, DBP, HbA1c), but not others (SBP, glucose, measures of insulin resistance) (table 3) [1]

A follow-up study looked at the satiety and subjective experiences (see link for responses) of the participants and found satiety to be equal but satiety per calorie to be higher on the Paleo diet [2]

This study looked various hormones and found that only leptin was significantly different, though glucagon was near significant (see table 2 for more).  The significant difference in leptin is not that surprising considering the Paleo diet resulted in significantly more weight loss than the diabetes diet


Palaeolithic diet
Diabetes diet
Delta diets
p
Leptin
(ng/ml)
5.1 ± 4.9
(2.1 to 8.0)
7.4 ± 8.3
(2.4 to 12)
−2.3 ± 4.6
(−5.1 to 0.4)
0.023
Insulin
(pg/ml)
248 ± 138
(165 to 332)
336 ± 327
(138 to 533)
−87 ± 240
(−232 to 58)
0.266
Glucagon
(pg/ml)
409 ± 40
(385 to 433)
431 ± 51
(400 to 463)
−22 ± 43
(−48 to 3.9)
0.089

They then looked at correlations between hormones with other hormones, metabolic markers and nutrients with the significant correlations being listed in table 3.  With this kind of data I place more value on the correlations between hormones and metabolic markers rather than nutrients, as the former can potentially inform on the biological mechanisms (for example there are significant correlations between glucagon and some of the insulin measurements (unsurprisingly), although leptin wasn’t significantly correlated with weight loss), while the latter is really just an observational study within an RCT (and you can sometimes find just about anything depending on the intervention)

No comments:

Post a Comment