UCSF

Background/objectives

Formulas developed to estimate diet-dependent net acid excretion (NAE) generally agree with measured values for typical Western diets. Whether they can also appropriately predict NAE for 'Paleolithic-type' (Paleo) diets-which contain very high amounts of fruits and vegetables (F&V) and concurrent high amounts of protein is unknown. Here, we compare measured NAEs with established NAE estimates in subjects with Type 2 diabetes (T2D).

Subjects/methods

Thirteen subjects with well-controlled T2D were randomized to either a Paleo or American Diabetes Association (ADA) diet for 14 days. Twenty-four hour urine collections were performed at baseline and end of the diet period, and analyzed for titratable acid, bicarbonate and ammonium to calculate measured NAE. Three formulas for estimating NAE from dietary intake were used; two (NAE_diet R or L) that include dietary mineral intake and sulfate- and organic acid (OA) production, and one that is empirically derived (NAE_diet F) only considering potassium and protein intake.

Results

Measured NAE on the Paleo diet was significantly lower than on the ADA-diet (+31±22 vs 112±52 mEq/day, P=0.002). Although all formula estimates showed similar and reasonable correlations (r=0.52-0.76) with measured NAE, each one underestimated measured values. The formula with the best correlation did not contain an estimate of dietary OA production.

Conclusions

Paleo-diets are lower in NAE than typical Western diets. However, commonly used formulas clearly underestimate NAE, especially for diets with very high F&V (as the Paleo diet), and in subjects with T2D. This may be due to an inappropriate estimation of proton loads stemming from OAs, underlining the necessity for improved measures of OA-related proton sources.