Identification of biomarkers for intake of protein from meat, dairy products and grains: a controlled dietary intervention study.
Study Goal
The researchers aimed to identify potential biomarkers for dietary protein intake from dairy products, meat, and grain to estimate their consumption in epidemiological studies.
Results Summary
The study could not identify biomarkers for dairy protein intake, unlike meat and grain proteins, for which specific amino acid combinations were predictive.
Population
Thirty men and seventeen women (average age 22 ± 4 years).
Effective Dosage
Approximately 14% of energy intake from dairy protein.
Duration
1 week per dietary intervention (total of 3 weeks, including run-in).
Interactions
None mentioned
| Intervention | Direction | Endpoint | Population | Dosage | Impact | Claim # |
|---|---|---|---|---|---|---|
high-protein diet with approximately 14 en% originating from meat | no change | intake of meat protein | thirty men and seventeen women (22 (SD 4) years) | 98% of variation in intake explained | A very good prediction could be made for the intake of meat protein | #1 |
high-protein diet with approximately 14 en% originating from grain | no change | intake of grain protein | thirty men and seventeen women (22 (SD 4) years) | 75% of variation explained | made a good prediction for dietary grain protein | #2 |
high-protein diet with approximately 14 en% originating from dairy products | no change | dairy protein intake | thirty men and seventeen women (22 (SD 4) years) | - | could not identify biomarkers | #3 |
In the present controlled, randomised, multiple cross-over dietary intervention study, we aimed to identify potential biomarkers for dietary protein from dairy products, meat and grain, which could be useful to estimate intake of these protein types in epidemiological studies. After 9 d run-in, thirty men and seventeen women (22 (SD 4) years) received three high-protein diets (aimed at approximately 18% of energy (en%)) in random order for 1 week each, with approximately 14 en% originating from either meat, dairy products or grain. We used a two-step approach to identify biomarkers in urine and plasma. With principal component discriminant analysis, we identified amino acids (AA) from the plasma or urinary AA profile that were distinctive between diets. Subsequently, after pooling total study data, we applied mixed models to estimate the predictive value of those AA for intake of protein types. A very good prediction could be made for the intake of meat protein by a regression model that included urinary carnosine, 1-methylhistidine and 3-methylhistidine (98% of variation in intake explained). Furthermore, for dietary grain protein, a model that included seven AA (plasma lysine, valine, threonine, α-aminobutyric acid, proline, ornithine and arginine) made a good prediction (75% of variation explained). We could not identify biomarkers for dairy protein intake. In conclusion, specific combinations of urinary and plasma AA may be potentially useful biomarkers for meat and grain protein intake, respectively. These findings need to be cross-validated in other dietary intervention studies.