The Journal of Nutrition. First published ahead of print November 26, 2014 as doi: 10.3945/jn.114.195438.

The Journal of Nutrition

Methodology and Mathematical Modeling

Protein Digestibility-Corrected Amino Acid

Scores and Digestible Indispensable Amino
Acid Scores Differentially Describe Protein
Quality in Growing Male Rats1–3
Shane M Rutherfurd,4* Aaron C Fanning,5 Bruce J Miller,5 and Paul J Moughan4
4
Riddet Institute, Massey University, Palmerston North, New Zealand; and 5Fonterra Research and Development Centre, Palmerston
North, New Zealand

Abstract
Background: The FAO has recommended replacing the protein digestibility–corrected amino acid score (PDCAAS) with
the digestible indispensable amino acid score (DIAAS).

Downloaded from jn.nutrition.org by guest on October 24, 2017

Objective: The objective of this study was to compare aspects underlying the calculation of the DIAAS and PDCAAS,
including 1) fecal digestibility vs. ileal digestibility, 2) using a single nitrogen digestibility value for all amino acids, and 3) the
effect of truncation. Truncated PDCAAS and untruncated DIAAS values calculated as formally defined were also compared
and DIAAS data presented for 14 dietary protein sources.
Methods: Semisynthetic wheat starch–based diets were formulated to contain the test protein (as consumed by humans)
source (whey- and soy-protein isolates, milk-, whey-, rice- and pea- protein concentrates, cooked kidney beans, roasted
peanuts, cooked peas, corn-based breakfast cereal, cooked rice, cooked rolled oats, and wheat bran) as the sole nitrogen
source and with an indigestible marker (titanium dioxide). Growing male rats (;250 g bodyweight) were given a basal
casein-based diet from day 1 to day 7 and then allocated (n = 6) to the test diets for day 8 to day 14 before ileal digesta were
collected after the rats were killed. Total feces were collected from day 11 to day 14.
Results: True fecal nitrogen digestibility was different (P < 0.05; 10% difference on average) from true ileal nitrogen
digestibility for 11 of the 14 protein sources. True ileal nitrogen digestibility was different (P < 0.05) from true ileal amino
acid digestibility for almost half of the indispensable and conditionally indispensable amino acids (differences ranged from
0.9% to 400%). DIAAS values ranged from 0.01 for a corn-based cereal to 1.18 for milk protein concentrate.
Conclusion: Untruncated PDCAAS values were generally higher than a DIAAS values, especially for the poorer quality
proteins; therefore, the reported differences in the scores are of potential practical importance for populations in which
dietary protein intake may be marginal. J Nutr doi: 10.3945/jn.114.195438.

Keywords: PDCAAS, DIAAS, protein quality, ileal, amino acid

Introduction
An accurate assessment of the nutritional quality of dietary
proteins is of fundamental importance. Although there are many
means of determining dietary protein quality, the FAO/WHO
(1) have previously recommended the protein digestibility–
1
This study was funded by the Primary Growth Partnership, a New Zealand
Government funding initiative.
2
Author disclosures: SM Rutherfurd, BJ Miller, and PJ Moughan, no conflicts of
interest. AC Fanning is a current employee of Fonterra Co-operative Group
Limited, a New Zealand dairy company.
3
Supplemental Tables 1 and 2 are available from the ‘‘Online Supporting
Material’’ link in the online posting of the article and from the same link in the
online table of contents at http://jn.nutrition.org.
* To whom correspondence should be addressed. E-mail: s.m.rutherfurd@
massey.ac.nz.
corrected amino acid score (PDCAAS)6 as a suitable method.
The score is based on the ratio of the amount of the first-limiting
dietary indispensable amino acid in the protein source to the
amino acid requirement of the 1–2-y-old child corrected for
protein digestibility based on true fecal nitrogen digestibility and
using the growing rat as a model for the adult human. PDCAAS
values >1 for both whole foods and ingredients are rounded
(truncated) to 1 (1).
Although the PDCAAS method has proven to be useful, it has
been criticized on several counts (2–6). First, the PDCAAS uses
6
Abbreviations: DIAAS, digestible indispensable amino acid score; MPC, milk
protein concentrate; PDCAAS, protein digestibility–corrected amino acid score;
SPI, soy protein isolate; WPC, whey protein concentrate; WPI, whey protein
isolate.

ã 2015 American Society for Nutrition.

Manuscript received June 4, 2014. Initial review completed July 8, 2014. Revision accepted October 30, 2014. 1 of 8
doi: 10.3945/jn.114.195438.
Copyright (C) 2014 by the American Society for Nutrition
fecal rather than ileal estimates of protein digestibility, yet ileal
digestibility estimates are more accurate compared with fecal
digestibility estimates, as discussed by Moughan (7) and Schaafsma
(4), because fecal digestibility estimates are confounded by the
metabolic activity of the microbiota of the hindgut (8, 9).
Second, protein digestibility values (based on the digestibility
of nitrogen) are less accurate than amino acid digestibility
values (based on the digestibility of individual amino acids)
for describing protein quality because they do not reflect the
digestibility of amino acids individually (4, 5). Third, metabolic
fecal nitrogen does not accurately account for the endogenous
proteins that are lost into the gut lumen and are present in ileal
digesta and feces. Fourth, protein ingredient sources that have
PDCAAS values >1, but that have been truncated to 1, cannot
be adequately ranked as sources of the first-limiting amino acid
(3–5). Fifth, the use of the amino acid requirements of the 1–2-y-
old child to estimate PDCAAS values for all humans and the
accuracy of those requirement estimates have been questioned
(10, 11).
Taking the above criticisms into account, the FAO (12) has
proposed that the PDCAAS be replaced by the digestible in-
dispensable amino acid score (DIAAS), which is based on true
ileal amino acid digestibility determined for each amino acid
<100 g/kg protein, the vitamin and mineral mix and marker only were
added to the protein source. The ingredient composition of the basal and
test diets is presented in Supplemental Table 1. All diets met the nutrient
requirements of the growing rat, with the exception of protein (13). The
animal trial using 84 male Sprague-Dawley rats weighing ;250 g was
conducted as described by Rutherfurd and Moughan (14), except that
the rats were fed a casein-based basal diet from day 1 to day 7, and then
were fed their respective test diets from day 8 to day 14. From day 11 to
day 14, total feces were collected from each rat. The feces were freeze-
dried and the nitrogen and titanium content determined. On day 15 of
the study, the rats were killed using asphyxiation with carbon dioxide gas
followed by decapitation and terminal ileal digesta samples collected as
described by Rutherfurd and Moughan (14). The digesta were freeze-
dried in preparation for the analysis of amino acids, nitrogen, and
titanium dioxide. Inspection of digesta at collection indicated that
coprophagy had not occurred.
Chemical analyses. Amino acid contents of the protein sources, diets,
and digesta samples were determined as described by Rutherfurd et al.
(15). Reactive Lys was determined by using the method of Rutherfurd
and Moughan (16). The weight of each amino acid was calculated by
using free amino acid molecular weights.
Crude protein was determined as nitrogen multiplied by 6.25 (12, 17).
Nitrogen was determined in duplicate using the LECO total combustion
method (18) on a LECO TruSpec CN (Carbon/Nitrogen) Determinator

Downloaded from jn.nutrition.org by guest on October 24, 2017

individually, and Lys availability (12) estimates, using non-
truncated scores for food ingredients.
The aim of this study was to compare aspects underlying the
calculation of the DIAAS and PDCAAS, including the use of
digestibility values determined at the fecal vs. ileal level (true
fecal nitrogen digestibility compared with true ileal nitrogen
digestibility) and the suitability of using a single digestibility
value for all amino acids (true ileal nitrogen digestibility
compared with true ileal amino acid digestibility). Truncated
PDCAAS and untruncated DIAAS values calculated as formally
defined (i.e., each based on the prescribed amino acid reference
pattern) were compared.
(LECO Corporation). The titanium dioxide content of the diets and
digesta samples was determined using the method of Short et al. (19).
Data Analysis. Amino acids and nitrogen (micrograms per gram dry
matter intake) in the terminal ileal digesta and true ileal nitrogen and
amino acid digestibility were calculated by using equations that are fully
documented elsewhere (15). Endogenous ileal amino acid flows were
based on those determined in the growing rat by using the enzyme
hydrolyzed protein/ultrafiltration method and reported by Rutherfurd
and Moughan (20). The estimates of endogenous ileal amino acid flows
were the means of flows determined by using different protein hydrol-
TABLE 1 True ileal nitrogen digestibility and true fecal nitrogen
digestibility determined in growing male rats for the 14 protein
sources ranked for ileal nitrogen digestibility1

Methods True nitrogen

digestibility, % Overall
Fourteen food protein sources were chosen to represent a cross-section of
foods and food ingredients that likely varied in their digestibility. The Protein source Ileal Fecal SEM2 P Overestimation,2 %
dairy proteins whey protein isolate (WPI) 8855, milk protein concentrate
(MPC) 4850, whey protein concentrate (WPC) 392, and acid casein were WPI 99.0 102 0.54 ,0.001 2.7
obtained from the Fonterra Co-operative Group. Rice protein concen- PPC 97.2 99.0 1.45 0.18 —
trate (Oryzatein 90) was obtained from Axiom Foods, soy protein isolate WPC 95.3 99.8 0.90 ,0.001 4.8
(SPI) A (Supro XF) and SPI B (Supro 670) were obtained from Solae, and SPI A 95.0 98.2 1.00 0.004 3.3
pea protein concentrate (Nutralys S85) was obtained from Roquette. SPI B 94.4 98.5 0.64 ,0.001 4.4
Kidney beans, roasted peanuts, whole peas, a corn-based breakfast MPC 92.1 98.1 1.10 ,0.001 6.6
cereal, rice, rolled oats, and wheat bran were obtained from local Roasted peanuts 90.9 98.4 1.71 ,0.001 8.2
supermarkets. The raw kidney beans, fresh garden peas, rice, and rolled
Cooked rolled oats 88.5 95.2 1.23 0.003 7.6
oats were all cooked as described below. The kidney beans were soaked
Cooked peas 88.4 89.0 1.51 0.66 —
overnight, then heated in water to 100°C and cooked for 45 min at that
temperature. The peas were heated in water to 100°C and then cooked Cooked kidney beans 79.6 80.4 4.47 0.85 —
for 4 min at that temperature. The rice was cooked using a commercially RPC 79.6 88.1 1.32 ,0.001 11
available rice cooker as described by the manufacturer (Kambrook). The Cooked rice 72.8 86.6 3.60 0.003 19
rolled oats were mixed with an equal volume of water, heated to ;100° Wheat bran 72.7 85.3 3.28 0.001 17
C, and cooked at that temperature for 15 min. All cooked materials were Corn-based breakfast 66.7 81.8 1.88 ,0.001 23
freeze-dried after cooking and all materials were ground through a 1mm cereal
mesh just prior to inclusion into the diets.
1
Values are means, n = 6, except for true ileal nitrogen digestibility, n = 3 for the
Rat study. National guidelines for the care and use of animals were cooked rice and n = 2 for corn-based breakfast cereal and cooked rolled oats, because
there was insufficient material to conduct the nitrogen analysis on all digesta samples.
followed and all experimental procedures involving rats were approved
MPC, milk protein concentrate; PPC, pea protein concentrate; RPC, rice protein
by the Massey University Animal Ethics Committee. concentrate; SPI, soy protein isolate; WPC, whey protein concentrate; WPI, whey
Fourteen semisynthetic wheat starch–based diets were formulated to protein isolate.
contain 100 g/kg protein, with each diet containing 1 protein source as 2
Overestimation (where differences were statistically significant) was calculated as
the sole source of protein. Titanium dioxide (3 g/kg) was included in the absolute difference between true fecal nitrogen digestibility and true ileal nitrogen
each diet as an indigestible marker. For protein sources that contained digestibility expressed as a percentage of true ileal nitrogen digestibility.

2 of 8 Rutherfurd et al.
ysates (casein, beef muscle protein, lactalbumin, zein, and SPI) in rats of
a body weight that was similar to those used in the present study.
Endogenous ileal nitrogen flow was that determined in rats fed a protein-
free diet as reported by Butts et al. (21). True fecal nitrogen digestibility
was calculated by following WHO guidelines (6), in which the metabolic
fecal nitrogen flow was that determined in rats fed a protein-free diet and
as reported by Hendriks et al. (22).
The PDCAAS was calculated as prescribed by the FAO/WHO/UNU
Expert Consultation (1) as follows:
PDCAAS ¼ Lowest uncorrected amino acid ratio
3 fecal crude protein digestibilityð%Þ
where the amount of test protein was based on crude protein (total N
content 3 6.25) content. The reference protein indispensable amino acid
profile was either the amino acid requirement pattern for the 1–2-y-old
child (6) or the amino acid requirement pattern for the 0.5–3-y-old child
(12). For comparison of the PDCAAS with the DIAAS, PDCAAS values
that were >1 were truncated to 1.
The DIAAS was calculated as prescribed by the FAO (12) as follows:
DIAAS ¼ lowest digestible indispensable amino acid ðithÞ reference ratio
0.5–3-y-old child (12), and the digestible amino acid content was
calculated as follows:
Digestible amino acid ðithÞ content ðg=kg proteinÞ
¼ Amino acid ðithÞ content ðg=kg proteinÞ
3True ileal amino acid ðithÞ digestibilityð%Þ
where i (1 2 n) refers to the dietary indispensable or conditionally
indispensable amino acids and where for Lys, estimates of availability
(digestibility of reactive Lys) rather than digestibility were used.
Statistical Analysis. True ileal nitrogen digestibility was compared
with true fecal nitrogen digestibility and true ileal nitrogen digestibility
was compared with the true ileal amino acid digestibility using 1-factor
ANOVA for each protein source singly (general linear model procedure)
by using SAS version 9.3 (SAS Institute) (23). True ileal nitrogen digest-
ibility was compared with true ileal amino acid digestibility using
orthogonal contrasts when overall significant differences were observed
(P < 0.05). In addition, a correlation analysis was conducted between
true ileal nitrogen digestibility and true fecal nitrogen digestibility by
using the CORR procedure in SAS version 9.3 (23).

Results
where the digestible indispensable amino acid reference ratio is calcu-
lated as follows: Gross protein and amino acid composition. The gross

Downloaded from jn.nutrition.org by guest on October 24, 2017

where the amount of test protein was based on crude protein (total N amino acid and protein contents of the 14 protein sources are
content 3 6.25) content, i (1 2 n) refers to the dietary indispensable presented in Supplemental Table 2.

mg of the ith digestible dietary indispensable amino acid in 1 g of the test protein
Digestible indispensable amino acid reference ratio ¼
mg of the ith dietary indispensable amino acid in 1 g of the reference protein

or conditionally indispensable amino acid, the reference protein Comparison between true ileal nitrogen digestibility and
amino acid profile was the amino acid requirement pattern for the true fecal nitrogen digestibility. True ileal nitrogen digestibility

TABLE 2 True ileal amino acid digestibility and Lys availability determined in growing male rats for the 14 protein sources1

Protein source, %
Cooked Cooked Corn-based
Cooked kidney Cooked rolled Wheat Roasted breakfast
MPC WPI WPC SPI A SPI B PPC peas beans rice oats bran peanuts RPC cereal

Asp2 96 99 97 96 95 97 91 76 77 84 68 92 79 47
Thr 94 99 94 95 92 96 88 70 70 82 65 87 80 62
Ser 83 98 93 97 95 97 89 77 73 84 70 91 81 72
Glu3 93 99 97 98 98 98 94 80 71 92 85 94 80 83
Ala 96 100 98 96 95 98 92 74 75 82 67 94 85 82
Cys 96 100 100 97 94 98 87 69 66 93 78 93 71 66
Val 94 99 97 95 94 96 90 73 80 85 68 91 83 69
Met 92 97 94 91 89 92 91 81 55 90 76 94 63 79
Ile 92 100 98 97 96 98 92 80 79 88 73 93 83 71
Leu 98 100 99 96 95 98 92 81 75 89 75 94 83 87
Tyr 99 100 99 98 97 98 92 78 74 89 73 95 83 84
Phe 99 100 99 97 96 98 93 81 80 90 76 96 84 83
His 98 99 98 98 97 96 92 79 82 88 80 95 83 72
Trp 92 98 97 90 90 91 83 72 80 79 71 79 80 25
Lys4 99 100 99 98 98 98 96 94 92 84 73 92 86 13
Arg 96 100 96 99 98 98 94 78 87 87 72 95 91 74
Mean amino acid digestibility 94 100 98 97 96 98 92 77 77 87 74 93 82 67
Overall SEM5 1.0 0.5 0.7 0.8 0.5 1.1 0.4 0.9 0.3 0.3 0.6 0.5 1.1 0.2
1
Values are means, n = 6. MPC, milk protein concentrate; PPC, pea protein concentrate; RPC, rice protein concentrate; SPI, soy protein isolate; WPC, whey protein concentrate;
WPI, whey protein isolate.
2
Asp consists of Asp + Asn.
3
Glu consists of Glu + Gln.
4
Lys availability based on reactive Lys determined by using the guanidination method (16).
5
Overall SEM for true ileal amino acid digestibility across amino acids.

Determining dietary protein quality for humans 3 of 8

(ranked from highest to lowest) and true fecal nitrogen digest-
ibility values for the 14 protein sources are given in Table 1.
Overall, there was a strong correlation between true ileal nitrogen
digestibility and true fecal nitrogen digestibility (r = 0.91, P <
0.001). When each protein source was examined individually,
there was no difference (P $ 0.05) between true ileal nitrogen
digestibility and true fecal nitrogen digestibility for pea protein
concentrate, cooked kidney beans, and cooked peas, but true fecal
nitrogen digestibility overestimated (P < 0.01) true ileal nitrogen
digestibility from 2.7% to 8.2% for some of the highly digestible
protein sources (i.e. those having a true ileal nitrogen digestibility
value >90%), including WPI, WPC, SPI A, SPI B, MPC, and
roasted peanuts, and from 7.6% to 22.7% for some of the more
poorly digested protein sources (i.e. those having a true ileal
nitrogen digestibility <90%), including rice protein concentrate,
cooked rice, corn-based breakfast cereal, cooked rolled oats, and
wheat bran.
Comparison between true ileal amino acid digestibility
and true ileal nitrogen digestibility. True ileal amino acid

TABLE 3 Statistical comparison between true ileal nitrogen digestibility and true ileal amino acid digestibility for the dietary
indispensable and dietary conditionally indispensable amino acids determined in growing male rats for the 14 protein sources1

Amino acid
Thr Cys Met Val Ile Leu Tyr Phe His Trp Lys2 Overall significance3

MPC ,0.001
P values 0.014 0.14 0.13 0.005 0.41 ,0.001 ,0.001 ,0.001 ,0.001 0.91 ,0.001
Underestimation 2.4 — — 2.0 — 5.8 6.7 7.3 5.7 — 6.7
WPI ,0.001
P values 0.5 0.015 0.006 0.73 0.27 0.16 0.26 0.08 0.49 0.11 0.07
Underestimation — 1.9 21.6 — — — — — — — —

Downloaded from jn.nutrition.org by guest on October 24, 2017

WPC ,0.001
P values 0.16 ,0.001 0.05 0.003 0.003 ,0.001 ,0.001 ,0.001 0.031 0.027 0.016
Underestimation — 4.7 — 1.4 2.9 3.8 3.9 4.1 2.9 1.3 3.6
SPI B ,0.001
P values 0.004 0.82 ,0.001 0.89 0.011 0.07 ,0.001 0.002 0.001 0.008 0.006
Underestimation 22.2 — 25.5 — 1.3 — 2.9 1.9 2.4 25.0 4.0
SPI A ,0.001
P values 0.76 0.17 0.001 0.35 0.04 0.15 0.009 0.19 0.005 0.017 0.004
Underestimation — — 24.4 — 1.7 — 3.0 2.1 3.2 25.1 3.3
PPC ,0.001
P values 0.24 0.08 ,0.001 0.04 0.3 0.06 0.2 0.049 0.13 0.006 0.06
Underestimation — — 26.1 20.87 — — — 0.99 — 26.3 —
Cooked peas ,0.001
P values 0.76 0.62 0.06 0.21 ,0.001 ,0.001 0.003 ,0.001 0.001 0.12 0.001
Underestimation — — — — 3.9 4.4 4.0 4.6 4.2 — 7.8
Cooked kidney beans 0.027
P values 0.17 0.14 0.07 0.93 0.09 0.06 0.20 0.09 0.17 0.021 0.043
Underestimation — — — — — — — — — 211 15
Cooked rice ,0.001
P values 0.08 0.023 0.012 0.43 0.029 0.29 0.65 0.047 0.018 0.19 0.009
Underestimation — 210 231 — 7.6 — — 9.5 11 — 21
Cooked rolled oats4 0.05
P values — — — — — — — — — — —
Underestimation — — — — — — — — — — —
Wheat bran ,0.001
P values 0.002 0.040 0.08 0.08 0.26 0.041 0.85 0.005 0.11 0.52 0.89
Underestimation 212 6.2 — — — 2.5 — 4.8 — — —
Roasted peanuts ,0.001
P values 0.002 0.005 0.002 0.26 0.002 0.009 0.011 0.012 0.05 0.10 0.44
Underestimation 24.1 1.9 3.7 — 2.5 3.4 4.0 5.1 — — —
RPC ,0.001
P values 0.62 ,0.001 ,0.001 0.002 0.002 0.002 0.009 ,0.001 0.015 0.80 0.027
Underestimation — 212 227 4.2 4.2 3.7 4.3 4.9 4.4 — 6.9
Corn-based breakfast cereal ,0.001
P values 0.14 0.17 0.18 0.29 0.40 0.06 0.10 0.09 0.17 0.036 0.017
Underestimation — — — — — — — — — 2160 2400
1
P values for the 14 protein sources are comparisons of the true ileal digestibility of nitrogen and the individual amino acid. Underestimation is the percentage underestimation of
true ileal amino acid digestibility by true ileal nitrogen digestibility. Negative values represent an overestimation. MPC, milk protein concentrate; PPC, pea protein concentrate;
RPC, rice protein concentrate; SPI, soy protein isolate; WPC, whey protein concentrate; WPI, whey protein isolate.
2
Lys availability based on reactive Lys determined by using the guanidination method (16)
3
Compared across true ileal amino acid digestibility and true ileal nitrogen digestibility for each protein source separately.
4
No P values are presented for the individual amino acids since the overall P value for cooked rolled oats was $0.05.

4 of 8 Rutherfurd et al.
digestibility and Lys availability values for the 14 protein sources
were determined (Table 2). True ileal digestibility values for
glycine were not reported, because endogenous ileal glycine
flows (used to correct apparent glycine digestibility values to
true values) may be overestimated when determined by using the
enzyme hydrolyzed protein/ultrafiltration method (14). Mean
true ileal digestibility across all of the determined amino acids
(excluding glycine) ranged from 67% for the corn-based break-
fast cereal to 100% for WPI. True ileal amino acid digestibility
ranged across the dietary indispensable and conditionally
indispensable amino acids within each protein source with the
mean difference across amino acids being 18% units across the
14 protein sources.
True ileal amino acid digestibility was compared for each
amino acid individually with true ileal nitrogen digestibility for
each protein source and the outcome of the statistical analysis is
shown in Table 3. There was no overall difference (P $ 0.05) be-
ween true ileal digestibility of dietary indispensable/conditionally
indispensable amino acids and true ileal digestibility of nitrogen
for cooked rolled oats, but for the other protein sources, true
ileal nitrogen digestibility and true ileal amino acid digestibility
were different (P < 0.05) for between 2 and 9 amino acids
(mean = 6 amino acids across protein sources) out of the 11
were calculated based on the amino acid requirement pattern for
the 0.5–3-y-old child (12). The reference ratios were $1 for all
of the dietary indispensable amino acids within each protein
source for only MPC and WPI. When the lowest digestible
indispensable amino acid reference ratio (DIAAS) was compared
with the lowest non-truncated protein digestibility–corrected
amino acid reference ratio (PDCAAS), the PDCAAS over-
estimated the DIAAS by between 2.0% for WPI and WPC and
515% for the corn-based breakfast cereal (mean overestimation,
7% across protein sources, excluding the corn-based breakfast
cereal). The PDCAAS underestimated the DIAAS by 6% for
cooked rice, whereas for cooked peas the latter 2 ratios were
similar.
Comparison between the DIAAS and PDCAAS as formally
defined. The PDCAAS determined as described by the FAO (1)
and corrected for fecal crude protein digestibility, truncated, and
based on the amino acid requirement pattern for the 1–2-y-old
child (6), and the DIAAS determined as prescribed by the
FAO (12) and corrected for true ileal amino acid digestibility,
untruncated, and based on the amino acid requirement pattern
for the 0.5–3-y-old child (12) are presented in Table 5. The
PDCAAS underestimated the DIAAS by 15% and 8% for MPC

Downloaded from jn.nutrition.org by guest on October 24, 2017

dietary indispensable and conditionally indispensable amino
acids examined for each protein source. For the latter amino
acids, true ileal nitrogen digestibility generally underestimated
true ileal amino acid digestibility (mean absolute differences
across amino acids ranged from 1.8% for WPI to 280% for the
corn-based breakfast cereal, with a mean difference across all
protein sources of 27%).
Amino acid reference ratios. Untruncated amino acid refer-
ence ratios determined based on true ileal amino acid digest-
ibility and Lys availability or true fecal nitrogen digestibility
alone are presented in Table 4. For Table 4, all reference ratios
and WPI, respectively. In contrast, the PDCAAS overestimated
the DIAAS by between 3% and 574% (mean overestimation,
59%) for the remaining 12 protein sources.
Discussion
A fundamental criticism of the PDCAAS is that it uses fecal
rather than ileal digestibility. A strong correlation was found
between true fecal nitrogen digestibility and true ileal nitrogen
digestibility, but true fecal nitrogen digestibility generally over-
estimated true ileal nitrogen digestibility. Another criticism of

TABLE 4 Untruncated digestible indispensable amino acid reference ratios calculated using true ileal amino acid digestibility values
and lowest untruncated PDCAAS calculated using true fecal nitrogen digestibility values determined in growing male rats for the 14
protein sources1

Amino acid reference ratio

Cooked Cooked Corn-based
Cooked kidney Cooked rolled Wheat Roasted breakfast
MPC WPI WPC SPI B SPI A PPC peas beans rice oats bran peanuts RPC cereal

Thr 1.56 1.80 2.53 1.13 1.30 1.28 1.12 0.936 0.757 0.884 0.595 0.574 1.01 0.652
Met + Cys 1.18 2.29 1.71 0.906 0.898 0.822 0.579 0.588 1.04 1.95 0.888 0.833 1.22 0.975
Val 1.55 1.21 1.29 1.02 1.11 1.24 0.870 0.791 0.927 0.872 0.542 0.560 1.12 0.707
Ile 1.81 2.22 2.35 1.38 1.59 1.63 1.25 1.20 1.10 1.13 0.689 0.788 1.16 0.881
Leu 1.77 2.57 1.93 1.13 1.29 1.37 1.04 1.01 0.989 1.10 0.664 0.766 1.11 2.05
Tyr + Phe 2.39 1.71 1.43 1.65 1.85 1.99 1.36 1.43 1.63 1.66 0.927 1.31 1.83 1.75
His 1.60 1.09 0.973 1.18 1.37 1.34 1.01 1.25 1.09 1.11 1.04 1.07 1.03 1.07
Trp 1.94 3.35 2.74 1.69 1.67 1.12 1.47 1.50 1.85 1.75 1.74 1.28 1.38 0.228
Lys2 1.77 2.51 2.03 0.987 1.16 1.50 0.992 1.07 0.595 0.542 0.411 0.434 0.371 0.012
Lowest digestible indispensable 1.18 1.09 0.973 0.906 0.898 0.822 0.579 0.588 0.595 0.542 0.411 0.434 0.371 0.012
amino acid reference ratio, DIAAS
Lowest protein digestibility–corrected 1.25 1.12 0.990 0.974 0.943 0.860 0.575 0.624 0.562 0.611 0.479 0.464 0.382 0.071
amino acid reference ratio,3 PDCAAS
1
The digestible indispensable amino acid reference ratio was calculated using the amino acid requirement pattern for the 0.5–3-y-old child (grams per kilogram protein) (12).
DIAAS, digestible indispensable amino acid score; MPC, milk protein concentrate; PDCAAS, protein digestibility–corrected amino acid score; PPC, pea protein concentrate; RPC,
rice protein concentrate; SPI, soy protein isolate; WPC, whey protein concentrate; WPI, whey protein isolate.
2
Based on reactive Lys content and Lys availability determined by using the guanidination method (16).
3
The lowest PDCAA ratio was calculated using the amino acid requirement pattern for the 0.5–3-y-old child (grams per kilogram protein) (12). For protein sources for which Lys
was the limiting amino acid, the ratio was based on total Lys content determined by using conventional amino acid analysis.

Determining dietary protein quality for humans 5 of 8

TABLE 5 The DIAAS and PDCAAS determined in growing male sources when compared with using food specific factors.
rats for the 14 protein sources1 Moreover, PDCAAS and DIAAS values will tend to be over-
estimated when using nitrogen-to-protein conversion factors if
DIAAS PDCAAS there is substantial nonprotein nitrogen present in the protein
MPC 1.18 1.00 source. Overall, the validity of using the generalized factor of
WPI 1.09 1.00 6.25, food specific factors, or other methods, such as the
WPC 0.973 1.00 summation of amino acids, to estimate the protein content for
SPI B 0.906 1.00 determining the DIAAS warrants further discussion. It is of note
SPI A 0.898 0.979 that although the absolute PDCAAS and DIAAS values can
PPC 0.822 0.893 change depending on whether a generalized factor or food-
Cooked peas 0.579 0.597 specific factors are used, the relative differences between the
Cooked kidney beans 0.588 0.648 PDCAAS and DIAAS for any given food are not greatly affected
Cooked rice 0.595 0.616 (data not shown).
Cooked rolled oats 0.542 0.670 The present study used only a relatively small sample set of
Wheat bran 0.411 0.525 foods (14 protein sources), but even within that sample set,
Roasted peanuts 0.434 0.509 meaningful differences between the PDCAAS and DIAAS (Table
RPC 0.371 0.419 4) were observed when calculated by using the same amino acid
Corn-based breakfast cereal 0.012 0.078 requirement pattern. For example, the PDCAAS differed from
the DIAAS by >6% for half of the protein sources tested and by
1
The DIAAS was the lowest digestible indispensable amino acid reference ratio (12), >12% for 3 protein sources. The observed differences for some
where the digestible indispensable amino acid reference ratio was calculated as
of the protein sources are particularly relevant for populations
shown for Table 4. The lowest protein digestibility–corrected amino acid ratio was
calculated using the amino acid requirement pattern for the 1–2-y-old child (grams per with a marginal protein intake, in whom the consequences of
overestimating protein quality may be severe. These results

Downloaded from jn.nutrition.org by guest on October 24, 2017

kilogram protein) (6). This is a different ratio than that used for calculating the PDCAAS
in Table 4 and reflects the recommended scoring profile for the PDCAAS as formally support the implementation of the DIAAS. When the PDCAAS
defined by the FAO (1). For protein sources in which Lys was the limiting amino acid, and DIAAS were determined based on their respective pre-
the ratio was based on total Lys content determined using conventional amino acid
analysis. Protein digestibility–corrected amino acid ratios .1 were truncated to 1.
scribed amino acid requirement pattern [the WHO (6) for
DIAAS, digestible indispensable amino acid score; MPC, milk protein concentrate; the PDCAAS and the FAO (12) for the DIAAS] and where
PDCAAS, protein digestibility–corrected amino acid score; PPC, pea protein concen- a PDCAAS >1 was truncated to 1, the difference between
trate; RPC, rice protein concentrate; SPI, soy protein isolate; WPC, whey protein the PDCAAS and DIAAS across all protein sources excluding the
concentrate; WPI, whey protein isolate.
corn-based breakfast cereal was 13% (Table 5). The latter
difference is a function of 1) differences between true ileal amino
the PDCAAS method centers on the use of nitrogen digestibility, acid digestibility and true fecal nitrogen digestibility, 2) differ-
rather than the digestibility of individual amino acids, to correct ences in the amino acid requirement pattern prescribed for the
amino acid scores, which can lead to either an underestimate DIAAS and PDCAAS and 3) the fact that the PDCAAS uses
or overestimate of protein quality depending on the relative truncation for scores >1, whereas the DIAAS does not. The
difference between the digestibility of nitrogen and the digest- fact that the differences between the PDCAAS and DIAAS
ibility of the first-limiting amino acid (4, 5, 24). Statistically (Table 5) were often much greater than the differences shown in
significant and practically meaningful differences were found Table 4 highlights the importance of knowing the amino acid
between true ileal amino acid digestibility and true ileal nitrogen requirement values (reference pattern) with accuracy. The larger
digestibility within protein sources.
The true ileal amino acid digestibility values determined in
the rat were generally similar to comparable published values TABLE 6 Dietary digestible indispensable amino acid reference
determined in both the growing rat and adult human where such ratios and DIAAS for 2 individual protein sources (a milk protein
data are available (14, 15, 20, 25, 26). The crude protein content concentrate and a corn-based breakfast cereal) and a combination
was estimated here based on the nitrogen content. However, of the 2 protein sources1
protein content can also be estimated based on the sum of the
amino acids. The protein content estimated based on the sum Corn-based breakfast MPC:cereal
of the amino acids (bound molecular weights) was close to the MPC cereal 60:402
crude protein content (estimated using nitrogen-to-protein Thr 1.56 0.652 1.20
conversion factors specific for each protein source) for most of Met + Cys 1.18 0.975 1.10
the protein sources (results not shown). For cooked peas, cooked Val 1.55 0.707 1.22
kidney beans, roasted peanuts, the corn-based breakfast cereal, Ile 1.81 0.881 1.44
and wheat bran, however, the sum of the determined amino Leu 1.77 2.05 1.88
acids accounted for only ;80% of the estimated crude protein Tyr + Phe 2.39 1.75 2.14
content. The lower recovery of amino acids in relation to the His 1.60 1.07 1.38
crude protein content is likely mainly due to the presence of Trp 1.94 0.228 1.26
nonprotein nitrogen, particularly for the less refined protein Lys 1.77 0.012 1.07
sources. For determining both the PDCAAS (1) and DIAAS (12), Overall range3 1.21 2.04 1.07
the generalized nitrogen-to-protein conversion factor (6.25), DIAAS 1.18 0.012 1.07
rather than the use of food specific factors, is recommended.
1
However, both PDCAAS and DIAAS values will be affected by DIAAS, digestible indispensable amino acid score; MPC, milk protein concentrate.
2
Protein combination based on 60% protein from the MPC and 40% protein from the
the accuracy of the protein content estimate. For example, using corn-based breakfast cereal (scores calculated based on values for the milk protein
the nitrogen-to-protein conversion factor of 6.25 will tend to concentrate and the corn-based breakfast cereal reported herein).
overvalue dairy protein sources and undervalue plant protein 3
Calculated as follows: Overall range = maximum ratio 2 minimum ratio.

6 of 8 Rutherfurd et al.
differences between the PDCAAS and DIAAS in Table 5 compared
with Table 4 demonstrate the outcome of the collective recom-
mendations of the FAO (12) in arriving at the new protein quality
score, the DIAAS.
The use of truncation may undervalue high-quality protein
ingredients and does not demonstrate the usefulness of a protein
for the combining of protein sources to alter the amino acid
balance of the diet (i.e., the complementary effect). In this study,
when MPC and WPI were compared with the other protein
sources, the DIAAS (nontruncated scores) for the other protein
sources were on average 46% and 50% lower than for MPC or
WPI, respectively, but the PDCAAS (truncated scores) was on
average only 37% lower than for MPC and WPI. Clearly, the use
of truncation significantly undervalued the higher-quality pro-
tein ingredients compared with lower quality protein ingredi-
ents. In addition, when the PDCAAS (truncated scores) is used,
the nutritional value of MPC and WPI was deemed to be
identical, yet the DIAAS (nontruncated scores) for MPC and
WPI were different (1.18 and 1.09, respectively). Clearly,
truncated scores are not a suitable measure for assessing the
nutritive value of high-quality protein ingredients. The latter
conclusion is supported by Schaafsma (4), who commented that
using truncated PDCAAS values based only on the limiting
amino acid underestimates the power of a high-quality protein
to balance the dietary indispensable amino acid content of
poorer-quality proteins.
The large difference observed for the corn-based breakfast
cereal (515%) reflects in part the observed very low true ileal
availability of Lys (13%) in comparison with true ileal nitrogen
digestibility (67%). In the latter case, true ileal nitrogen diges-
tibility clearly markedly overestimated true ileal Lys availability
and therefore would not be suitable for determining the protein
digestibility–corrected amino acid reference ratio for Lys (the
dietary first-limiting amino acid) for the corn-based breakfast
cereal. The very low ileal availability for Lys was most likely due
to Maillard reactions that can occur during processing and
highlights the severe impact processing can have on protein
quality in some foods. For amino acids like Lys that can be
chemically modified during processing, traditional amino acid
analysis methods can considerably overestimate the content of
the unmodified amino acids in processed foods and lead to
inaccuracy in their digestibility estimates (16, 27, 28). Conse-
quently, specific analytic strategies that determine amino acid
content and availability may be required for some amino acids.
Such methods exist for Lys, and the PDCAAS has been criticized
because digestible total Lys content rather than digestible
reactive Lys (available Lys) content is used (3). The DIAAS, on
the other hand, recommends the use of digestible reactive Lys
rather than digestible total Lys. In this study, the difference
between digestible total Lys and available Lys content was
practically meaningful (>5%) for more than half of the 14
protein sources examined (SM Rutherfurd, unpublished results).
The latter may be particularly important for processed protein
sources when Lys is the first-limiting amino acid.
Information about the capability of a protein source to
improve the amino acid balance of other poorer-quality protein
sources can be lost when using the scores (ratios for the limiting
amino acid only). This point is highlighted in Table 6, in which
the DIAAS based on the limiting amino acids in MPC and the
corn-based breakfast cereal were 1.18 and 0.01, respectively.
However, when combined in a ratio of 60:40 MPC protein to
cereal protein (the ratio is based on the relative contribution of
protein and is similar to that in an average Western breakfast
consisting of a bowl of the cereal with milk), the Lys (which is
not the limiting amino acid in MPC) in the MPC complements
that of the corn-based breakfast cereal and the DIAAS of the
resulting protein mixture increases to 1.07. The power of MPC
to supplement the corn-based breakfast cereal is reflected in the
Lys ratio of 1.77 for the MPC, not its DIAAS of 1.22.
In conclusion, true fecal nitrogen digestibility differed from
true ileal nitrogen digestibility for most of the protein sources
and in many cases this difference was large. Moreover, true ileal
amino acid digestibility varied considerably across the dietary
indispensable amino acids within protein sources and, as a
result, a single nitrogen digestibility value should not be used as
a proxy for true ileal amino acid digestibility. There were some
large differences between the PDCAAS and DIAAS when the
scores were calculated as prescribed by the FAO, and these
differences were due to the way in which the digestibility
correction was made, the amino acid reference pattern used, and
truncation. When the scores were not truncated and a common
reference pattern was used, so that the difference between scores
was due to differences in the digestibility correction alone,
statistically significant differences between the PDCAAS and
DIAAS persisted. In this case, the PDCAAS was generally higher
than the DIAAS, especially for the poorer-quality proteins, and
therefore the reported differences in the scores are of potential
Downloaded from jn.nutrition.org by guest on October 24, 2017
practical importance for populations in which dietary protein
intake may be marginal.
Acknowledgments
SMR, ACF, BJM, and PJM designed the research; SMR
conducted the research and analyzed the data; and SMR and
PJM wrote the paper. All authors read and approved the final
manuscript.
References
1. FAO/WHO. Protein quality evaluation: Report of the Joint FAO/WHO
Expert Consultation, FAO food and nutrition paper no. 51. Rome
(Italy): FAO; 1991.
2. Fenwick RM, Knighton DR, Moughan PJ. Protein quality measurement
by the PDCAAS technique. Int Dairy Fed Nutr Newslett. 1995;4:40–3.
3. Gilani GS, Xiao C, Lee N. Need for accurate and standardized
determination of amino acids and bioactive peptides for evaluating
protein quality and potential health effects of foods and dietary
supplements. J AOAC Int 2008;91:894–900.
4. Schaafsma G. The protein digestibility-corrected amino acid score
(PDCAAS)-A concept for describing protein quality in foods and food
ingredients: A critical review. J AOAC Int 2005;88:988–94.
5. Schaafsma G. The protein digestibility–corrected amino acid score.
J Nutr 2000;130:1865S–7S.
6. WHO. Protein and amino acid requirements in human nutrition. Report
of a joint FAO/WHO/UNU Expert Consultation. WHO technical
reports series no. 935. Geneva (Switzerland): WHO; 2007.
7. Moughan PJ. Amino acid availability: aspects of chemical analysis and
bioassay methodology. Nutr Res Rev 2003;16:127–41.
8. Mason VC, Just A, Bech-Anderson S. Bacterial activity in the hindgut of
piglets. 2. Its influence on apparent digestibility of nitrogen and amino
acids. Z Tierphysiol Tierernahr Futtermittelkd 1976;36:310–24.
9. Rowan AM, Moughan PJ, Wilson PJ, Maher K, Tasman-Jones C.
Comparison of ileal and faecal digestibilities of dietary amino acids in
adult humans and evaluation of the pig as a model for animal digestion
studies in man. Br J Nutr 1994;71:29–42.
10. Millward DJ, Jackson AA, Price G, Rivers JPW. Human amino acid and
protein requirements: Current dilemmas and uncertainties. Nutr Res
Rev 1989;2:109–32.
11. Young VR, Borgonha S. Nitrogen and amino acid requirements: the
Massachusetts Institute of Technology amino acid requirement pattern.
J Nutr 2000;130:1841S–9S.
Determining dietary protein quality for humans 7 of 8
12. FAO Dietary protein quality evaluation in human nutrition. Report of
an FAO Expert Consultation. FAO food and nutrition paper no. 92.
Rome (Italy): FAO; 2013.
13. NRC Nutrient requirement of the laboratory rat. In: Nutrient require-
ments of laboratory animals. 4th ed. Washington, DC: National
Academies Press; 1995.
14. Rutherfurd SM, Moughan PJ. The digestible amino acid composition of
several milk proteins: Application of a new bioassay. J Dairy Sci
1998;81:909–17.
15. Rutherfurd SM, Bains K, Moughan PJ. Available lysine and digestible
amino acid contents of proteinaceous foods of India. Br J Nutr
2012;108:S59–68.
16. Moughan PJ, Rutherfurd SM. A new method for determining digestible
reactive lysine in foods. J Agric Food Chem 1996;44:2202–9.
17. Gilani GS, Sepehr E. Protein digestibility and quality in products
containing anti-nutritional factors are adversely affected by old age in
rats. J Nutr 2003;133:220–5.
18. AOAC Official methods of analysis. 18th ed. Washington, DC: AOAC; 2007.
19. Short FJ, Gorton P, Wiseman J, Boorman KN. Determination of
titanium oxide added as an inert marker in chicken digestibility studies.
Anim Feed Sci Technol 1996;59:215–21.
20. Rutherfurd SM, Cui J, Goroncy AK, Moughan PJ. Dietary protein
structure affects endogenous ileal amino acids but not true ileal amino
acid digestibility in growing male rats. J Nutr 2014. In press.
8 of 8 Rutherfurd et al.
21. Butts CA, Moughan PJ, Smith WC. Protein nitrogen, peptide
nitrogen and free amino acid nitrogen in endogenous digesta nitro-
gen at the terminal ileum of the rat. J Sci Food Agric 1992;59:
291–8.
22. Hendriks WH, Sritharan K, Hodgkinson SM. Comparison of the
endogenous ileal and faecal amino acid excretion in the dog (Canis
familiaris) and the rat (Rattus rattus) determined under protein-free
feeding and peptide alimentation. J Anim Physiol Anim Nutr (Berl)
2002;86:333–41.
23. SAS SAS user’s guide: Statistics, version 9.1. Cary (NC): SAS Institute; 2007.
24. Darragh AJ, Hodgkinson SM. Quantifying the digestibility of dietary
protein. J Nutr 2000;130:1850S–6S.
25. Moughan PJ, Butts CA, Rowan AM, Deglaire A. Dietary peptides
increase gut endogenous amino acid losses in adult humans. Am J Clin
Nutr 2005;81:1359–65.
26. Gaudichon C, Bos C, Morens C, Petzke KJ, Mariotti F, Everwand J,
Benamouzig R, Daré S, Tomé D, Metges CC. Ileal losses of nitrogen and
amino acids in humans and their importance to the assessment of amino
acid requirements. Gastroenterology 2002;123:50–9.
27. Rutherfurd SM, Moughan PJ. Available versus digestible dietary amino
acids. Br J Nutr 2012;108:S298–305.
28. Rutherfurd SM, Moughan PJ. Determination of sulfur amino acids in
foods as related to bioavailability. J AOAC Int 2008;91:907–13.
Downloaded from jn.nutrition.org by guest on October 24, 2017