SCIENTIFIC ARTICLE

Australian Dental Journal 2008; 53: 140144

doi: 10.1111/j.1834-7819.2008.00023.x

Comparative analysis of three commercial saliva testing kits

with a standard saliva buffering test
Y Kitasako,* MF Burrow, M Stacey, L Huq, EC Reynolds, J Tagami*
*Cariology and Operative Dentistry, Department of Restorative Sciences, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan.
School of Dental Science, The University of Melbourne, Victoria, Australia.
Cooperative Research Centre for Oral Health Science, School of Dental Science, The University of Melbourne, Victoria, Australia.
Center of Excellence Program for Frontier Research on Molecular, Destruction and Reconstruction of Tooth and Bone, Tokyo Medical and
Dental University, Tokyo, Japan.

ABSTRACT
Background: In 1959, Ericsson developed a laboratory buffer capacity test. Because the Ericsson test is not practical for use
as a chair-side test, commercially available saliva buffering capacity tests have been developed for use in the dental office.
The purpose of this study was to evaluate the correlation between a modified Ericsson test and three commercially available
quantitative and colourimetric tests.
Methods: Stimulated saliva (by chewing paraffin wax) was collected from 113 patients. Individual saliva buffering capacity
was assessed with the following four different methods: modified Ericsson test; quantitative test using a hand-held pH meter;
paper strip; or liquid colourimetric test. The correlations of ranking results among the different tests were analysed using the
Spearman Rank Correlation Test, p < 0.001.
Results: Spearman Rank Correlation indicated significant positive coefficients between the modified Ericsson test and the
quantitative test (q = 0.857), the paper strip colourimetric test (q = 0.621) and the liquid-type colourimetric test
(q = 0.689).
Conclusion: The detection level of medium and high buffering capacity was test dependent. The quantitative test using a
hand-held pH meter showed a stronger positive correlation with the modified Ericsson test. The qualitative tests seemed less
reliable, particularly for patients classified as having a medium buffering capacity.
Key words: Saliva, caries risk, pH, buffer capacity.
(Accepted for publication 30 August 2007.)

INTRODUCTION
Salivary buffering capacity has been identified as one of
the many factors that may affect an individuals caries
risk.1 The ability of saliva to buffer acids is essential for
maintaining pH values in the oral environment above
the critical pH, thereby protecting teeth against demineralization.24 It is imperative that measurement of
salivary buffering capacity is accurate so that appropriate preventive management may be implemented for
individual patients.
The buffering capacity of stimulated whole saliva is
mainly determined by bicarbonate and phosphate ions,
concentrations of which are slightly higher in parotid
than in submandibular saliva.5 In 1959, Ericsson6
developed a laboratory buffer capacity test involving
the addition of hydrochloric acid and the elimination of
carbon dioxide by bubbling air through the saliva
sample. The resulting final pH was an acceptable
140

measure of salivary buffering capacity. This test has

become the gold standard for the assessment of saliva
buffering capacity.
However, the Ericsson test is not practical for use as
a chair-side test, and consequently other saliva buffering capacity tests have been developed for use in the
dental office. These tests are generally colourimetric,
using either a colourimetric paper strip, or colourimetric liquid to determine saliva buffering capacity. It has
been noted however, that the results from colourimetric
kits are sometimes inconclusive. Ericson and Bratthall7
indicated that the colour reaction was dependent on the
time the saliva sample was in contact with the colourchanging strip or solution.
To overcome this problem, it seems more useful if
saliva buffering capacity is measured using a quantitative method. Recently, a quantitative saliva buffering
capacity test has become available using a small handheld electronic pH meter.8,9 Quantitative pH analysis is
2008 Australian Dental Association

Analysis of saliva testing kits

performed by the simple process of collecting a saliva
sample and titrating lactic acid onto the electrode of the
pH meter. This quantitative method is now employed
for determining salivary pH and acid buffer capacity as
a chair-side test.
The purpose of this study was to evaluate the
correlation between a modified Ericsson test (quantitative laboratory gold standard test) and three commercially available kits (quantitative test using a hand-held
pH meter, and strip-type or liquid-type colourimetric
test). The null hypothesis is that the individual saliva
buffering capacity is not associated with different
methods for stimulated saliva buffering capacity.
MATERIALS AND METHODS

Modified Ericsson test (quantitative standard test)

The saliva sample was mixed by inverting the collection
tube twice, 1.0 mL of saliva was transferred to 3.0 mL
of HCl (0.005 N). Then a stream of nitrogen was
passed through the mixture for 20 minutes to eliminate
carbon dioxide from the sample. The final pH of the
solution was assessed using an electronic pH meter and
this final pH was determined as being the buffering
capacity of the saliva sample. In this study, individual
salivary buffering capacities obtained from the modified
Ericsson test were ranked into one of the following
three buffering capacity categories: high (above pH
5.5); medium (from pH 5.5 to pH 4.5); and low (below
pH 4.5).6

Study population

Quantitative test using a hand-held pH meter

The study population consisted of 113 subjects

(46 males and 67 females, aged from 21 to 66 years,
mean 37.8) who were all staff members of the School of
Dental Science, The University of Melbourne, Australia. Approval for the study was obtained from the
Human Ethics Research Committee of The University
of Melbourne, and signed informed consent was
obtained from each subject prior to entry into the study.

Saliva pH change was measured directly using the

hand-held electronic pH meter with a digital display.
After calibration using the supplied standard solutions of pH 4.0 and 7.0, 0.25 mL of saliva was
placed onto the pH-sensitive electrode to measure
the initial pH value within 30 seconds. Two hundred
and fifty microlitres of lactic acid (pH 3.0, 15 mM)
was then titrated into the test saliva and mixed for
30 seconds using the manufacturers auto-mixer.
Salivary buffering capacities were ranked into one
of the following three buffering capacity categories
specified by the manufacturer: high (above pH 5.8);
medium (from pH 4.8 to pH 5.7); and low (below
pH 4.7).

Saliva sampling and flow rate

A stimulated whole saliva sample was collected from
each subject. Subjects were seated and relaxed for
several minutes prior to saliva collection. A 1-gram
piece of unflavoured paraffin wax was chewed for
30 seconds and the saliva was collected and discarded.
The test sample was then continuously collected into an
ice-cooled vial for 5 minutes, whilst chewing the
paraffin wax. Saliva collection was taken at least
2 hours after meals and at least 1 hour after brushing
to minimize effects of the diurnal variability in salivary
composition.10
Saliva pH and buffering capacity
Individual saliva buffering capacity was assessed by four
different methods (Table 1): modified Ericsson test;
quantitative test using a hand-held pH meter; colourimetric paper strip test; and liquid colourimetric test.

Colourimetric paper strip test

Using a pipette, sufficient saliva was drawn from the
saliva sample, and one drop was dispensed onto each of
the three test pads on the test strip. The test pads
changed colour and after 2 minutes the final results
were read and classified following five standard colours
supplied by the manufacturer: Green (4 points);
Green Blue (3 points); Blue (2 points); Blue Red
(1 point); and Red (0 points). After calculation of the
total points, the salivary buffering capacity was ranked
into one of the following three categories: Very Low
(from 0 to 5 points); Low (from 6 to 9 points); and
Normal (from 10 to 12 points).

Table 1. Saliva test used

Modified Ericsson test

Checkbuf
Saliva Check Buffer test
Oral Tester Buffer

Manufacturer and location

Quantitative colourimetric

Equipment

Laboratory test (standard test)

J. Morita, Tokyo, Japan
GC Co., Tokyo, Japan
Tokuyama Dental Co., Tokyo, Japan

quantitative
quantitative
colourimetric
colourimetric

electronic pH meter, nitrogen gas

hand-held pH meter
paper strip
tube (liquid)

2008 Australian Dental Association

141

Y Kitasako et al.
Liquid colourimetric test

Table 3. Results of the buffer capacity findings

The pipette supplied with the test kit was used to

transfer 0.5 mL of saliva into the test tube containing
tartaric acid (approximately 0.1 per cent) supplied by
the manufacturer. The tube was sealed and then shaken
45 times to mix the saliva and reagent. The colour of
the reagent changed according to the buffering capacity. This colour was compared with a colour chart
supplied in the kit and from this salivary buffering
capacities were ranked into one of the following three
buffering capacity categories: Low (Yellow); Medium
(Orange or Red); and High (Purple).

Saliva test

Total
number

Modified Ericsson test

Quantitative test using a
hand-held pH meter
Colourimetric paper strip test
Liquid colourimetric test

Rank
High Medium

Low

113
113

87
86

17
17

9
10

113
113

59
62

46
42

8
9

Table 4. Results of each category under different test

Modified Ericsson test
High

Statistical analysis
Each test ranked buffering capacity into three categories. However, the groupings given for each category
were not consistent among the three tests. Equivalent
categories were determined as indicated in Table 2.
For the purpose of discussion, in this study all results
in category 1 were designated Low, all results in
category 2 were designated Medium, and all results
in category 3 were designated High. The correlations
of ranking results among the different tests were
analysed using the Spearman Rank Correlation Test,
p < 0.001. All statistical calculations were performed
using the SPSS statistical software programme 10.01
(Chicago, II, USA).
RESULTS
For the modified Ericsson test (Standard Test), 87 of
113 subjects showed a high buffering capacity, 17
subjects medium buffering capacity, and the remaining
9 were classified as having a low buffering capacity
(Table 3). The Spearman Rank Correlation indicated
significant positive coefficients between the modified
Ericsson test and the quantitative test (q = 0.857), the
paper strip colourimetric test (q = 0.621), and the
liquid-type colourimetric test (q = 0.689).
Although the quantitative test showed the strongest
positive correlation with the modified Ericsson test
(q = 0.857), 4 out of 17 subjects who showed medium
buffering capacity in the modified Ericsson test were

Table 2. Categories for the buffering capacity among

each test
Category

Modified
Quantitative Colourimetric
Liquid
Ericsson test test using a paper strip test colourimetric
(standard test) hand-held
test
pH meter

1
2
3

Low
Medium
High

142

Low
Medium
High

Very low
Low
Low
Medium
Normal High High

Medium

Low

Saliva test
87
17
Quantitative test using a hand-held pH meter
High
83
3
Medium
4
13
Low
0
1

9
0
0
9

Colourimetric paper strip test

High
58
Medium
29
Low
0

1
16
0

0
1
8

Liquid colourimetric test

High
62
Medium
25
Low
0

0
17
0

0
0
9

classified as having either high or low buffering capacity

in the quantitative test (Table 4). For the paper strip
colourimetric test, 16 out of 17 cases showed medium
buffering capacity, the same as the Ericsson test.
However, 29 of 87 cases (33 per cent) were classified
as having a high buffering capacity with the modified
Ericsson test, but were classified as having medium
buffering capacity with the paper strip colourimetric
test. For the liquid colourimetric test, all 17 cases
showed medium buffering capacity, the same as the
Ericsson test. However, 25 of 87 cases (29 per cent)
that showed high buffering capacity for the modified
Ericsson test were classified as medium buffering
capacity for the liquid colourimetric test.
DISCUSSION
Ericsson6 reported that there were distinct differences
in saliva buffering capacity of caries-free and cariesactive patients. He developed a saliva buffering test
method that used the addition of hydrochloric acid
and also eliminated carbon dioxide that had dissolved
into the saliva by bubbling air through the saliva
sample. In his laboratory study, the resulting final pH
was an acceptable measure of saliva buffering capacity. In the current study, nitrogen was used to
eliminate dissolved carbon dioxide instead of the air
as used by Ericsson.6 This was done as a matter of
2008 Australian Dental Association

Analysis of saliva testing kits

convenience as cylinders of nitrogen were readily
available. Tubes were connected to the cylinder outlet
and the valve was opened to allow nitrogen to flow
through the saliva sample. This technique precluded
the need for a pump which would have been necessary
to pass air through the saliva. The modification to the
Ericsson method would not have affected the results of
the test.
In the case of the colourimetric methods, a change of
colour occurs which corresponds to a range of pH
values indicative of a buffer value. Although the paper
strip method is easy to use, the viscosity of sample
saliva seems to influence the final colour produced on
the paper. This is believed to be due to several possible
factors. First, the size of the drop of saliva from the
supplied syringe can vary depending on viscosity;
second, the more viscous the saliva, the less able it
seems to wet the paper; third, it is necessary to remove
excess saliva from the test strip, this is more difficult
with more viscous saliva samples. This results in there
possibly being more or less saliva for the volume of acid
in the colour pads on the strips that may result in
a variation of the final colour which can therefore
influence the final score. Another factor that makes
using the paper test difficult is discriminating the colour
of the sample, especially the blue-green colour range.
The outcome is patches of varying colours on the paper,
making it difficult to obtain an accurate result, particularly for a practitioner using the test for the first time.
In this study, 28 out of 113 saliva samples (25 per cent)
showed a high viscosity and recorded colour codes that
were difficult to assign points, hence making scoring
difficult. Therefore, it became difficult to conclude
which buffering capacity group these samples should be
assigned to. The liquid test is preferable for viscous
saliva samples as it is easy to mix the high viscosity
saliva with the solution.
The classification of high and medium buffering
capacity varied depending on whether the quantitative
or colourimetric method was used. Both colourimetric
tests detected the medium buffering capacity in similar
numbers as the modified Ericsson test. However,
approximately 30 per cent of medium buffering
capacity samples for both colourimetric tests were
classified as having high buffering capacity when
analysed with the modified Ericsson test. This would
indicate that the colourimetric tests potentially under
estimated the buffering capacity of some samples.
Although lactic acid used in the quantitative method is
safer than the HCl used in the modified Ericsson test,
the use of lactic acid could affect the classification of
medium buffering capacity due to its diminishing
dissociation at pH 4.0.6 The reason the pH groupings
for buffering capacity are different from the Ericsson
test is based on in-house tests comparing the use of
lactic acid and HCl. These tests concluded that a pH
2008 Australian Dental Association

range of 4.85.7 for lactic acid is equivalent to the pH

range of 4.55.5 for HCl (Horiba Ltd, Japan, personal
communication).
The preventive management of high and low buffering capacity groups is generally clear-cut.9,11,12 However, for those patients who show a medium buffering
capacity, it is less clear how a preventive dental
programme should proceed. As a number of the
medium buffering capacity patients may actually fall
into the high or low buffering capacity groups, it would
seem that these patients may need to be monitored
carefully to ensure they do not belong to one of the
other buffering capacity groups. Further long-term
clinical studies investigating patients who have medium
buffering capacity are required to determine if there is
any association between medium buffering capacity
and the progression of dental caries.
For the colourimetric method, colour matching with
supplied colour guides can be problematic.13 Operators colour perception is subjective and may be affected
by colour vision deficiencies, ambient lighting and
operator experience.14 This could lead to errors and
inconsistencies arising that may affect the formulation
of treatment plans.15 The results from this study
showed only minor differences between the results
obtained with the modified Ericsson test and the
quantitative method. The quantitative saliva buffer
capacity test using a hand-held pH meter eliminates the
subjectivity of the colourimetric saliva buffering capacity tests, and may provide a clearer indication to the
potential caries risk of patients.
CONCLUSIONS
Individual saliva buffering capacity can vary depending
on different evaluation methods used for stimulated
saliva buffering capacity. The hypothesis was rejected
as it seems that the detection level of medium and high
buffering capacity is test dependent. The quantitative
test using a hand-held pH meter showed a stronger
positive correlation with the modified Ericsson test.
The qualitative tests using a hand-held pH meter seem
less reliable, particularly for patients classified as
having a medium buffering capacity. It was demonstrated that approximately 30 per cent of subjects in
this group may belong to the high or low buffering
capacity groups.
ACKNOWLEDGEMENTS
This project was supported by Grant #17390507 and
#18592083 from the Japan Society for the Promotion
of Science, and the Center of Excellence Program for
Frontier Research on Molecular Destruction and
Reconstruction of Tooth and Bone in Tokyo Medical
and Dental University.
143

Y Kitasako et al.
REFERENCES
1. Messer LB. Assessing caries risk in children. Aust Dent J
2000;45:1016.
2. Mandel ID. The functions of saliva. J Dent Res 1987;66:623627.
3. Tenovuo J. Salivary parameters of relevance for assessing caries
activity in individuals and populations. Community Dent Oral
Epidemiol 1997;25:8286.
4. Larsen MJ, Jensen AF, Madsen DM, Pearce EI. Individual variations of pH, buffer capacity, and concentrations of calcium and
phosphate in unstimulated whole saliva. Arch Oral Biol
1999;44:111117.
5. Mandel ID, Wotman S. The salivary secretions in health and
disease. Oral Sci Rev 1976;8:2547.
6. Ericsson Y. Clinical investigations of the salivary buffering
action. Acta Odontol Scand 1959;17:131165.
7. Ericson D, Bratthall D. Simplified method to estimate salivary
buffer capacity. Scand J Dent Res 1989;97:405407.
8. Kitasako Y, Moritsuka M, Foxton RM, Ikeda M, Tagami J,
Nomura S. Simplified and quantitative saliva buffer capacity
test using a hand-held pH meter. Am J Dent 2005;18:147
150.
9. Moritsuka M, Kitasako Y, Foxton RM, Ikeda M, Tagami J,
Nomura S. Quantitative assessment for stimulated saliva flow
rate and buffering capacity in relation to different ages. J Dent
2006;34:716720.
10. Hong-Seop K, Sung-Woo L, Sung-Chang C, Young-Ku K. Oral
manifestations and salivary flow rate, pH, and buffer capacity in

144

patients with end-stage renal disease undergoing hemodialysis.

Oral Surg Oral Med Oral Pathol 1999;88:316319.
11. Mjor IA, Dahl JE, Moorhead JE. The age of restorations
at replacement in permanent teeth in general dental practice.
Acta Odontol Scand 2000;58:97101.
12. Zimmer S, Jahn KR, Barthel CR. Recommendations for the use of
fluoride in caries prevention. Oral Health Prev Dent 2003;1:4551.
13. Sproull RC. Color matching in dentistry. Part II. Practical
applications of the organization of color. J Prosthet Dent
1973;29:556566.
14. McMaugh DR. A comparative analysis of the colour matching
ability of dentists, dental students, and ceramic technicians. Aust
Dent J 1977;22:165167.
15. Stephen JD. Clinical color vision tests. Clin Exp Optom
2004;87:276293.

Address for correspondence:

Dr Yuichi Kitasako
Cariology and Operative Dentistry
Department of Restorative Sciences
Graduate School
Tokyo Medical and Dental University
5-45 Yushima 1-chome, Bunkyo-ku
Tokyo, Japan 113-8549
Email: kitasako.ope@tmd.ac.jp

2008 Australian Dental Association