j o u r n a l o f d e n t i s t r y 3 6 ( 2 0 0 8 ) 2 9 4 – 2 9 9

avai lable at www.sc iencedi rec t .com

journal homepage: www. int l .e lsev ierhea l th .com/ journals / jden

An in vivo investigation of associations between salivaproperties, caries prevalence and potential lesionactivity in an adult UK population

Sachin Varma, Avijit Banerjee *, David Bartlett

Department of Conservative Dentistry and Prosthodontics, King’s College London Dental Institute at Guy’s,

King’s College & St. Thomas’ Hospitals, Guy’s Dental Hospital, London Bridge, London SE1 9RT, UK

a r t i c l e i n f o

Article history:

Received 2 October 2007

Received in revised form

9 January 2008

Accepted 15 January 2008

Keywords:

Caries prevalence

Lesion activity

Caries detection

DMFT

Epidemiology

ICDAS

Saliva

a b s t r a c t

Objective: To investigate associations between prevalence and activity of intra-oral incipi-

ent, carious lesions and salivary properties tested using the Saliva Check kit (GC Corp.,

Belgium).

Methods: With ethical approval, 58 subjects with >16 teeth underwent clinical/radiographic

examination. Conventional decayed, missing, filled teeth/decayed, missing, filled surfaces

(DMFT/DMFS) indices and a more recently developed visual index described by International

Caries Detection and Assessment System (ICDAS) were used to ascertain caries prevalence.

Potential lesion activity was scored using an Ekstrand visual index. Saliva properties tested

included hydration, resting pH, stimulated flow and buffering capacity. Spearman’s rank

correlation was used to analyse data.

Results: No saliva parameters correlated significantly with DMFT/DMFS caries prevalence

scores (D3 threshold). The resting pH correlated negatively and significantly with the total

number of lesions (r = �0.267, p = 0.043), with ICDAS scores>1 (r = �0.333, p = 0.011) and with

mild lesions (r = �0.263, p = 0.046). A negative correlation was found between saliva buffer-

ing capacity and the potential activity of moderate lesions (ICDAS 3 and 4; r = �0.227,

p = 0.035).

Conclusions: There appeared to be a correlation between the resting pH of saliva and the

prevalence of early lesions as well as the saliva buffering capacity and the potential lesion

activity of moderate lesions. A difference was shown between lesion prevalence calculated

T(S) D3 versus the ICDAS D1 thresholding.

# 2008 Elsevier Ltd. All rights reserved.

using traditional DMF

1. Introduction

There have been few reports on the prevalence of incipient,

pre-cavitated lesions. Pitts1,2 described an epidemiological

‘‘iceberg’’ to explain caries progression with a D4 lesion

extending into the pulp and D1 representing pre-cavitated,

incipient enamel lesions. Lesions recorded at the D3

threshold are defined as those lesions involving dentine,

* Corresponding author at: Floor 26, Guy’s Dental Hospital, London BridgE-mail address: avijit.banerjee@kcl.ac.uk (A. Banerjee).

0300-5712/$ – see front matter # 2008 Elsevier Ltd. All rights reservedoi:10.1016/j.jdent.2008.01.009

cavitated or not. Pitts2 stated that the D3 threshold was

most commonly reported by classical epidemiological

studies surveying caries incidence, prevalence and aetiolo-

gical relationships.

The most common method of recording caries for

epidemiological studies is the DMFT index (decayed, missing,

filled teeth) and DMFS (decayed, missing, filled surfaces).3 The

problems with this index include the missing and filled

e, London SE1 9RT, UK. Tel.: +44 207 188 1577; fax: +44 207 188 7486.

d.

Fig. 1 – ICDAS scores of the detection and histological severity of the carious lesion observed (reproduced with kind

permission from K.R. Ekstrand).

j o u r n a l o f d e n t i s t r y 3 6 ( 2 0 0 8 ) 2 9 4 – 2 9 9 295

components not always being indicative of past caries

experience, leading to a real risk of overestimation of the

caries experience in a population.4 Traditionally, this index

has been used in epidemiological data for recording at the D3

threshold but opinion exists advocating recording at the D1

threshold in order to obtain a more realistic picture of the

extent of dental disease and its distribution.5–8 A more

recently developed International Caries Detection and Assess-

ment System (ICDAS) (Fig. 1) has been recognised for use in

epidemiological studies.1,9–11

Investigators have developed classification systems linking

visual and/or radiographic appearance to lesion histology as

well as attempting to predict potential caries activity by

recording the state of the carious lesion at a snapshot in

time.5,6,8,12–14 It has been reported that operator opinion

regarding lesion activity is variable and unreliable using visual

and tactile investigation alone.15 Bleeding on probing near a

carious site can be predictive for lesion activity; plaque alone

however has a less predictive power.7 Potential lesion activity

might be calculated using weighted numerical values for

Fig. 2 – Final determination of lesion activity using weighted sc

(with kind permission from K.R. Ekstrand).

lesion appearance, lesion location in relation to a cariogenic

plaque stagnation area, surface integrity and association to

gingival bleeding (Fig. 2).

Investigations of salivary properties (including flow rate

and buffering) and their association with caries prevalence in

adult populations have been documented in the past with

conflicting results.16–20 Only one report has shown a positive

correlation between a low buffering capacity and caries

activity but this analysis was subjective and operator-

dependent.17 The advent of a simple chairside saliva check

kit (GC Saliva check, GC Corp., Belgium) analysing levels of

labial hydration of unstimulated saliva, the resting pH, volume

of stimulated saliva and its buffering capacity together with

the development of the new visual indices, creates the

potential for using these markers to predict lesion activity.

The aim of this study was to investigate the null hypothesis

that there is no association between the prevalence or activity

of intra-oral incipient, carious lesions and the saliva proper-

ties, tested using the Saliva Check kit (GC Corp.) in an adult UK

population.

oring >7 = potentially active and 0–7 = potentially in-active

Table 1 – Mean and S.D. for caries prevalence usingconventional DMFT(S) (D3 threshold) and ICDAS scoring

Mean per patient S.D.

DMFT (D3 cut-off) 18.02 6.94

DFT (D3 cut-off) 12.21 5.40

DMFS (D3 cut-off) 58.78 28.16

DFS (D3 cut-off) 30.88 18.90

No. of lesions recorded at D1 4.33 3.23

ICDAS scores above 1 3.64 3.04

ICDAS scores above 2 1.28 1.61

ICDAS scores above 3 0.67 1.05

ICDAS scores above 4 0.43 0.75

ICDAS scores above 5 0.17 0.17

Score 1 + 2 (pre-cavitated lesion) 1.12 1.11

Score 3 + 4 (moderate lesion) 0.37 0.71

Score 5 + 6 (severe lesion) 0.13 0.29

No. of active lesions 1.55 2.35

j o u r n a l o f d e n t i s t r y 3 6 ( 2 0 0 8 ) 2 9 4 – 2 9 9296

2. Materials and methods

Fifty-eight patients presenting to the Restorative Assessment

Clinic at King’s College London Dental Institute at Guy’s

Hospital for routine care by undergraduate students were

recruited with ethical approval from the Guy’s and St.

Thomas’ Hospital Ethics Committee (04/Q0704/117). Subjects

were included if they had no removable dental prostheses, a

minimum of 16 teeth and had not eaten for 2 h prior to the

appointment. Subjects were at least 18 years of age but with no

upper age limit. The medical history of the patients was

assessed and those patients requiring antibiotic prophylaxis

and/or receiving medications that can cause hyposalivation

including antidepressants, diuretics, antihistamines, narco-

tics and b-adrenoreceptor agonists were excluded from the

study.21

2.1. Caries detection and diagnosis

The smooth and occlusal surfaces of teeth were cleaned with a

rotating bristle brush, dried and examined using magnifica-

tion (2�). As the subjects were due further treatment by

undergraduates, bite-wing radiographs were taken using film

holders to aid lesion detection on approximal tooth surfaces.

The radiographs were used to detect lesions at the D1 and D3

level. DMFT, DFT, DMFS and DFS scores at the D3 level were

recorded. International Caries Detection and Assessment

System scoring was used to detect caries on all tooth surfaces

in order to record both cavitated and pre-cavitated lesions

(Fig. 1). Expert training was given for the ICDAS9 visual and

Ekstrand activity indices and practised in a clinical pilot study

prior to definitive data collection.

2.2. Caries activity

An activity score for each lesion was calculated from the visual

appearance, in association with a plaque stagnation area,

lesion surface integrity (using a ball-ended CPITN probe) and

associated gingival bleeding for lesions located at the gingival

third of the crown. A final, weighted activity score was then

calculated, as per Fig. 2, where lesions scoring>7 were classed

as potentially active. All caries detection and scoring was

performed independently by two examiners with final agree-

ment decided at the time of examination.

2.3. Saliva analysis

Samples were taken between 9 a.m. and 12 noon. Using the GC

Saliva Check Kit (GC Corp., Belgium) the unstimulated flow

rate was measured visually, noting the time taken for a

salivary droplet to form on the lower lip. A time greater than

60 s was considered as an abnormally low (according to the

manufacturer’s instructions). Subjects were then asked to

pool their saliva in the floor of the mouth and then expectorate

over 30 s. A pH strip was dipped in the saliva and the colour

used to estimate the pH. To stimulate saliva subjects were

given paraffin wax to chew and saliva was collected for 5 min

in a measuring cup and the volume was calculated. From the

collected saliva a drop was placed on a buffering strip, left for

5 min and then measured on the pH scale provided.

2.4. Reproducibility

Intra-examiner reproducibility was confirmed by randomly re-

examining seven patients (12% of the total) and re-recording

the visual indices for caries detection, activity and saliva

properties. Re-examination was carried out on the same day at

different times.

2.5. Statistical analysis

The highest score was recorded for teeth with multiple lesions

on multiple surfaces. The scores were classified by recording

the total number of lesions per patient and dividing this by the

number of teeth per patient to give a proportionate score. The

scores were grouped into mild or pre-cavitated enamel lesions

(ICDAS 1 and 2), moderate open or closed dentine lesions

(ICDAS 3 and 4), and severe cavitated lesions with possible

pulpal involvement (ICDAS 5 and 6). Statistical analysis was

carried out using Stata 8 statistical software (Stata Co., TX,

USA). The sample size permitted the use of the arithmetic

mean as an average and standard deviation as an indication of

spread. Correlations between salivary properties and caries

prevalence and activity were analysed with Spearman’s rank

correlations. Statistical significance was set at p < 0.05, r-

values were calculated to show positive or negative correla-

tions. Reproducibility was determined by using the Kappa

statistic.

3. Results

All 58 subjects with a mean age of 52.7 years (standard

deviation (S.D.) 15.6; age range 22–79 years) comprising 26

males (44.8%) and 32 females (55.2%) successfully completed

the study. The average number of teeth per patient was

26.1(3.2). Of these 13(22.4%) reported smoking.

3.1. Caries detection

Table 1 summarises the mean (S.D.) for caries prevalence in

the sample population using DMFT(S) and ICDAS. The mean

j o u r n a l o f d e n t i s t r y 3 6 ( 2 0 0 8 ) 2 9 4 – 2 9 9 297

DMFT score per patient was 18.02 (S.D. 6.94) and the mean

DMFS score was 58.78 (S.D. 5.40) using a D3 threshold. The

mean total number of lesions detected using the ICDAS

scoring at D1 threshold was 4.33 (S.D. 3.23). The mean number

of pre-cavitated lesions was 1.12 (S.D. 1.11). The mean for

moderate lesions was 0.37 (S.D. 0.71) and the mean for severe

lesions was 0.13 (S.D. 0.29). The mean number of active lesions

diagnosed was 1.55 (S.D. 2.35).

3.2. Saliva examination

The saliva hydration scores had a mean of 32.5 s per patient

(S.D. 19.5). The mean pH was calculated as 6.8 (S.D. 0.4).

Stimulated volume was recorded over 5 min and a mean was

5.9 ml (S.D. 2.8), calculated as millilitre/minute the mean was

1.18 ml/min. The mean buffer score was 9.7 (S.D. 2.5).

3.3. DMFT (D3) caries detection scores and salivaryproperties

Table 2 summarises the Spearman correlations of the DMFT

(D3) prevalence scores as a proportion of the teeth against

salivary properties. There were no statistically significant

correlations observed.

3.4. ICDAS caries scores and salivary properties

Table 3 shows the correlations between salivary properties

and caries detection recorded as a proportion of teeth for

Table 2 – DMFT(S) caries prevalence scoring at the D3 level corfound)

DMFT r( p)

Hydration (s) unstimulated 0.149(0.265) �0

pH �0.004(0.974) 0

Volume (ml) stimulated �0.076(0.570) �0

Buffer 0.052(0.699) 0

Table 3 – Spearman’s rank correlations with ICDAS visual ind

Shaded cells show statistical significance.

cumulative scores above 0–5 and grouped scores for mild,

moderate and severe lesions, using ICDAS. The hydration,

stimulated volume and buffering parameters investigated did

not correlate significantly with the visual index scores tested.

The resting pH correlated significantly and negatively with the

total lesions (p = 0.043, r = �0.267), scores above 1 ( p = 0.011,

r = �0.333) and mild lesions (p = 0.046, r = �0.263).

3.5. Potential lesion activity correlated with salivaryproperties

Table 4 shows the Spearman’s rank correlations for salivary

properties and activity of the lesion as a proportion of teeth.

There was a significant negative correlation between saliva

buffering and the activity of moderate lesions (ICDAS 3 and 4).

3.6. Kappa reproducibility

The kappa scores for intra-examiner reproducibility were

0.65–1.00, with a mean score of 0.83 indicating a high level of

agreement.22

4. Discussion

As far as the authors are aware, there is no current published

evidence of resting salivary pH correlating with caries

prevalence or activity in adults. The only evidence reported

on pH has been done on stimulated salivary pH with one study

related with the salivary properties tested (no significance

DFT r(p) DMFS r(p) DFS r(p)

.227(0.086) �0.005(0.973) �0.216(0.1041)

.125(0.350) �0.085(0.524) 0.082(0.541)

.016(0.904) �0.150(0.263) �0.003(0.998)

.176(0.187) �0.104(0.437) 0.142(0.290)

ex scores as a proportion of teeth and salivary properties

Table 4 – Spearman’s rank correlations with active lesions’ ICDAS scores as a proportion of teeth and the salivaryproperties

Shaded cell shows statistical significance.

j o u r n a l o f d e n t i s t r y 3 6 ( 2 0 0 8 ) 2 9 4 – 2 9 9298

showing a negative correlation between stimulated salivary

pH and caries prevalence23 and another showing a positive

correlation, caries prevalence both detected at the D3 thresh-

old.24 In the latter study those patients with a low resting pH

had a larger number of pre-cavitated lesions and may reflect

that this population was more likely to develop early lesions.

Pre-cavitated lesions are in a constant state of dissolution and

repair depending on a critical pH and pH fluctuations can

cause a loss of mineral from the tooth when the pH drops and a

gain in mineral when the pH rises.25

Plaque pH and caries activity have been previously

investigated26 and there is evidence to suggest a fall in plaque

pH below a critical point will cause caries on certain prone

individuals.27 Whether the plaque pH was a factor in

influencing the resting salivary pH is not known and requires

further investigation. Moderate or severe lesions had no

significant correlation with caries prevalence. For moderate

and severe lesions other aetiological factors may override any

saliva effects, as caries is a multifactorial disease.28

Previous methods to measure unstimulated saliva were

described by Navazesh and Christensen29 but they did not

include using the lower labile minor salivary gland hydration

time. The estimation of buffering capacity has been investi-

gated using similar methods using the Dentobuff (Orion

Diagnostics, Helsinki).30 Widodo et al. reported the results of a

study on a group of patients with dental erosion using the GC

Saliva Check Kit (GC Corp., Belgium).31 Although, the methods

used to record the saliva properties are unique to this product

the data recorded in this study appear to correlate with

previously published literature.15,31–35

Results from this study showed no correlations with

hydration, stimulated volume and stimulated buffering with

the caries scores using the modified visual indices and

supports the findings from studies.19,36,37 There has been only

one study in the literature and that was on a child population

which found a low unstimulated flow rate associated with an

increase in DFS.34 This is contrary to the correlations found

between hydration and caries detection in this study and may

reflect the age difference between the sample populations.

The resting pH was found to be negatively and significantly

correlated to the total lesions recorded at the D1 threshold,

scores above 1 and the pre-cavitated scores using the ICDAS

visual index (Table 3). However, there were no correlations

with pH and caries scores recorded with DMFT/DMFS indices

detecting caries at the D3 level (Table 2). This finding was

possibly due to the lack of sensitivity of the DMFT/DMFS

indices used conventionally at the D3 thresholding.1,38

Potential activity of the ICDAS 3 and 4 lesions did show a

negative correlation with saliva buffering capacity. Wiktorsson

et al. reported that lesion activity was correlated to saliva but

used an operator-based, subjective opinion to determine the

activity of the lesion.17 The results presented require further

development to validate the scoring system and to investigate

these potentially clinically useful associations, especiallyuseful

when performing clinical caries risk assessments. The follow-

ing conclusions from this study were deduced:

1. S

alivary parameters (hydration, resting pH, stimulated flow

and stimulated buffering) as tested by the Saliva Check kit

(GC Corp.) showed no correlations with caries prevalence

scored using DMFS/DMFT at the D3 threshold.

2. T

here was a negative correlation with resting pH and total

carious lesions detected at the pre-cavitated threshold. Pre-

cavitated lesions alone correlated negatively with resting pH.

3. T

he potential activity of moderate (ICDAS 3 and 4) lesions

did negatively correlate with salivary buffering.

The null hypothesis was therefore rejected. The clinical

implications of the findings of this study indicate that the

resting saliva pH may be a predictor in patients developing

pre-cavitated lesions. In terms of lesion activity, the buffering

capacity might help to predict the potential activity of dentine

lesions present but further experimentation and validation is

required.

j o u r n a l o f d e n t i s t r y 3 6 ( 2 0 0 8 ) 2 9 4 – 2 9 9 299

Acknowledgement

The authors would like to acknowledge the assistance

provided by Dr. Ron Wilson with the statistical analysis

completed in this study.

r e f e r e n c e s

1. Pitts NB. Modern concepts of caries measurement. Journal ofDental Research 2004;83:C43–7.

2. Pitts NB. Clinical diagnosis of dental caries: a Europeanperspective. Journal of Dental Education 2001;65:972–8.

3. World Health Organisation. A guide to oral healthepidemiological investigations. Geneva: Oral Health Unit;1997.

4. Broadbent JM, Thompson WM. For debate: problems withthe DMF index pertinent to dental caries data analysis.Community Dental and Oral Epidemiology 2005;33:400–9.

5. Ekstrand KR, Ricketts DNJ, Kidd EAM. Reproducibility andaccuracy of three methods for assessment ofdemineralisation depth on the occlusal surface: an in vitroexamination. Caries Research 1997;31:224–31.

6. Ekstrand KR, Ricketts DNJ, Kidd EAM, Qvist V, Schou S.Detection, diagnosing, monitoring and logical treatment ofocclusal caries in relation to lesion activity and severity: anin vivo examination with histological validation. CariesResearch 1998;32:247–54.

7. Ekstrand KR, Bruun G, Bruun M.. Plaque and gingival statusas indicators for caries progression on approximal surfaces.Caries Research 1998;32:41–5.

8. Nyvad B, Machiulskiene V, Baelum V. Reliability of a newcaries diagnostic system differentiating between active andinactive caries lesions. Caries Research 1999;33:252–60.

9. International Caries Detection and Assessment SystemCoordinating Committee. International Caries Detectionand Assessment System (ICDAS II) Workshop held inBaltimore, MD, March 12–14, 2005. Available http://www.dundee.ac.uk/dhsru/docs/ICDAS%20II%20criteria%20document%20September%2010.doc [accessed 25/1/06].

10. Burt BA, Kolker JL, Sandretto AM, Yuan Y, Sohn W, Ismail AI.Dietary patterns related to caries in a low-income adultpopulation. Caries Research 2006;40:473–80.

11. Sohn W, Ismail A, Amaya A, Lepkowski J. Determinants ofdental care visits among low-income African-Americanchildren. Journal of the American Dental Association2007;138:309–18.

12. Ekstrand KR, Ricketts DNJ, Kidd EAM. Occlusal caries:pathology, diagnosis and logical management. Dental Update2001;28:380–7.

13. Ricketts DNJ, Ekstrand KR, Kidd EAM, Larsen T. Relatingvisual and radiographic ranked scoring systems for occlusalcaries detection to histological and microbiologicalevidence. Operative Dentistry 2002;27:231–7.

14. Kidd EAM, Banerjee A, Ferrier S, Longbottom C, Nugent Z.Relationships between a clinical-visual scoring system andtwo histological techniques: a laboratory study on occlusaland approximal carious lesions. Caries Research 2003;37:125–9.

15. Ekstrand KR, Ricketts DNJ, Longbottom C, Pitts NB.Visual and tactile assessment of arrested initial enamelcarious lesions: an in vivo pilot study. Caries Research2005;39:173–7.

16. Rask PI, Emilson CG, Krasse B, Sundgerg H. Dental cariesand salivary and microbial conditions in 50–60-year-oldpersons. Community Dental Oral Epidemiology 1991;19:93–7.

17. Wiktorsson A-M, Martinsson T, Zimmerman M. Salivarylevels of lactobacilli, buffer capacity and salivary flow raterelated to caries activity among adults in communities withoptimal and low water fluoride concentrations. SwedishDental Journal 1992;16:231–7.

18. Mazengo MC, Soderling E, Alakuijala P, Tiekso J, Tenovuo J,Simell O, Hausen H. Flow rate and composition of wholesaliva in rural and urban Tanzania with special reference todiet, age, and gender. Caries Research 1994;28:468–76.

19. Mazengo MC, Tenovuo J, Hausen H. Dental caries in relationto diet, saliva and cariogenic microorganisms in Tanzaniansof selected age groups. Community Dental Oral Epidemiology1996;24:169–74.

20. Powell LV, Leroux BG, Persson RE, Kiyak HA. Factorsassociated with caries incidence in an elderly population.Community Dental Oral Epidemiology 1998;26:170–6.

21. Tenovuo J.. Human saliva clinical chemistry and microbiology,vol. 1. Florida: CRC Press; 1989. p 26–73.

22. Landis JR, Koch GG. The measurement of observeragreement for categorical data. Biometrics 1977;33:159–74.

23. Johansson I, Saellstrom AK, Rajan BP, Parameswaran A.Salivary flow and dental caries in Indian children sufferingfrom chronic malnutrition. Caries Research 1992;26:38–43.

24. Crossner C-G, Holm A-K. Saliva tests in the prognosis ofcaries in children. Acta Odontologica Scandinavia 1976;35:135–9.

25. Manji F, Fejerskov O, Nagelkerke NJD, Baelum V. A randomeffects model for some epidemiological features of dentalcaries. Community Dental Oral Epidemiology 1991;19:324–8.

26. Fejerskov O, Scheie AA, Manji F. The effect of sucrose onplaque pH in the primary and permanent dentition ofcaries-inactive and -active Kenyan children. Journal of DentalResearch 1992;71:25–31.

27. Stephan RM. Intra-oral hydrogen-ion concentrationsassociated with dental caries activity. Journal of DentalResearch 1944;23:256–7.

28. Fejerskov O, Kidd E. Dental caries. The disease and its clinicalmanagement. Oxford/Blackwell Munksgaard; 2003.

29. Navazesh M, Christensen CM. A comparison of wholemouth resting and stimulated salivary measurementprocedures. Journal of Dental Research 1982;61:1158–62.

30. Ericson D, Bratthall D. Simplified method to estimatesalivary buffer capacity. Scandinavian Journal of DentalResearch 1989;97:405–7.

31. Widodo G, Wilson R, Bartlett D. Oral clearance of an acidicdrink in patients with erosive tooth wear compared withthat in control subjects. International Journal of Prosthodontics2005;18:323–7.

32. Ericsson Y, Hardwick L. Individual diagnosis, prognosis andcounselling for caries prevention. Caries Research 1978;12:94–103.

33. Edgar WM. Saliva: its secretion, composition and functions.British Dental Journal 1992;172:305–12.

34. Vehkalahti M, Nikula-Sarakorpi E, Paunio I. Evaluation ofsalivary tests and dental status in the prediction of cariesincrement in caries-susceptible teenagers. Caries Research1996;30:22–8.

35. Dawes C. Factors influencing salivary flow rate andcomposition. In: Edgar M, Dawes C, O’Mullane D, editors.Saliva and oral health. 3rd ed. London: BDJ Books; 2004. p. 32–49.

36. Lundgren M, Emilson CG, Osterberg T. Root caries and somerelated factors in 88-year-old carriers and non-carriers ofStreptococcus sobrinus in saliva. Caries Research 1998;32:93–9.

37. Gabris K, Nagy G, Madlena M, Denes Z, Marton S, KeszthelyiG, Banoczy J. Associations between microbiological andsalivary caries activity tests and caries experience inHungarian adolescents. Caries Research 1999;33:191–5.

38. Nyvad B. Diagnosis versus detection of caries. CariesResearch 2004;38:192–8.