Synnove leikanger\

Effects of food preservatives on Espen Bjertness2 and

Anne Aamdal Scheie 3

growth and metabolism of

'Department of Operative Dentistry, Dental
Faculty, University of Oslo, 2Section for
Epidemiology, Department of Community
Medicine, National Institute of Public Health,

plaque bacteria in vitro and Oslo, ' Department of Microbiology, Dental

Faculty, University of Oslo, Oslo, Norway

In VIVO

Leikanger S. Bjertness E, Scheie AAa: Effects of food preservatives on grow th

and metabolism of plaque bacteria in vitro and in vivo. Scand J Dent Res 1992;
100: 371--6.

The aims of the present study were to assess the consumption of food preserva-
tives during the l ast decades, and to study the effect of the prese rvatives, sorbic
a nd benzoic acid , on growth and glycolysis of oral bacteria in vitro , and on acid
forma tion by dental plaque in vi vo. Five consumption reports from the Central
Bureau of Statistics of Norway were used to estimate alterations in consumption


of staple food containing the two preservatives. A modified broth dilution method

was used to determine the MIC values of the preservatives against Streptococcus
sobrinus and Streptococcus sanguis. Extracellular 1 4C-glycolytic metabolites
were studied by HPLC a n a lyses. Plaque-pH meas urements were used to assess
possible effects on acid production. The consumption reports indicated increased

Key words: benzoic acid; dental caries; dental
co nsumption of preservatives. The in vitro testing suggested that lega l concentra- plaque; food preservatives; plaque pH; sor-
tions of prese rvatives may inhibit the growth of oral streptococci. However, the bic acid; streptococci; S. sa nguis; S. sobrinu s
preservatives did not inhibit in vitro glycolysis at tested concentrations. in vivo Anne Aamdal Scheie, Department of Micro-
testing with similar concentrations (0.4% w / v) showed a significant effect. A biology, Dental Faculty, University of Oslo, Box
higher concentration (2% w / v potassium sorbate) had a tendency to inhibit 1052- Blindern, 0316 Oslo, Norway
acid-formation by dental plaque even more. Accepted for publication 8 December 1991

Many industrialized countries have seen a pro-
nounced decline in the occurrence of dental caries
during the last decades (1, 2). This improvement
in dental health has been recorded since the late
sixties among children and adolescents in Norway
(3) and later also among adults (4, 5).
The most obvious explanation for this change
has been increased use of fluorides, particularly
in fluoride dentifrices. However, declining caries
prevalence was seen even before fluorides were ex-
tensively used (6-9).
If fluorides alone were not responsible for the
reduced caries prevalence there must be other con-
tributory factors. There is no data to indicate im-
proved oral hygiene (5), nor has the diet changed
dramatically (10). Some authors have suggested
that the oral microflora have become less patho-
genic (11) due, for instance, to extensive use of
antibiotics, but data does not substantiate such a
concept. On the other hand, increased consump-
tion of foodstuffs containing preservatives may,
theoretically, have contributed.
The general effect of preservatives is based on
inhibition of bacterial growth and metabolism (12).
It is the undissociated form of the preserving or-
ganic acids which is able to pass the cell membrane.
The antimicrobial effect of these substances is
therefore strongly pH-dependent and preservatives
are most effective in an acidic environment. During
a carbohydrate challenge to dental plaque, undis-
sociated preservative molecules may enter the bac-
teria, inhibit intracellular enzyme systems and in-
terfere with acid production.
The aims of the present study were 1) to deter-
mine whether the consumption of preservatives
had increased during the last decades, 2) to assess
the effect on bacterial growth and metabolism of
the two most commonly used preservatives, potas-
sium sorbate and sodium benzoate, on two selected
oral streptococci, and 3) to test whether these
agents have the potential to inhibit acid production
in dental plaque in vivo.
Material and methods
Consumption of food containing preservatives
Data from five reports concerning food consump-
tion from the Central Bureau of Statistics of Nor-

372 Leikanger et al.
way, during the period of 1958-85, were used to
estimate consumption of staple food (13). The esti-
mations of changes in consumption of the preser-
vatives sorbic acid and benzoic acid were based on
rules and regulations for food additives (14).
Questionnaires were sent to food-processing in-
dustries asking about sal es per year, market share,
and amount of sorbic and benzoic acid added to
their products during the l ast 20 yr. Together with
reports from the Central Bureau of Statistics of
Norway and the rules and regulations for food
additives (14) we were able to evaluate whether
changes in consumption of the two preservatives
during the last two decades had occurred.
Effect on bacterial growth
Streptococcus sanguis ATCC 10556 (NCTC 7863)
and Streptococcus sobrinus OMZ 176 (NCTC
10992), were grown in a static batch culture (37 C,
20 h) in BHI broth (BHI, Difco, Laboratories,
Detroit, MI, USA). The optical density was adjust-
ed to 0.8 at 650 nm for S. sobrinus and to 1.0 for
S. sanguis (Uvikon 810 kontron spectrophotome-
ter). The tests were performed using a modified
broth dilution method (15) in steril e microtiter
trays (Costar, Bradway, Cambridge, MA). A vol-
ume of 200 111 of sterile-filtered solutions of the
preservatives sodium benzoate and potassium sor-
bate (Sigma Chemical Company, St. Louis, MO,
USA) in BHI was added to the first well in each
row (Pipetman, Gilson, France) and serial 0.625
dilutions in BHI were made. Each well was inocu-
lated with 75 111 inoculum of the bacteria diluted
in BHI; S. sanguis 1:1000 and S. sobrinus 1:800.
After growth (24 h, 37 C) sedimented bacteria
were redispersed by shaking the trays for 20 min.
Bacterial growth was expressed as optical density
which was read in a microplate reader (MR 710,
Dynatech lab, Chantilly, VA). The minimum Inhib-
itory Concentration (MIC) was defined as the low-
est concentration of a test agent which inhibited
visible growth of the bacteria (OD=O.OlO). The
agents were tested in triplicate and the experiment
was repeated four times (S. sobrinus) and five times
(S. sanguis).
Effect on bacterial metabolism:
Non-preservative-preexposed cells - S. sobrinus
OMZ 176 was grown in a static batch culture at
3r C, and harvested by centrifugation in l ate expo-
nential phase. The harvested cells were washed
twice, and resuspended in 70 mM NaCl, 70 mM
KCl, and 5 mM MgC12 , pH 6.6 (control cells), or
in the same sal t solution containing either 30 or 60
mM sodium benzoate or 30 or 60 mM potassium
sorbate (final concentrations). The bacterial den-
sity was adjusted by optical density measurements
(OD = 2.0 at 680 mM). 200 111 of each of the re-
spective cell suspensions were preincubated in a
water bath for 15 min at 37 C. Glycolysis was then
initiated by adding 14C-glucose (final concentration
0.2 mM, specific activity 15 11Ci/llmol). Glycolysis
was stopped by centrifugation after 1 h. The super-
natants containing extracellular 1 4C-glucose metab-
olites were analyzed by HPLC analysis connected
to an on-line radioactivity detector.
Cells grown in presence of preservatives - S. sobri-
nus was grown in 200 111 BHI or in BHI containing
20 mM (final concentration) of either sodium ben-
zoate or potassium sorbate (14 h, 3r C). 14C-glu-
cose (final concentration 0.2 mM, specific activity
15 11Ci/llmol, was added to all cultures). The cul-
tures were harvested and the supernatant anal yzed
by HPLC.
Dental plaque was collected in phosphate buffer
(50 mM, pH 6.8), centrifuged (6000 rpm, 10 min,
4C), attd resuspended in a salt solution (pH 6.5).
The suspensions were ultrasonically dispersed (2 x
10 s, Sonifier-B12, Branson Sonic Power Company,
CT, USA) and distributed into 6 tubes, and washed
twice in salt solution. The average amount of
protein was 3.83 mg (SD 1.5) / ml. Glycolysis was
initiated as described above. The supernatants con-
kg
80 -----------, , _ FLOUR , MEAL and
' BAKERY PROD.
IFRUITS IBERRIES I
60
POTATOES
I MEAT
40
I VEGETABLES
!MINERAL WATERS!
20 FISH
SUGAR
!EDIBLE OILS
IFATsl
t:===------------- I CHEESE
SWEETS
58 76 79 82 85 year
Fig. 1. Cons umption of foodstuffs (k g/ person / year). - Food
s in which preservatives are allowed.--- - Foods wh ich regula
rly do not, or seldom, contain preservatives.


Food preservatives and dental plaque 373

Table I PS
OD 6 30 SB

MIC- values for sodium ben zoat e and potassium sorbat e against CONTROL
S. sobrinu s OM Z 176 and S. sa nguis 10556
S. sobrinus OMZ1 76 S. sanguis I0556 0.150
Test-agents MIC-value MIC-value

Sodium benzoate 58 mM l 2 102 mM O 0.100
Pota ssium sorbate 41 mM 8 72mM + ll
Mean SE.
0.050

tammg extracellular 14C-glycolytic metabolites

20 40 60 80 10 0 mM
were collected for HPLC analyses after 1 h glycol-
ysis. Fig. 3. Growth meas ured by optica l density (OD) of S. sanguis
10556 in Brain Hea rt Infusion (BHI) without addition of preser-
va tives (control) and in BHI with different concentra tion s of
Dental plaque prese rvatives, (PS - potassium sorbate, SB - sodium ben zoate).
Shaded area is sta ndard e rror (SE) of three parallels in five
Plaque pH measurements - Five medically healthy ex periments.
dental students with healthy gingival conditions
and no caries cavities volunteered for the study.
The subjects were instructed to avoid mechanical ven , CT, USA) connected to an Orion Research
oral hygiene and fluoride dentifrice during a 4-day millivolt-meter (Boston, MA, USA, model EA
plaque accumulation period . The test subjects used 940). A porous glass electrode (Beetrode MEPH-
sucrose-containing chewing-gum for at least 5 min l ,WP Instruments) was used as a reference elec-
every 2 h during daytime in order to standardize trode. A reference saltbridge was created by placing
plaque accumulation. At least 10 days of habitual the reference electrode and one of the test subject's
oral hygiene was instituted between the exper imen- fingers into a 3 M KCl solution.
tal periods. Plaque pH was monitored in situ with The measurements were performed on two
a palladium touch microelectrode (Beetrode diam- maxillary mesiobuccal surfaces of teeth posterior
eter 0.1 mm MEPH-I;WP Instruments, New Ha- to the canines in each test subject. The resting pH-
value of the 4-day-old plaque was measured and
its pH response to a 1-min mouthrinse with glucose
OD 630 (10 ml 5% w / v) served as individual control values.
pH was measured at the same sites 2, 5, 10, and

15 min after the rinse. Resting pH was measured

0.250 again after 1 h and as soon as the prerinse pH

values were regained , the rinsing and pH measure-

ments were repeated using a 5% w / v glucose solu-
0.200 tion containing 0.4% w / v of potassium sorbate as

the mouthrinse. The tests were also repeated on a

PS limited scale (3 subjects) using 2% w / v potassium
0.150 SB sorbate or 0.4% w / v of sodium benzoate. In order

CONTROL to compare pH responses over time between con-

trol and test, we calculated the area under the curve
0.100 as defined by: the curve connecting pH measure-
ments conducted at time intervals from 0 to time
15, a horizontal line through pH 4, a vertical line
0.050 through time 0 and time 15 (Area 15). Differences
were tested using a paired l-test.

Results
20 40 60 mM Consumption of food containing preservatives

Fig . 2. Growth meas ured by optical densi ty (OD) of S. sobrinus There has been a change in consumption of dif-
OMZ 1 76 in "Brain Heart Infu si on " (BHI) without addition ferent foodstuffs in the period 1958 to 1985 (Fig. 1).
of preservatives (control ), and in BHI with different concentra-
tions of preservatives (PS - potassium sorbate, SB - sodium
Among foods in which preservatives are allowed ,
benzoate). Shaded area is sta ndard error (SE) of three para ll el s consumption shows clear increase for meat and
in four experiments. meat products, and for soft drinks, and a slight


374 Leikanger et al.

X 10-6 mmol GLUCOSE/mg PROTEIN than sodium benzoate against both S. sobrinus
3.5 OMZ 176 and S. sanguis 10556. S. sobrinus OMZ
0 CONTROL [Z] 30 mM SB [S2l 60 mM SB 176 was more sensitive than S. sanguis 10556 to
o:D 30 mM PS rz::::l 60 mM PS the two preservatives (Table 1, Figs. 2 and 3).
3 Metabolism - The main metabolite produced by
S. sobrinus OMZ 176 (Fig. 4) and by dental pl aque
(not shown) was lactate. Only small amounts of
2.5
acetate was produced (not shown). Glucose up-
take, and lactate production were not affected by
the preservatives either in S. sobrinus OMZ 176
2
(non-preservative- preexposed cells) (Fig. 4) or in
dental plaque (not shown). S. sobrinus OMZ 176
1.5 grown with potassium sorbate or sodium benzoate
(20 mM) and 1 4C-glucose (0.2 mM) for 14 h showed
a lower glucose uptake and lactate production than
control cells (Fig. 5).

0.5 Plaque pH measurements

Compared to a mouthrinse with glucose (5%), a

o l_=======l single mouthrinsing with potassium sorbate (0.4%
GLUCOSE CONSUMED LACTATE FORMED w/v) and glucose (5%) caused a small, but signi-
Fig. 4. Glucose uptake and product i on of lactate in resting cells ficant, reduction in the pH-drop in plaque as eval-
of S. sobrinus OMZ 176 ex posed to either 30 or 60 mM sod ium uated by the area under the curves (Fig. 6). A more
benzoate (SB) or potassium so rba te (PS) (given as glucose pronounced inhibition was seen at 2% potassium
eq uival ents / mg protein).{
sorbate (w /v, n = 3) (Fig. 7), but the difference in
response between 0.4% and 2% was not significant.

increase for fruit and berries, vegetables and

cheese; consumption of foodstuffs which regularly

do not contain preservatives (wheat flour and bak-
ery goods, fish and fish products) has decreased
(Fig. 1 ).
X 10-6 mmol GLUCOSE/mg PROTEIN
3.5
Furthermore a certain shift in the consumers' [Z] 20 mM SB [D 20 mM PS
choice of foodstuffs seel'l}W to have occurred . The
consumption of foods in which sorbic acid and/ or 3
benzoic acid may legally be added has increased

(for instance juice, jam, pickled vegetables, tomato
ketchup, mayonnaise, etc.). Regulations in Norway 2.5

concerning legal amounts of sorbic acid and benzo-

ic acid in foodstuffs have not changed appreciably
since the sixties (max. 3 g/kg). The response from 2
the food processing industries supported the as-
sumption-s drawn from reports by the central Bu-
reau of Statistics of Norway (13). 1.5

Bacteriologic tests

Growth studies - The MIC-values of potassium

sorbate was 41 mM against S. sobrinus OMZ 176 0.5
and 72 mM against S. sanguis 10556. The corre-
sponding values for sodium benzoate were 58 mM
against S. sobrinus OMZ 176 and 102 mM against
LACTATE FORMED
S. sanguis 10556 (Table 1). Both preservatives re-
Fig. 5. Glucose uptake a nd l actate production in S. sobrinus
duced bacterial growth considerably even at con- OMZ 176 grow n with sod ium benzoate (20 mM - SB) o r
centrations lower than the MIC values (Figs. 2 and potassium so rba te (20 mM - PS) an d 14C-glucose (0.2 mM) for
3). Potassium sorbate was slightly more effective 14 h (given as glucose equiva lents / mg protein).

pH
7.5.------------------------------------.
-e- 5'11. GLUCOSE
-8- 0.4'!1o KS 11'11. GLUCOSE
7 preservatives in food.
6.5
6 method for pH measurements used in the present
5.5
5
4.5 ------.------,-------.------.-----
-5 0 5 10 15 20
TIME MIN
Fig. 6. pH-response (mean of two sites in 5 subjects) in dental
plaque after one mouthrinse with potassium sorbate (KS) (0.4%
w / v + glucose (5% w / v) or 5% w / v glucose.
Discussion
Central Bureau of Statistics of Norway registers
food consumption (13) but does not give any infor-
mation regarding the use of food preservatives.
It is therefore impossible to calculate the exact
consumption or possible changes in consumption
of some food preservatives. However, the data
clearly indicate increased consumption of foods in
which preservatives may legall y be added. More-
over, there has been a strong tendency towards
increased use of processed food in industrialized
countries. It is therefore likely that consumption
of food preservatives through the daily food intake
has increased. On the other hand, the content of
preservatives in various food items seems to have
remained stable.
The microdilution broth method is well accepted
(15) and seems appropriate for testing the antibac-
terial effect of sodium benzoate and potassium sor-
bate against oral streptococci in vitro. We found
that the MIC-values of the preservatives were at
the top limits of the concentrations legally added
to foods. However, bacterial growth was inhibited
substantiall y even at lower concentrations (Figs. 2
and 3). Potassium sorbate was slightly more effec-
tive than sodium benzoate against both bacterial
strains tested. A probable explanation is that sorbic
acid has a higher pK value (pK = 4.76) than benzo-
ic acid (pK = 4.18). Interestingly, S. sobrinus, which
Food preservatives and dental plaque 375
is a member of the mutans streptococci , was more
sensitive to the preservatives than S.sanguis was.
The MIC-value of potassium sorbate against S.
sobrinus (41 mM) is close to the legal amount of
Dental students were chosen for the clinical in-
vestigation because they were expected to cooper-
ate well. A limitation may be, however, their high
dental health awareness, very good oral hygiene,
and low share of mutans streptococci (16). The
study has previously been described by S0DERLING
et al. (17) and evaluated by ScHEIE et al. (18).
We tested the effect of the food preservatives in
combination with glucose because we found that
simu ltaneous supply of these substances would
mimic a real life situation. Potassium sorbate had
a significant effect on pH-drop at concentrations
corresponding to legal amounts of the preserva-
tives (0.4 w / v) (Fig. 6). However, the clinical im-
pact of this inhibition is debatable. At a higher
concentration (2% w / v potassium sorbate), we ob-
served a greater inhibition of pH-drop in plaque
in the three subjects tested (Fig. 7). The preserva-
tives thus have a potential to inhibit the acid pro-
duction in dental plaque, but the question is wheth-
er such concentration will ever be reached in dental
plaque under real life conditions during intake of
preservative-containing food.
There are several factors which may modulate
pH
7.5 ,------------
-e- 5'11. GLUCOSE
-8- 2'!1o KS 15'11. GLUCOSE
7
6. 5
6
5.5
5
4.5 ------r------.-------,------.-----
-5 0 5 10 15 20
TIME MIN
Fig. 7. pH- response (mean of two sites in 3 s ubjects) in d enta l
pl aque after one mouthrinse with pota ssium so rbate (KS) (2%
w / v +glucose (5% w / v) or 5% w / v glucose.

376 Leikanger et al.
potential effects of the preservatives in the oral
cavity. Release of active molecules from the food
and their diffusion into the plaque is crucial. It is
possible that dilution in saliva prevents any effect.
On the other hand, it is also possible that undissoci-
ated molecules pass the cell membrane. At intracel-
lular pH levels, the molecule will become undissoci-
ated and thus accumulate as fluoride does (19). In
that way, continuous suppl y of food which con-
tains preservatives could inhibit plaque metabo-
lism.
In conclusion, the consumption of preservatives
appears to have grad ually increased during the last
decades. The results from in vitro testing suggest
that these preservatives may have inhibiting effects
on caries-pathogenic bacteria at concentrations
near legal amounts. In vivo testing of acid produc-
tion in dental plaque showed a significant effect of
potassium sorbate added in legal amounts, but the
clinical importance is debatable. A higher concen-
tration h ad an obvious inhibiting effect on pH-
drop. The observed effects justify further microbio-
logical and clinical testing of these substances in
order to elucidate their potential to affect plaque
metabolism and thereby caries development.
16. AxELSSON P, KR!STOFFERSEN K, KARLSSON R , BRATTHALL
References
I . GLASS RL. The first internation al conference on the declin-
ing prevalence of dental caries. The evidence a nd the impact
on dental education , dental research , and dental practice.
J Dent Res 1982; 61: 1303-83.
2. CARLOS JP. Epidemiologic trends in caries: impact on adults
a nd the aged. In: GuGGENHEIM B, ed. Cariology t oday.
Basel: Karger 1984: 24-32.
3. FEIIR F VON DER. Evidence of decreasing caries prevalence
in Norway. J Dent Res 1982; 61: 1331 -5.
4. BIERUM P, HOLST D, RISE J . Tannhelseutvik/ingen
1973-1983. En rapport fra 1hmdelagsundersek elsen.
Helsed ir. Oslo, 1984.
5. BJERTNESS E, ERIKSEN HM, HANSEN BF. Ca ries prevalence
of 35-year-old Oslo citizens in 1973 a nd 1 984. Communit y
Dent Oral Epidemiol 1986; 14: 277-82.
6. MARTHALER TM. Explanations for cha ngin g patterns of
disease in the western world . In: GuGGENHEIM B, ed. Cariol-
og y toda y. Basel, Karger: 1984; 1 3-23.
7. JACKSO N D. Has the decline of dental caries in English
children made water Ouo ridat i o n both unnecessary and
unecon omic? Br Dent J 1 987: 162: 1 70-3.
8. BOW E N WH. Impact on research. In: GUGGENHEIM B, ed.
Ca riology today. Basel, Karger: 1984; 49-55.
9. BTRKELA N D JM, LOKEN P. 0ken de Ouorprofy l akse - mindre
karies i Norge. Tidskr Nor Lcegeforen 1978; 98: 11 55-6.
10. MARTHALER TM. Changes in the prevalence of dental car-
ies: How much can be attributed to changes in diet? Caries
Res 1990; 24 (Suppl 1): 3-15.
II. KLocK B, KRASSE B. Caries s ta tus and microbia l conditions
in children in 1 973 and 1 984. Scand J Dent Res 1987; 95:
1 3-7.
1 2. LUECK E. Antimicrobial food additives; c/raracteristics-uses-
effects. Springer-Verlag, New York, 1980.
1 3. THE CENTRAL BUREAU OF STATISTICS OF NORWAY. Norges
offentlige stat istikk. Forbruksundersokelsen; c:'ir 1958, 1976,
1979, 1982 og 1985 .
14. NJERINGSMIDDELLOVEN AV 1 9. MA l 1 933. Bestemmelser om
tilsetningsstoff til nceringsmidle1: Sosialdepartmentet,
Helsedir. 1 983.
1 5. JONES RN , BARRY AL, GAVAN TL, WASHINGTO N II JA.
Susceptibility tests: microdilution a nd macrodi lution broth
procedures. In: LENETTE EH , ed. Manual of clinical microbi-
ology. 4th ed., Washington D.C.: American Society for
Microbiology, 1985; 972-7.
D. A 30 month lon gitudin a l study of th e effects of some ora l
hygi en e meas ures on st reptococcus muta ns and approximal
dental caries. J Dent Res 1 987; 66: 761 -6.
1 7. Si:iDERLING E, MAKI NEN KK , CHEN CY, PAPE HR JR,
LOESCHE W JR, MAKI NEN P-L. Effect of sorbitol , xy litol
and xyl itol / sorbi tol chewing g ums on dental plaque. Caries
Res 1 989; 23: 378-84.
1 8. SCHEIE A AA, FEJERSKOY 0, LfNGSTR0M P, BIRKHED D,
MANJ F. Use of palladium touch microelectrodes under
field conditions for in vivo assessment of dental plaque pH
in children. Caries Res 1992: 26: 44-52.
1 9. HAMILTON I , BOWDEN G. Effect of fluoride on ora l micro-
organisms. In: EKSTRAND J, FEJERSKOV 0, SiLVERSTONE
LM, eds. Fluoride in dentistr y. Copenh agen: Munksgaard:
1988; 77-103.


This document is a scanned copy of a printed document. No warranty is given about

the accuracy of the copy. Users should refer to the original published version of the
material.