The effect of dentifrice quantity and toothbrushing behaviour on oral delivery and retention of fluoride in vivo

Abstract

While toothpaste F⁻ concentration and rinsing regimen have well-characterised impacts on fluoride’s effectiveness, other aspects of brushing regimen have much less well-established effects, in particular, dentifrice quantity and brushing duration. An in vivo study (n = 42) of oral fluoride delivery (i.e. oral disposition post-brushing), and retention (i.e. concentration of F⁻ in saliva post-brushing, a known efficacy predictor), was performed to compare effects observed with those of dentifrice F⁻ concentration and rinsing regimen. Subjects brushed with a NaF-silica dentifrice (Aquafresh Advanced, 1,150 ppm F⁻) or a control dentifrice (250 ppm F⁻, same base), for 45, 60, 120 or 180 seconds with 0.5 or 1.5 g dentifrice, and rinsed with 15 ml water once or three times in a cross-over design. The F⁻ concentration was measured in post-brushing expectorate, rinse and toothbrush washing samples, and in saliva between 5–120 minutes after brushing. Using 1.5 g versus 0.5 g dentifrice increased F⁻ in all samples: oral retention of F⁻ was almost doubled by this increase. Increasing duration of brushing had more complex effects. The amount of F⁻ in the expectorate increased but decreased in both rinse and toothbrush washing samples. Oral F⁻ retention increased, but only in the period 30–120 minutes after brushing. Over the ranges investigated, the order of importance on oral F⁻ retention was: dentifrice F⁻ concentration > quantity > rinsing regimen > brushing duration. Hence, increasing dentifrice quantity and, to a lesser extent, the duration of brushing, can elevate oral fluoride post-brushing. Evidence is accumulating that the importance of these variables to fluoride efficacy may have been underestimated.

INTRODUCTION

Toothbrushing with fluoride toothpaste is the primary method by which individuals reduce their risk of dental caries¹. Fluoride reduces caries incidence by: (1) binding directly to the hydroxyapatite crystals at the enamel surface, forming a fluoridated layer that is less prone to demineralisation during exposure to plaque acid; and (2) promoting remineralisation of partly demineralised enamel when present in fluid overlying the lesion, even at very low concentration2., 3.. Fluoride dentifrices containing 1,000–1,500 ppm F⁻ have been shown convincingly to be effective in reducing the incidence of caries in adolescents, giving a typical 25% reduction in caries in this age group¹. It is likely that this effect builds further over the years with continued use⁴.

Fluoride can also have negative effects. Excessive exposure to fluoride as a young child from over-ingestion of dentifrices can result in fluorosis of the teeth, which is normally visible as white mottling on the permanent dentition but may appear as brown mottling in more severe cases. Young children are vulnerable because their permanent dentition is developing under the gum surface at this time, and fluoride exposure can interfere with the developmental process. The risk of fluorosis is exacerbated by young children’s relatively poor dexterity and understanding, with the result that much of the dentifrice applied is swallowed rather than expectorated after brushing. In 2- to 3 year-olds, when children are most vulnerable to fluorosis⁵, 59–65% of the fluoride is ingested⁶. This subject has been well-reviewed7., 8..

From a caries prevention perspective, therefore, the aim of toothbrushing is to get high enough exposure to topical fluoride for long enough for it to protect against caries effectively, without an unacceptable risk of fluorosis⁹.

The need to control the risk of fluorosis has resulted in regulation of dentifrice fluoride concentrations in countries around the world, with 1,500 μg F⁻ per g of dentifrice the maximum any country allows for general sale. These concerns have also resulted in dentifrice pack label requirements advising that children age 6 years and under should ‘use a pea-sized amount’ for twice-daily toothbrushing. In one study of children for whom toothbrushing was the only source of fluoride, those who claimed to always brush with a pea-sized amount of dentifrice did not experience fluorosis¹⁰. However, there is no clear description of how much a pea-sized amount actually is.

The European Union Scientific Committee On Cosmetic Products and Non-Food Products Intended For Consumers¹¹ recommends a pea-sized amount should be taken as 0.25 g – less than 0.2 ml of a conventional dentifrice formulation – and this amount was also used in a previous study in the USA of the effect of quantity of paste on intraoral fluoride concentrations in children, but no reference for choice of that amount was provided¹².

While much information exists on the toxicological effects of fluoride and the level of exposure likely to produce harmful effects⁷, there is much less information on the relationship between quantity of dentifrice used and incidence of caries, and the data that do exist appear contradictory. Three separate clinical studies that recorded subjects’ use of dentifrice found no evidence for a link between quantity used and incidence of caries; in retrospective analyses13., 14., 15., although dentifrice quantity was not controlled, nor measured on more than one occasion, in any of the clinical studies. In contrast, studies of the effect of dentifrice quantity on fluoride delivery and remineralisation potential, in which quantity was prospectively controlled or carefully monitored, have consistently shown strong positive relationships12., 16., 17..

Varying levels of information exist on other elements of toothbrushing regimen. The effect of toothbrushing duration on fluoride efficacy has received little attention. Evidence suggests that fluoride penetration through plaque¹⁸ and uptake to demineralised enamel¹⁹ are time-dependent on the scale of normal brushing. Consistent with these results, a single-use study demonstrated that enamel remineralisation in situ is influenced by the duration of brushing¹⁷.

In contrast, F⁻ concentration has consistently been linked to the anti-caries efficacy of a dentifrice²⁰. The degree of water rinsing after brushing has also been clearly linked to the anti-caries effectiveness of a dentifrice: four studies have reported that those who rinsed more thoroughly had greater caries experience13., 21., 22., 15.. In isolated contrast, however, the single published prospective study of the effect of rinsing behaviour after brushing on the incidence of caries, performed in a school environment without control of home brushing habits, failed to find a significant link²³.

The aim of the present study was therefore to investigate the effect of dentifrice quantity and brushing duration on intraoral fluoride concentrations during and after brushing, across the normal range of use. The importance of these variables was estimated by comparing their impact to factors known to influence F⁻ effectiveness, specifically dentifrice F⁻ concentration²⁰ and rinsing behaviour13., 21., 22., 15.. As intraoral fluoride concentrations after brushing have been shown to correlate with incidence of caries24., 25., 26., some inference on the likely importance of these brushing behaviour characteristics to incidence of caries might be made.

The values of these oral hygiene regimen variables were chosen to span the range of common practice from potentially lower-efficacy regimens (small amount of dentifrice, short duration of brushing, multiple rinsing) to potentially higher-efficacy regimens (large amount of dentifrice, long duration of brushing, single rinsing). For dentifrice quantity, while a ‘pea-sized’ amount of dentifrice has been indicated to be 0.25 g (EU)²⁷, this is in practice a very small amount and parents of children in developed markets typically dose between 0.3 and 0.5 g28., 27., 29., 30.. Furthermore, the present study was performed in adults, who have larger oral cavities than young children. Hence, 0.5 g was chosen to represent the lower end of the normal range of product use. A 1.5 g dose, representing approximately a full brush-head of silica dentifrice, was taken as the upper end of the normal range.

Estimates of average brushing duration vary from just over 30 seconds to just over 60 seconds31., 32., 33., 34., 35., 36., 37., 38., 39.. A relatively recent study⁴⁰ determined an average of 46 seconds from a home-use study involving 173 US adults. From this survey, a value of 45 seconds was taken as a consensus estimate of average brushing time. There appears to be a general consensus among oral health-care professional bodies that individuals should spend at least 2–3 minutes carefully brushing their teeth twice a day⁴¹, so a value of 180 seconds was chosen as a higher-end estimate of brushing duration, with intermediate values of 60 seconds and 120 seconds also included.

Two rinsing regimens were chosen in this study: once with 15 ml water and three times with 15 ml water, to approximate regimens previously reported to lead to differences in experience of caries. Two studies compared experience of caries in those who used a cup to rinse (higher risk) with those who used other methods (lower risk)21., 22.. The only caries study that quantified rinsing was that of Sjogren and Birkhed¹⁵, who identified low-incidence and high-incidence caries groups, where those with low incidence rinsed an average of 1.5 times with 7 ml water, and those with high incidence rinsed an average of 3.6 times with 19 ml water.

In a previous, preliminary study, we investigated the effects of brushing duration and dentifrice quantity on oral clearance of F⁻¹⁷. In that study, subjects wore an intraoral appliance to allow determination of effects on enamel blocks, supported in the appliance. No appliance was worn in the present investigation, avoiding this potentially interfering factor. The study design also allowed the effect of these behavioural factors to be compared with those of rinsing and dentifrice F⁻ concentration in the same study.

The hypothesis underpinning this work was that dentifrice quantity and duration of brushing are important factors determining intraoral fluoride concentrations during and after brushing, and that these elements of brushing regimen may affect the caries benefits from fluoride dentifrices. The importance of these elements should be balanced against their potential to influence the risk of fluorosis. Results obtained in this adult population should have application from early childhood to old age⁴².

MATERIALS AND METHODS

This was a single-centre, randomised, examiner and laboratory analyst blinded, nine-treatment, nine-period crossover study in normal healthy adult subjects.

Treatments

The study assessed the effect of four target variables on oral delivery and retention of fluoride after toothbrushing:

• •Amount of dentifrice (0.5 g and 1.5 g)
• •Duration of brushing (45 seconds, 60 seconds, 120 seconds and 180 seconds when brushing with 1.5 g dentifrice; and 45 seconds, 60 seconds and 120 seconds when brushing with 0.5 g dentifrice)
• •Concentration of F⁻ in the dentifrice (1,150 ppm F⁻ and 250 ppm F)
• •Rinsing regimen (1 × 15 ml or 3 × 15 ml, for 10 seconds per rinse).

This was achieved by subjects undertaking nine separate brushing regimens, consisting of combinations of these target variables, as set out in Table 1.

Table 1.

Study treatment regimens. Rinses were performed for 10 seconds in all cases

Quantity of dentifrice used, g	Duration of brushing , seconds	Rinsing procedure, ml	F− concentration in dentifrice, ppm
0.5	45	1 × 15	1150
1.5	45	1 × 15	1150
0.5	60	1 × 15	1150
1.5	60	1 × 15	1150
0.5	120	1 × 15	1150
1.5	120	1 × 15	1150
1.5	180	1 × 15	1150
1.5	60	3 × 15	1150
1.5	60	1 × 15	250

Before each test occasion, subjects completed a ‘washout’ period by following their usual dental hygiene practices for 4 days, then using only a fluoride-free dentifrice for 2 days leading up to the visit. On each test day, subjects undertook one of the brushing treatment regimens at the study site under supervision of a trained study technician in randomised order.

Subjects brushed the occlusal surfaces of their back teeth for 10–15 seconds with the specified quantity of the specified dentifrice, moving continuously from quadrant to quadrant until the dentifrice was well dispersed. For the remaining specified brushing time, subjects brushed all surfaces of their teeth covering all quadrants in their usual brushing manner. They then expectorated, and rinsed for the specified number of times. Saliva samples were collected over the following 2 hours.

Sample collection

After brushing, subjects immediately expectorated as much of the dentifrice slurry as possible into a collection container. Rinsing was performed immediately, with rinse samples being expectorated into a separate container. For the extended rinse treatment regimen, subjects rinsed immediately twice more. All three rinse samples were combined. The toothbrush was placed into 10 ml of deionised water and allowed to soak for at least 1 hour.

Saliva samples of target approximately 5 ml volume were collected over a 5-minute period. The first sample was collected immediately before brushing, then further samples at 5, 15, 30, 60 and 120 minutes after brushing.

Test products

The study test treatments were: (1) Aquafresh Advanced dentifrice (GlaxoSmithKline Consumer Healthcare, Clifton, NJ, USA) containing 1,150 ppm F^–, as NaF; (2) an identical dentifrice formulation, except that it contained 250 ppm F⁻ as NaF. The 1,150 ppm F⁻ formulation was used via eight combinations of quantity, brushing duration and rinsing regimen (see Table 1). The 250 ppm F⁻ formulation was used only in ‘standard’ conditions: 1.5 g for 60 seconds with one 15 ml rinse.

Fluoride analysis of brushing samples

The fluoride concentration in saliva samples and of the post-brushing expectorates, rinses and toothbrush washing samples were determined by the method of Taves⁴³, as modified by Martinez–Mier⁴⁴.

Safety assessment

Safety assessments (oral soft tissue examination) and adverse events (AEs) were reported to assess the subjects’ tolerance of the study procedures and treatment regimens.

Measurements and statistical analysis methods

The effect of the variables listed above was assessed on the following measures:

• •The amount of F⁻ in the expectorate, i.e. (expectorate weight) × (F⁻ concentration in sample)
• •The amount of F⁻ in the rinse sample, i.e. (rinse sample weight) × (F⁻ concentration in sample). For the extended rinsing regimen, the samples were pooled and a single overall weight and F concentration were determined
• •The amount of F⁻ remaining on the brush, i.e. (toothbrush washing weight) × (F⁻ concentration in sample)
• •The net amount of F⁻ delivered to the oral cavity during brushing, given by the equation: a

  $\begin{array}{cc}{F}_{\mathrm{delivered}}^{-}\hfill & =F_{\mathrm{dentifrice}}^{-}-(F_{\mathrm{expectorate}}^{-}+F_{\mathrm{rinse}}^{-}\hfill \\ & +F_{\mathrm{toothbrush}\mathrm{washings}}^{-})\hfill \end{array}$

• •The total fluoride retained in the oral cavity after brushing and rinsing, estimated as the area under the complete measured saliva clearance curve (i.e. AUC_{5–120 min})
• •The total fluoride bound in the oral cavity after brushing and rinsing, estimated as the area under the measured saliva clearance curve after 30 minutes (i.e. AUC_{30–120 min}). This allows a 30-minute period for any fluoride retained but not bound in the oral cavity to be rinsed out by normal saliva flow45., 46.. The difference between the ‘bound F⁻’ and ‘total F⁻’ pools is denoted as the ‘unbound F⁻’ pool.

The variables AUC_{5–120 min} and AUC_{30–120 min} were calculated using a trapezoidal rule. That is, the area under the salivary F⁻ concentration versus time plot was calculated as the sum of trapezia drawn between the F⁻ concentration values at adjacent sampling times and their intercepts on the x-axis. To account for non-Gaussian distribution of data, AUC data were log₁₀-transformed before statistical analysis. Study treatment regimens were compared using a mixed-model analysis of variance (anova). The model included a random effect for subject and fixed effects for study period and treatment regimen. The analysis of data involved estimating least-square means for the study measurements. These were used to compare the effects of the different treatment regimens on the study measures. Selected treatment comparisons were performed using two-sided significance tests.

The overall P-value for the treatment effect was also determined. No adjustment for multiplicity was made as a primary comparison was predefined. While the focus of this manuscript is on the effects of dentifrice quantity on the total fluoride retained in the oral cavity (i.e. AUC_{5–120 min}), the primary comparison identified in the protocol was the comparison of the logAUC_{30–120 min} values for 45 seconds brushing versus 120 seconds brushing, with 1.5 g 1,150 ppm F⁻ dentifrice.

RESULTS

Recruitment and demography

Forty-two subjects, in good general and dental health, between the ages of 18 and 65 years, with a normal saliva flow rate and willing to comply with all study procedures and visit requirements, were screened and randomised. Forty-one subjects were included in the per-protocol population of which 39 completed all study treatments. Partial data for 11 subjects were excluded from the analysis because their saliva or expectorate sample volume was less than 1 ml, or they were allocated to incorrect treatment.

Of the randomised subjects, 13 were male and 29 were female. Subjects were aged 21 to 58 years, with a mean age of 35.9 years (SD 11.95). Thirty-three were white, two were Asian, six were black or African-American and one was multiple-race.

Oral disposition of F⁻ immediately after brushing

Table 2 summarises the results and Table 3 the statistical comparisons for the effect of brushing regimen on the disposition of F⁻ after brushing.

Table 2.

Effect of brushing regimen on F⁻ content of samples collected immediately after brushing (unadjusted mean ± SE)

Treatment	n	Expectorate, μg	Rinse, μg	Toothbrush washings, μg	Amount of F− delivered to oral cavity, μg
45 seconds 0.5 g,1 × 15 (1,150 ppm F)	37	244 ± 15.2	104 ± 6.2	181 ± 14.7	134 ± 12.2
45 seconds 1.5 g, 1 × 15 ml (1,150 ppm F)	35	701 ± 49.8	253 ± 14.8	350 ± 16.7	612 ± 48.1
60 seconds 0.5 g, 1 × 15 ml (1,150 ppm F)	38	266 ± 12.4	97 ± 3.6	149 ± 9.1	118 ± 12.9
60 seconds 1.5 g, 1 × 15 ml (1,150 ppm F)	36	801 ± 45.4	238 ± 12.1	337 ± 24.4	531 ± 42.3
120 seconds 0.5 g, 1 × 15 ml (1,150 ppm F)	39	325 ± 12.4	80 ± 4.2	104 ± 7.2	127 ± 12.9
120 seconds 1.5 g, 1 × 15 ml (1,150 ppm F)	40	1,031 ± 46.9	173 ± 12.1	195 ± 16.4	483 ± 40.7
180 seconds 1.5 g, 1 × 15 ml (1,150 ppm F)	40	1,183 ± 41.7	114 ± 7.4	135 ± 12.0	481 ± 40.5
60 seconds 1.5 g, 3 × 15 ml (1,150 ppm F)	38	781 ± 51.1	372 ± 22.2	289 ± 19.6	505 ± 51.2
60 seconds1.5 g, 1 × 15 ml (250 ppm F)	35	156 ± 10.6	72 ± 3.7	90 ± 5.8	90 ± 12.1

The amount of F⁻ delivered to the oral cavity was calculated by subtracting the amount in the expectorate, rinse and toothbrush washings from the theoretical dose. The theoretical doses were: 1725 μg F⁻ for 1.5 g of 1,150 ppm F⁻, 575 μg F⁻ for 0.5 g of 1,150 ppm F⁻ and 375 μg F⁻ for 1.5 g of 250 ppm F⁻ dentifrice.

Table 3.

Treatment comparisons of the effect of dentifrice quantity, duration of brushing, rinsing regimen and concentration of F⁻ in the dentifrice on expectorate, rinse, toothbrush washings, and the amount of fluoride delivered to the oral cavity (using adjusted means). The difference is first-named treatment minus second-named treatment, such that a positive difference favours the first-named treatment

Comparison	Expectorate		Rinse		Toothbrush washings		Amount of F− delivered to oral cavity
	Difference, μg (95% CI*)	P-value	Difference, μg (95% CI*)	P-value	Difference, μg (95% CI*)	P-value	Difference, μg (95% CI*)	P-value
120 seconds versus 45 seconds (0.5 g 1 × 15 ml, 1,150 ppm F)	85.49 (12.9, 158.1)	0.0211	−25.94 (−51.5, −0.4)	0.0465	−86.33 (−117.7, −54.9)	<0.0001	−3.09 (−76.9, 70.7)	0.9344
120 seconds versus 45 seconds (1.5 g 1 × 15 ml, 1,150 ppm F)	324.09 (250.9, 397.3)	<0.0001	−80.97 (−106.7, −55.3)	<0.0001	−156.64 (−187.8, −125.5)	<0.0001	−112.24 (−185.1, −39.4)	0.0026
1.5 g versus 0.5 g (45 seconds 1 × 15 ml, 1,150 ppm F)	479.61 (404.8, 554.4)	<0.0001	149.37 (123.4, 175.3)	<0.0001	163.66 (132.2, 195.1)	<0.0001	460.58 (385.7, 535.5)	<0.0001
1.5 g versus 0.5 g (60 seconds 1 × 15 ml, 1,150 ppm F)	548.73 (475.1, 622.4)	<0.0001	142.32 (116.6, 168.1)	<0.0001	184.94 (153.7, 216.2)	<0.0001	402.40 (329.2, 475.6)	<0.0001
1.5 g versus 0.5 g (120 seconds 1 × 15 ml, 1,150 ppm F)	718.22 (647.2, 789.3)	<0.0001	94.34 (69.0, 119.7)	<0.0001	93.36 (62.1, 124.6)	<0.0001	351.43 (279.6, 423.3)	<0.0001
1.5 g versus 0.5 g (45–120 seconds, 1,150 ppm F)	582.19 (539.8, 624.6)	<0.0001	128.68 (113.8, 143.5)	<0.0001	147.32 (129.2, 165.4)	<0.0001	404.81 (362.4, 447.3)	<0.0001
1,150 ppm F versus 250 ppm F (60 seconds 1.5 g, 1 × 15 ml)	667.11 (591.9, 742.3)	<0.0001	166.44 (140.6, 192.3)	<0.0001	245.74 (214.4, 277.1)	<0.0001	415.03 (340.3, 489.8)	<0.0001
1 × 15 ml versus 3 × 15 ml rinses (60 seconds 1.5 g, 1,150 ppm F)	29.20 (−44.6, 103.0)	0.4367	−131.55 (−157.3, −105.8)	<0.0001	47.57 (16.3, 78.8)	0.0030	11.12 (−62.2, 84.4)	0.7656

^*Confidence intervals for the treatment difference between the true means of the two treatments being compared. Significant differences are in bold type.

Increasing the quantity of dentifrice threefold from 0.5 g to 1.5 g significantly increased F⁻ concentrations across expectorate, rinse and toothbrush washing samples, and increased the net amount of F⁻ calculated to be delivered to the oral cavity more than fourfold (Table 2).

Increasing the duration of brushing significantly increased the F⁻ in the expectorate, but reduced F⁻ in the rinse and in the toothbrush washings, for both 0.5 g and 1.5 g doses. When brushing with 1.5 g dentifrice, the overall impact of these changes was to reduce the net amount of F⁻ delivered to the oral cavity as brushing duration increased (P = 0.0026). There was no overall effect when brushing with 0.5 g dentifrice (P = 0.9344).

Increasing rinsing from one to three times increased the F⁻ recovered in the combined rinse sample by more than 50% (P < 0.0001). However, this increased removal of fluoride did not lead to a significant decrease in the net amount of F⁻ calculated to be delivered to the oral cavity (P = 0.7656).

Increasing F⁻ concentration in the dentifrice from 250 ppm F⁻ to 1,150 ppm F⁻ also increased F⁻ concentrations across expectorate, rinse and toothbrush washing samples. The net amount of F⁻ calculated to be delivered to the oral cavity increased fivefold (P < 0.0001).

Effect of brushing regimen on saliva F⁻ clearance

Tables 4and5, and Figure 1., Figure 2., Figure 3. summarise the results and statistical comparisons for the effect of brushing regimen on salivary F⁻ clearance.

Table 4.

Effect of brushing regimen on oral F⁻ retention, as measured by areas under the curve, AUC_{5–120 min} (estimating ‘total F⁻’ retained) and AUC_{30–120 min} (estimating ‘bound F⁻’ retained). Values are unadjusted mean ± standard error

Treatment	AUC5–120 min, μg × min	SE	AUC30–120 min, μg × min	SE
1,150 ppm, 45 seconds 0.5 g, 1 × 15 ml	17.4	0.358	4.9	0.057
1,150 ppm, 45 seconds 1.5 g,1 × 15 ml	32.4	0.306	8.7	0.049
1,150 ppm, 60 seconds 0.5 g,1 × 15 ml	16.6	0.301	4.9	0.051
1,150 ppm, 60 seconds 1.5 g,1 × 15 ml	34.7	0.324	8.9	0.053
1,150 ppm, 120 seconds 0.5 g, 1 × 15 ml	19.5	0.419	6.3	0.066
1,150 ppm, 120 seconds 1.5 g, 1 × 15 ml	33.9	0.292	10.0	0.046
1,150 ppm, 180 seconds 1.5 g, 1 × 15 ml	33.1	0.355	10.0	0.047
1,150 ppm, 60 seconds 1.5 g,3 × 15 ml	20.0	0.285	7.2	0.046
250 ppm, 60 seconds 1.5 g, 1 × 15 ml	9.3	0.318	3.4	0.050

SE, standard error.

Table 5.

Treatment comparisons: effect of dentifrice quantity, brushing duration, rinsing regimen, and concentration of F in the dentifrice on log₁₀ area under the curve (AUC_{30–120 min} and log₁₀AUC_{5–120 min}) using adjusted means. Difference is first-named treatment minus second-named treatment, such that a positive difference favours the first-named treatment

Comparison	Log10AUC5–120 min		Log10AUC30–120 min
	Difference, 95% CI*	P-value	Difference, 95% CI*	P-value
1.5 g versus 0.5 g (45 seconds 1 × 15 ml, 1,150 ppm F)	0.28 (0.21, 0.35)	<0.0001	0.27, (0.20, 0.35)	<0.0001
1.5 g versus 0.5 g (60 seconds 1 × 15 ml, 1,150 ppm F)	0.32 (0.25, 0.39)	<0.0001	0.27, (0.19, 0.35)	<0.0001
1.5 g versus 0.5 g (120 seconds 1 × 15 ml, 1,150 ppm F)	0.26 (0.19, 0.33)	<0.0001	0.22, (0.14, 0.29)	<0.0001
1.5 g versus 0.5 g (45–120 seconds, 1,150 ppm F)	0.29 (0.25, 0.33)	<0.0001	0.25, (0.21, 0.30)	<0.0001
1150 ppm F versus 250 ppm F (60 seconds 1.5 g, 1 × 15 ml)	0.58 (0.51, 0.65)	<0.0001	0.44, (0.36, 0.51)	<0.0001
1 × 15 ml versus 3 × 15 ml rinses (60 seconds 1.5 g, 1,150 ppm F)	0.25 (0.18, 0.32)	<0.0001	0.10, (0.03, 0.18)	0.0076
120 seconds versus 45 seconds (0.5 g 1 × 15 ml, 1,150 ppm F)	0.05 (−0.02, 0.12)	0.1804	0.12, (0.04, 0.19)	0.0026
120 seconds versus 45 seconds (1.5 g 1 × 15 ml, 1,150 ppm F)	0.02 (−0.05, 0.09)	0.5063	0.06, (−0.02, 0.14)	0.1305
Overall brushing duration response (0.5 g and 1.5 g doses)		0.4219		0.0063
Linear duration response: 1.5 g dose (45–180 seconds)		0.6612		0.1926

s, seconds.

^*Confidence intervals for the treatment difference between the true means of the two treatments being compared. Significant differences are given in bold type.

Figure 1.

Mean saliva F⁻ concentration after treatment as a function of rinsing regime, dentifrice F⁻ concentration and dentifrice quantity (60 seconds brushing). Error bars represent between-subject standard error.

Figure 2.

Mean saliva F⁻ concentration after treatment as a function of duration of brushing for 1.5 g dose (1 × 15 ml rinse). Error bars represent between-subject standard error.

Figure 3.

Mean saliva F⁻ concentration after treatment as a function of duration of brushing for 0.5 g dose (1 × 15 ml rinse). Error bars represent between-subject standard error.

Dentifrice quantity had a substantial impact on oral retention of F⁻. Compared with brushing with 0.5 g, brushing with 1.5 g dentifrice significantly increased the ‘total F⁻’ retained by about 90%, as measured by AUC_{5–120 min} (86% at 45 seconds brushing, 109% at 60 seconds brushing and 74% at 120 seconds brushing; AUC data shown in Table 4)_. Increasing dentifrice quantity on ‘bound F⁻’, as measured by AUC_{30–120 min}, gave a quite similar figure of 73% increase (78% at 45 seconds brushing, 82% at 60 seconds brushing and 59% at 120 seconds brushing).

There was little evidence of a relationship between duration of brushing (across the range 45 seconds to 180 seconds) and ‘total F⁻’ retained (P = 0.4219, averaged across dentifrice quantities, Table 5). However, regarding the equivalent calculation for the ‘bound F⁻’ measure, a modest but significant positive relationship with duration of brushing was observed (P = 0.0063, also Table 5). This relationship appeared to be driven by the results for 0.5 g dentifrice, for which increasing the duration of brushing from 45 seconds to 120 seconds increased ‘bound F⁻’ by 28.6% (P = 0.0026). When subjects used 1.5 g dentifrice, this increase in the duration of brushing led to a 14.9% increase in ‘bound F⁻’, which was not significant (P = 0.1804).

Rinsing once rather than three times also increased ‘total F⁻’ retained, by 73.8% in this study. Using a 1,150 ppm F⁻ dentifrice rather than a 250 ppm dentifrice increased ‘total F⁻’ retained 3.7-fold.

DISCUSSION

Effect of quantity of dentifrice on delivery and clearance of F⁻ after brushing

Increasing the quantity of paste used from 0.5 g to 1.5 g increased F⁻ concentrations in all samples taken immediately after brushing (Table 2). The degree of increase was relatively least on toothbrush washings (about twofold, when averaged across brushing durations), which may be explained by there being only a limited amount of dentifrice that can become trapped in the bristles: the more dentifrice is used, the lower the proportion lost in the toothbrush bristles. Interestingly, this appeared to lead to a higher proportion of the fluoride dose being delivered to the oral cavity: for example, for 60 seconds brushing with the 1,150 ppm F⁻ dentifrice, the net amount of F⁻ delivered to the oral cavity was 30.8% (531/1,725 μg) of the dose from 1.5 g dentifrice, but only 20.5% of the dose from 0.5 g dentifrice (118/575 μg; Table 2).

The importance of dose of toothpaste to salivary fluoride concentrations after brushing is indicated by the near-doubling in ‘total F⁻’ retained in the oral cavity (as measured by AUC_{5–120 min}) when the quantity of dentifrice increased from 0.5 g to 1.5 g (ratio = 1.90 when averaged across brushing durations; Table 4). The size of this effect correlates closely with a previous study comparing these dosages in subjects wearing an intraoral appliance¹⁷, and also with a study in which a fourfold increase in the dentifrice dose (from 0.25 g to 1.0 g) in children aged 4–5 years was reported to produce an approximately threefold increase in salivary F⁻ concentrations in the 2 hours after brushing¹². In the present study, the dose effects on ‘bound F⁻’ and ‘total F⁻’ were quite similar, as reflected in the consistent separation of profiles for the 1.5 g and 0.5 g treatments shown in Figure 1.

One might expect a close relationship between the calculation of the net amount of F⁻ delivered to the oral cavity by brushing and the total amount retained (i.e. AUC_{5–120 min}). However, the net amount of F⁻ delivered is calculated as the difference between two larger numbers (the total in the dentifrice applied and the total recovered in post-brushing samples), and is therefore not a highly robust estimate. Furthermore, the AUC only records from 5 minutes after brushing onwards, after a substantial amount of the F⁻ delivered that remains unbound or loosely bound will already have been cleared, presumably due to swallowing.

Effect of the duration of brushing on delivery and clearance of F⁻ after brushing

Duration of brushing had clear effects on oral delivery of F⁻. Combined data for 0.5 g and 1.5 g doses show that increasing the duration of brushing led to the following differences (Table 3):

• •A reduction in F⁻ remaining trapped in the brush
• •An increase in F⁻ removed from the mouth in the expectorate
• •A reduction in F⁻ present in the rinse
• •A reduction in net amount of F⁻ calculated to be delivered to the oral cavity during brushing (though this was only statistically significant for 1.5 g dentifrice).

These effects may be explained by longer brushing providing more time to dissolve and disperse paste trapped in the bristles of the brush. Individuals brushing for only 45 seconds with 1.5 g dentifrice left 20.3% (350/1725 μg) of the toothpaste in the brush, while those brushing for 180 seconds left only 7.8% (114/1725 μg; Table 2). For the ‘minimum application’ 0.5 g dose/45 seconds brushing regimen, 31.5% (181/575 μg) of the F⁻ remained in the brush.

Longer brushing stimulates saliva production, owing to flavour and the physical stimulation of the bristles on oral surfaces. This increased volume will rinse the mouth more thoroughly, which increases the F⁻ content of the expectorate, leaving less to be removed by the rinse. Overall, although longer brushing increases the time for fluoride to bind to oral surfaces, it slightly reduced the net amount of F⁻ calculated to be delivered to the oral cavity, attributable to improved efficiency of rinsing (Tables 2 and 3).

In contrast, the effects of increasing brushing duration on F⁻ retention were much more subtle: there was a modest increase on ‘bound F⁻’ retained in the oral cavity, but no measurable effect on ‘total F^–’ retained. This conclusion is illustrated by the similarity of the salivary clearance profiles in Figures 2 and 3. This result appears to conflict with the findings of the oral appliance study of Zero and co-workers¹⁷, in which a significant linear dose-response was observed across the 30–180 seconds range. In that study, the increase in AUC_5–120 between 45 seconds and 180 seconds brushing with 1.5 g of an 1,150 ppm F was 33.1% (84.93/64.78 μg × min, P < 0.05; Table 2 in that article); in this study it was only 2.3% (33.1/32.4 μg × min; Table 4).

The contrasting impact in this study of the duration of brushing on the ‘total F⁻’ and ‘bound F⁻’ pools implies differential effects on these two measures. If, as argued above, increasing the duration of brushing has the effect of rinsing the oral cavity more thoroughly, then the amount of ‘unbound F⁻’ would be expected to drop. However, the increased exposure time of oral surfaces to high concentrations of F⁻ during longer brushing times might also be expected to increase fluoride binding, and hence increase ‘bound F⁻’. On close examination, the data behind Figure 2 provide some support for this argument: the longest (180 seconds) duration of brushing gave numerically the lowest fluoride concentration in the first (5 minutes) saliva sample yet the highest fluoride concentration in the last (120 minutes) saliva sample. Further examination of the relationship between duration of brushing and the content of oral fluoride reservoirs is needed to draw firm conclusions in this area.

Effect of rinsing regimen on delivery and clearance of F⁻ after brushing

The 56.3% increase in F⁻ in the combined rinse sample obtained from rinsing three times rather than once (372/238 μg; Table 2) did not, somewhat surprisingly, translate to a meaningful difference in the net amount of F⁻ delivered to the oral cavity in this study (P = 0.7656; Table 3). This finding appeared to be linked to a decrease in the amount recovered in the toothbrush washings in the three-rinse treatment, which has no obvious explanation.

The rinsing regimen did have a major impact on oral retention of F⁻ in this study. Rinsing three times instead of once reduced the ‘total F⁻’ retained by 42.5% (34.7/20.0 μg × min, P < 0.0001; Table 4), and the ‘bound F⁻’ by 18.7% (8.9/7.2 μg × min, P = 0.0076; Table 4). While no statistical comparison of the effects of rinsing on ‘bound F⁻’ with those on ‘total F⁻’ was performed, a greater effect of increased rinsing on ‘total F⁻’ can be rationalised: the ‘total F⁻’ measure encompasses ‘unbound F⁻’ as well as ‘bound F⁻’, and the ‘unbound F⁻’ pool should be more vulnerable to the short-term high-dilution effects of increased rinsing than ‘bound F⁻’⁴⁷.

Effect of dentifrice F⁻ concentration on delivery and clearance of F⁻ after brushing

Increasing the F⁻ concentration in the dentifrice from 250 ppm to 1,150 ppm greatly increased oral delivery of F⁻, as shown by an effect on F⁻ content across expectorate, rinse and toothbrush washings, broadly in proportion to the concentration difference in the dentifrice applied. Similar effects were observed on oral retention of F⁻. Figure 1 shows, for the low and high dentifrice F⁻ concentrations, a near-constant ratio in the F⁻ concentration in saliva between the two treatments across the 2-hour post-brushing sampling period.

IMPLICATIONS FOR IMPACT OF BRUSHING REGIMEN ON EFFECTIVENESS OF FLUORIDE DENTIFRICES

It is widely believed that the anti-caries effects of F⁻ are driven primarily by elevations in oral F⁻ concentrations over a period of hours after product use, as F⁻ is slowly released from oral soft-tissue reservoirs47., 2.. Brushing behaviours that maximise concentrations of F⁻ in saliva after brushing should therefore be encouraged, unless they create other issues such as an increase in fluoride toxicity. A specific analysis was performed to determine the order of importance of the treatment variables (over the ranges investigated) in this study on F⁻ concentrations in saliva after brushing, by comparing area under the salivary clearance curves for 60 seconds brushing (Table 6, plots shown in Figure 1). This analysis found that, for ‘total F⁻’ retained, the order was: F⁻ concentration > dentifrice quantity ≥ rinsing regimen. For ‘bound F⁻’, the order was the same except that the effect of quantity of dentifrice became highly statistically significantly more important than the effect of rinsing regimen. Duration of brushing significantly affected F⁻ concentrations only in the 30–120 minutes after brushing.

Table 6.

Comparison of effects of different toothbrushing variables on fluoride retained in the oral cavity after toothbrushing (using adjusted mean treatment values)

Comparison of effects	LogAUC5–120 min (‘total F−’)		LogAUC30–120 min (‘bound F−’)
	Difference, 95% CI*	P-value	Difference, 95% CI*	P-value
Dose of dentifrice versus number of rinses	0.066 (−0.003, 0.136)	0.0591	0.164 (−0.089, 0.240)	<0.0001
Concentration of fluoride versus number of rinses	0.330 (0.261, 0.399)	<0.0001	0.334 (0.259, 0.409)	<0.0001
Concentration of fluoride versus dose of dentifrice	0.264 (0.194, 0.333)	<0.0001	0.169 (0.094, 0.245)	<0.0001

^*Confidence interval for the treatment difference between the true means of the two treatments being compared. Significant differences are given in bold type.

If F⁻ concentration in oral fluids over hours after brushing is the key parameter driving the anti-caries effects of fluoride, then it would be expected that F⁻ concentration in the dentifrice, brushing quantity and rinsing behaviour should all correlate with incidence of caries. However, only F⁻ concentration (raising F⁻ from 250 ppm to 1,150 ppm)²⁰ and rinsing behaviour13., 21., 22., 15. have been related to the incidence of caries. Three of these studies on the effect of rinsing behaviour on caries incidence also examined the effect of the amount of dentifrice used, but no correlation was found13., 14., 15.. It is important to note here that the analysis of rinsing regimens and dentifrice quantities were not matched between studies. Although it is possible that, in the present study, the effect of rinsing may have been underestimated or that of dentifrice quantity overestimated because of the values chosen, the rinse conditions used here are similar to those reported by Sjogren and Birkhed¹⁵ to be clinically relevant to caries incidence, and the 0.5–1.5 g range of dentifrice quantities tested in this study is by no means extreme.

This situation raises the question: Does the oral clearance profile offer any clues why rinsing regimen can exert a clear effect on anti-caries effectiveness of a dentifrice when its impact on overall fluoride retention is relatively modest? Possibly; it may be important that rinsing regimen was found to affect ‘unbound F⁻’ considerably more than ‘bound F⁻’ in this study, as also noted previously⁴⁷. This suggests that the effect of highly elevated F⁻ concentrations in the minutes after brushing may have been underestimated, in comparison with the effect of very low F⁻ concentrations in the hours after brushing.

It is also possible that there is some confounding factor in the caries studies either interfering with the observation of a real relationship with dentifrice quantity, or exaggerating the effect of rinsing. Further investigation is required to understand if such a confounding factor exists.

The data presented in this study provide limited support for an effect of duration of brushing on the effectiveness of F⁻ from toothbrushing, in line with a previous study that found a significant, but again modest, link between brushing duration and both oral retention of F⁻ and degree of in situ remineralisation¹⁷. Duration of brushing is already strongly positively associated with plaque removal⁴⁸, which is in turn strongly associated with reducing the risk of gingival disease⁴⁹. It is this link that drives the recommendation by many dental associations to brush for 2–3 minutes.

In summary, this study substantially strengthens the evidence indicating that quantity of dentifrice used (within the range of common practice) can influence oral F⁻ retention, and so is potentially important to the anti-caries effects of F⁻ dentifrices. A pea-sized amount of toothpaste is appropriately recommended for young children to reduce the risk of enamel fluorosis. However, when the child is able to properly expectorate, our results add to the existing evidence suggesting that oral health-care professionals should consider recommending brushing with a larger amount of toothpaste to reduce caries risk. Increasing the time spent brushing can also increase F⁻ retention and may also provide F⁻ performance benefits, without bringing any increased risk of fluorosis.

Acknowledgements

The authors thank Mairead North for advice on the statistical analysis.

Conflict of interest

Authors Creeth, Bosma and Butler are employed by GlaxoSmithKline Consumer Healthcare. Authors Zero and Mau were funded by GlaxoSmithKline Consumer Healthcare to conduct this study.