2.1

Participants

Participants (n = 600) aged 18 years or older, were recruited between May 2014 and March 2016 following referral by their general dental practitioners (GDP) for erosive tooth wear (cases, n = 300) or general treatment (controls, n = 300) to the restorative clinics at King’s College London Dental Institute. All participants had a minimum of 20 teeth (10 in each jaw). Participants were excluded if they had missing anterior teeth, anterior crowns/bridges or cavitated caries on more than one tooth. A history of eating disorders or gastro-oesophageal reflux, bruxism or prescribed xerostomic/heartburn medication, pregnancy, involvement in other research within the past 30 days or inability to speak or understand the English language also excluded the participant from this study. Bruxism was diagnosed from the clinical appearance, signs and symptoms of attritive tooth wear; those with flattened surfaces without erosive lesions were excluded.

According to a pilot study and a previous study within our group , a minimum sample size of 490 participants (245 in each group) was needed. This calculation assumed the proportion of adults with high dietary acid intake (3+ times/day) was 55% among cases and 40% among controls (expected odds ratio of 2.25), case-control ratio of 1-to-1, 90% statistical power and 95% significance level.

2.2

Data collection

Data collection procedures were identical for cases and controls. The Basic Erosive Wear Examination (BEWE) index graded tooth wear on the buccal, occlusal and palatal/lingual surfaces of each tooth excluding third molars and was used to differentiate the groups. This ordinal scale graded tooth wear from 0 to 3 (0 = no wear, 1 = early surface loss, 2 = surface loss < 50% or specific defect, 3 = surface loss > 50%). Teeth with restorations involving >50% of the tooth, traumatised or carious teeth were excluded from the assessment. Teeth were examined under good lighting in a dental chair but without magnification. A sextant BEWE score was calculated by recording the highest score from any surface in each sextant. The cumulative BEWE score was calculated by summing each sextant BEWE score within the range from 0 to 18 . A trained and calibrated dentist carried out all clinical examinations. Inter-examiner and intra-examiner Kappa scores, calculated on duplicate examinations of 30 patients against a calibrated BEWE expert (DB), were 0.85 and 0.75 respectively.

Erosive tooth wear cases were defined as those with a BEWE score of 12 or higher and at least one score of 3 in a sextant whereas controls were defined as those with a BEWE score of 10 or lower and no score of 3 on any surface of any tooth (clinically classified as no or mild erosive tooth wear). Controls were age-matched on a 1:1 ratio with cases. Frequency matching was used, to yield controls with the same distribution over six age groups (18–25, 26–35, 36–45, 46–55, 56–65 and 66+ years). Cases that could not be matched were excluded from the study.

After recruitment, a trained interviewer questioned participants on potential risk factors for erosive tooth wear using an adapted version of a previously validated questionnaire . Participants were asked about the frequency, timing (with meals or between meals) and duration of consumption of fruits, fruit drinks, carbonated beverages and other acidic drinks , the type of holder (cup, glass, bottle, can), whether they had a habit of retaining drinks in their mouth (sipping, swishing or holding) prior to swallowing, whether they reported to brush within 10 min of consuming something acidic and any associated symptomatology (i.e. sensitive teeth). Total daily frequency of acid intake was estimated as the sum of fruit (apples, citrus, grapes, berries and any other fruit) and acidic drinks (carbonated drinks, fruit drinks, any other acidic drinks e.g. fruit teas, wine). For timing, total daily frequency of acid intake was separated into four indicators, daily frequency of fruit intake with and between meals and daily frequency of acidic drink intake with and between meals. Duration of a single acid intake was divided into three categories (<5 min, 5–10 min and ≥10 min). Questions were standardised and participants were given the opportunity to clarify questions, ensuring comprehension before answering. Each questionnaire lasted between 5 and 10 min.

2.3

Data analysis

All analyses were performed in the Statistical Package for Social Sciences version 22 for Windows (IBM Corporation, Armonk, New York, USA). Cases and controls were compared in terms of their demographic and clinical characteristics, using the Chi-square test for categorical variables and the t -test for continuous measures. Cases and controls were also compared in terms of risk factors (total daily acid intake, fruit intake with meals, fruit intake between meals, acidic drink intake with meals, acidic drink intake between meals, duration of fruit and acidic drink consumption, brushing within 10 min of acid intake and alternate drinking habits) using the Chi-square test. Variables to be included in the logistic regression model were manually selected based upon prior theory and the research hypothesis. The variables included were: the frequency of fruit and acidic drink consumption with meals and between meals, duration of consumption, brushing after acid intake and alternate drinking habits prior to swallowing.

Using the presence or absence of severe erosive wear (cases and controls) as the dependent variable, the unadjusted, sex-and-age-adjusted and fully adjusted associations of the included variables with erosive tooth wear were estimated using unconditional binary logistic regression and reported using odds ratios (OR). As frequency matching does not require the use of conditional logistic regression during analysis , age (in the pre-defined groups), was included as a potential confounder.

2.1

Participants

Participants (n = 600) aged 18 years or older, were recruited between May 2014 and March 2016 following referral by their general dental practitioners (GDP) for erosive tooth wear (cases, n = 300) or general treatment (controls, n = 300) to the restorative clinics at King’s College London Dental Institute. All participants had a minimum of 20 teeth (10 in each jaw). Participants were excluded if they had missing anterior teeth, anterior crowns/bridges or cavitated caries on more than one tooth. A history of eating disorders or gastro-oesophageal reflux, bruxism or prescribed xerostomic/heartburn medication, pregnancy, involvement in other research within the past 30 days or inability to speak or understand the English language also excluded the participant from this study. Bruxism was diagnosed from the clinical appearance, signs and symptoms of attritive tooth wear; those with flattened surfaces without erosive lesions were excluded.

According to a pilot study and a previous study within our group , a minimum sample size of 490 participants (245 in each group) was needed. This calculation assumed the proportion of adults with high dietary acid intake (3+ times/day) was 55% among cases and 40% among controls (expected odds ratio of 2.25), case-control ratio of 1-to-1, 90% statistical power and 95% significance level.

2.2

Data collection

Data collection procedures were identical for cases and controls. The Basic Erosive Wear Examination (BEWE) index graded tooth wear on the buccal, occlusal and palatal/lingual surfaces of each tooth excluding third molars and was used to differentiate the groups. This ordinal scale graded tooth wear from 0 to 3 (0 = no wear, 1 = early surface loss, 2 = surface loss < 50% or specific defect, 3 = surface loss > 50%). Teeth with restorations involving >50% of the tooth, traumatised or carious teeth were excluded from the assessment. Teeth were examined under good lighting in a dental chair but without magnification. A sextant BEWE score was calculated by recording the highest score from any surface in each sextant. The cumulative BEWE score was calculated by summing each sextant BEWE score within the range from 0 to 18 . A trained and calibrated dentist carried out all clinical examinations. Inter-examiner and intra-examiner Kappa scores, calculated on duplicate examinations of 30 patients against a calibrated BEWE expert (DB), were 0.85 and 0.75 respectively.

Erosive tooth wear cases were defined as those with a BEWE score of 12 or higher and at least one score of 3 in a sextant whereas controls were defined as those with a BEWE score of 10 or lower and no score of 3 on any surface of any tooth (clinically classified as no or mild erosive tooth wear). Controls were age-matched on a 1:1 ratio with cases. Frequency matching was used, to yield controls with the same distribution over six age groups (18–25, 26–35, 36–45, 46–55, 56–65 and 66+ years). Cases that could not be matched were excluded from the study.

After recruitment, a trained interviewer questioned participants on potential risk factors for erosive tooth wear using an adapted version of a previously validated questionnaire . Participants were asked about the frequency, timing (with meals or between meals) and duration of consumption of fruits, fruit drinks, carbonated beverages and other acidic drinks , the type of holder (cup, glass, bottle, can), whether they had a habit of retaining drinks in their mouth (sipping, swishing or holding) prior to swallowing, whether they reported to brush within 10 min of consuming something acidic and any associated symptomatology (i.e. sensitive teeth). Total daily frequency of acid intake was estimated as the sum of fruit (apples, citrus, grapes, berries and any other fruit) and acidic drinks (carbonated drinks, fruit drinks, any other acidic drinks e.g. fruit teas, wine). For timing, total daily frequency of acid intake was separated into four indicators, daily frequency of fruit intake with and between meals and daily frequency of acidic drink intake with and between meals. Duration of a single acid intake was divided into three categories (<5 min, 5–10 min and ≥10 min). Questions were standardised and participants were given the opportunity to clarify questions, ensuring comprehension before answering. Each questionnaire lasted between 5 and 10 min.

2.3

Data analysis

All analyses were performed in the Statistical Package for Social Sciences version 22 for Windows (IBM Corporation, Armonk, New York, USA). Cases and controls were compared in terms of their demographic and clinical characteristics, using the Chi-square test for categorical variables and the t -test for continuous measures. Cases and controls were also compared in terms of risk factors (total daily acid intake, fruit intake with meals, fruit intake between meals, acidic drink intake with meals, acidic drink intake between meals, duration of fruit and acidic drink consumption, brushing within 10 min of acid intake and alternate drinking habits) using the Chi-square test. Variables to be included in the logistic regression model were manually selected based upon prior theory and the research hypothesis. The variables included were: the frequency of fruit and acidic drink consumption with meals and between meals, duration of consumption, brushing after acid intake and alternate drinking habits prior to swallowing.

Using the presence or absence of severe erosive wear (cases and controls) as the dependent variable, the unadjusted, sex-and-age-adjusted and fully adjusted associations of the included variables with erosive tooth wear were estimated using unconditional binary logistic regression and reported using odds ratios (OR). As frequency matching does not require the use of conditional logistic regression during analysis , age (in the pre-defined groups), was included as a potential confounder.