Skip to main content
. 2014 Nov 7;9(11):e111123. doi: 10.1371/journal.pone.0111123

Table 3. Case-control studies included in this systematic review presented according to their quality score.

Study Country (Publication language) Local setting Initial sample (final sample) Gender of the sample Patients' mean age at dental examination (range in years) Tooth erosion index (calibration/Kappa) Type of eating disorder (diagnostic criteria) Statistics (adjusted for confounders) Outcomes (OR; 95% CI) or (p-value) Quality score
Dynesen et al., 2008 [21] Denmark (English) Cases: Psychiatric clinic and University Controls: University 40: 20 cases and 20 controls Female Cases: 23.8±4 (18–33) Controls: 23.1±2 (20–30) Larsen et al. modified (K = 0.64) BN‡‡ (DSM IV‡‡‡‡) Multiple regression (age, salivary flow rate and acidic drinks) TE†† score was significantly higher in BN‡‡ group compared with the control- group (p = 0.019) 9(10)
Ohrn et al., 1999 [35] Sweden (English) Cases: Psychiatric clinic Controls: College of nursing 152: 100 cases and 52 controls (133: 81 cases and 52 controls) Male and female Cases: 25 (17–47) Controls: 24 (19–41) Eccles modified by Lussi et al. (NR) AN, BN‡‡, EDNOS (DSM III-R‡‡‡) t-test, Mann-Whitney test and logistic regression (age, number of years suffering from eating disorders) The eating disorders were associated with TE†† severity (<0.001) The period suffering from binge-eating was associated with TE†† (p<0.01) 9(10)
Järvinen et al., 1991 [28] Finland (English) Cases and controls: Metropolitan Helsinki area 206: 106 cases (with TE) and 100 controls (without TE) Male and female Cases: 33.1 (13–73) Controls: 36.3 (17–83) Eccles and Jenkins (1974) (NR) --- Logistic regression (age, gender) Prevalence of ED: 7% of the patients from the case-group were suffering from AN. In the logistic model the practice of vomiting was associated with TE†† (OR = 31; 95%CI = 3–300) 9(10)
Johansson et al., 2012 [4] Sweden (English) Cases: Eating Disorder Clinic Controls: Public Dental Health Clinic 108: 54 cases and 54 controls Male and female Cases:21.5(10–50) Controls: NR Eccles (1979) modified (training and calibration was performed) AN, BN‡‡, EDNOS (NR) Bivariate tests, conditional logistic regression (age) Eating disorders associated with TE†† (OR:8.5; 95%CI: 2.1 – 34.4) Vomiting/binge eating behaviors associated with TE†† (OR = 5.5; 95%CI = 1.3–22.9) 8(10)
Emodi-Perlman et al., 2008 [22] Israel (English) Cases: Weight and Eating Disorders Center Controls: School of Dental Medicine 136: 86 cases and 50 controls (127: 79 cases and 48 controls) Female Cases: 23.46 ± 3.54 (18–35) Controls: 24.58 ±3.01 (18–36) Johansson et al. (1993) (NR) AN, BN‡‡, EDNOS (NR) Chi-square, ANOVA and Tukey's test (no) Vomiting and non-vomiting groups had higher degree of TE†† than controls (p<0.001) There was no difference between TE†† degree among vomiting and non-vomiting groups (p>0.05) 8(10)
Rytömaa et al., 1998 [20] Finland (English) Cases: University Hospital Controls: Universities Dental Services and Colleges 140: 35 cases and 105 controls Female Cases: 25.3 ± 6.8 Controls: 25.7 ±7.0 Eccles and Jenkins & Järvinen et al. (K = 0.74–0.94) BN‡‡ (DSM III-R‡‡‡) Chi-square and t-test (no) BN‡‡ were associated with TE†† (p<0.01) Prevalence of TE††: 11% among controls; 63% among BN group 8(10)
Robb et al., 1995 [19] England (English) Cases: Psychiatric institutions Controls: dental attenders 244: 122 cases and 122 controls NR NR TWI developed by Smith and Knight (NR) AN (purging and restrictive type) and BN‡‡ (NR) Student's t-test (no) AN (both types) and BN‡‡ were associated with TE†† (p<0.005) 7(10)
Milosevic & Slade, 1989 [32] England (English) Cases and controls: Medical School and School of Dentistry 108: 58 cases and 50 controls Male and female Cases:(16–43) Controls: (15–39) TWI developed by Smith and Knight (NR) AN, BN‡‡ without SIV†††, BN‡‡ with SIV††† (DSM III-R‡‡‡) Chi-square, ANOVA (no) Prevalence of TE††: 6% among controls 33% among AN patients 28% among BN‡‡ without SIV††† patients42% among BN‡‡with SIV††† The EDs were associated with TE†† (p<0.001) 7(10)
Jones & Cleaton-Jones, 1989 [31] South Africa (English) Cases and controls: Private dental office 33: 11 cases and 22 controls Female Cases:29.8±8.4 Controls: 28.9 ± 9 Own criteria (NR) BN‡‡(NR) Chi-square (no) Prevalence of TE††: 7% among controls 69% among BN‡‡ patients BN‡‡ was associated with TE††(p<0.001) 7(10)
Howat et al., 1990 [18] USA (English) Cases: University Eating Disorder Clinic Controls: University (other departments) 20: 10 cases and 10 controls (18: 8 cases and 10 controls) Female Cases:24.6 Controls: 22.2 NR BN‡‡ (DSM III-R‡‡‡) T-test, Fisher's exact test, ANOVA and Pearson Correlation (no) No difference between the presence of TE†† between the groups (p>0.05) 6(10)
Sivolella et al., 2000 [33] Italy (Italian) Cases: Eating Disorder Hospital Controls: NR 26: 14 cases and 12 controls Female Cases: 23.28 ± 4.9 Controls: 22.58 ± 1.8 NR AN, BN‡‡ (DSM IV‡‡‡‡) Fisher's exact test (no) AN and BN‡‡ were associated with TE†† (p = 0.009) Prevalence of TE††: 41.6% among controls 92.8% among BN‡‡ group 5(10)
Greenwood et al., 1988 [30] Ireland (English) Cases: University Hospital patients Controls: University Hospital staff 48: 24 cases and 24 controls Female AN group: (15–24) BN‡‡ group: (19–35) Controls: NR NR AN, BN‡‡ (NR) NR Prevalence of TE††: 9% among AN patients 30% among BN‡‡ patients All patients who SIV††† had TE†† 5(10)
Touyz et al., 1993 [34] Australia (English) NR 45: 30 cases and 15 controls Female AN group: 20.1 ± 8.3 BN‡‡ group: 19.1 ± 3.8 Controls: 22.1 ±3.3 NR AN, BN‡‡ (DSM III-R‡‡‡) Chi-square and t-test (no) AN and BN‡‡ were associated with TE†† (p-value was not reported) 3(10)

AN  =  anorexia nervosa; BN‡‡  =  bulimia nervosa; DSM III- R‡‡‡  =  Diagnostic and Statistical Manual of Mental Disorders, 3rd Edition, Revised; DSM IV‡‡‡‡  =  Diagnostic and Statistical Manual of Mental Disorders, 4th Edition; NR  =  not reported; TE††  =  tooth erosion; SIV†††  =  self-induced vomiting.