knitr::opts_chunk$set(
  fig.width = 12,
  fig.height = 8,
  fig.path = 'Figs/',
  echo = FALSE,
  include= TRUE,  # for the chunks to be displayed
  warning = FALSE,
  message = FALSE,
  out.width = "75%"
)# Use include=TRUE for the chunks to be displayed

Packages

Dataset

Change output variable

EDA

##  ecc2   n   percent
##     0 115 0.4243542
##     1 156 0.5756458
## 
##   0   1 
## 115 156

age

age by ecc

Mark baseline factors

toothbrushing_frequency = Twice daily and more

Regression model

delete the never in toothbrushing frequency

Omit NA in breastfeeding code

Drop levels

Table 1

Characteristic Overall, N = 2591 Boy, N = 1371 Girl, N = 1221
age
0 20 (7.7%) 9 (6.6%) 11 (9.0%)
1 7 (2.7%) 2 (1.5%) 5 (4.1%)
2 44 (17%) 23 (17%) 21 (17%)
3 90 (35%) 41 (30%) 49 (40%)
4 51 (20%) 34 (25%) 17 (14%)
5 42 (16%) 23 (17%) 19 (16%)
6 4 (1.5%) 4 (2.9%) 0 (0%)
7 1 (0.4%) 1 (0.7%) 0 (0%)
ecc2 149 (58%) 82 (60%) 67 (55%)
general_health_01
0 247 (95%) 129 (94%) 118 (97%)
1 12 (4.6%) 8 (5.8%) 4 (3.3%)
parental_brushing
Every night 102 (45%) 53 (43%) 49 (47%)
Occasionally 70 (31%) 38 (31%) 32 (30%)
Seldom or never 55 (24%) 31 (25%) 24 (23%)
Unknown 32 15 17
dental_hygienist_attendance
Regular twice year 7 (3.4%) 2 (1.9%) 5 (5.3%)
>2 43 (21%) 27 (25%) 16 (17%)
1 65 (32%) 40 (37%) 25 (26%)
0 88 (43%) 39 (36%) 49 (52%)
Unknown 56 29 27
dentist_attendance
Regular twice year 11 (5.4%) 3 (2.8%) 8 (8.4%)
>2 39 (19%) 21 (19%) 18 (19%)
1 63 (31%) 36 (33%) 27 (28%)
0 90 (44%) 48 (44%) 42 (44%)
Unknown 56 29 27
toothbrushing_frequency
Twice daily and more 123 (47%) 63 (46%) 60 (49%)
In evenings 94 (36%) 47 (34%) 47 (39%)
In mornings or less than once 34 (13%) 20 (15%) 14 (11%)
Never 8 (3.1%) 7 (5.1%) 1 (0.8%)
toothpaste
Fluoride 116 (45%) 63 (46%) 53 (43%)
Low Fluoride 129 (50%) 68 (50%) 61 (50%)
Fluoride Free 14 (5.4%) 6 (4.4%) 8 (6.6%)
improvement_hygiene 127 (49%) 66 (48%) 61 (50%)
sugar_liquid 174 (67%) 91 (66%) 83 (68%)
sugar_solid 204 (79%) 107 (78%) 97 (80%)
plaque 127 (49%) 75 (55%) 52 (43%)
bottle_feeding 127 (49%) 69 (50%) 58 (48%)
breastfeeding_code
<6 76 (29%) 39 (28%) 37 (30%)
>6 37 (14%) 18 (13%) 19 (16%)
>12 98 (38%) 59 (43%) 39 (32%)
>24 48 (19%) 21 (15%) 27 (22%)
mouth_breathing 45 (19%) 30 (23%) 15 (13%)
Unknown 17 7 10

1 n (%)

Regression table

m1 full

relevel for analysis

Characteristic N OR1 95% CI1 p-value
age 203 0.63 0.45, 0.85 0.005
gender
Boy 108 — —
Girl 95 0.60 0.26, 1.37 0.2
hygiene_combined_foranalysis
Twice daily and more Fluoride Always 28 — —
Twice daily and more Fluoride Recently 74 9.75 1.62, 190 0.039
Twice daily and more Fluoride Free 5 0.52 0.01, 22.6 0.7
In evenings Fluoride Always 4 11.8 0.44, 428 0.13
In evenings Fluoride Recently 22 15.5 2.06, 333 0.022
In evenings Low Fluoride 32 16.5 2.28, 347 0.017
Less than once per day and/or no fluoride 38 11.9 1.64, 249 0.035
parental_brushing
Every night 96 — —
Occasionally 63 1.38 0.57, 3.33 0.5
Seldom or never 44 9.33 2.80, 36.9 <0.001
plaque
No 105 — —
Yes 98 5.40 2.35, 13.1 <0.001
dental_hygienist_attendance
Regular twice year 7 — —
>2 43 0.54 0.05, 4.76 0.6
1 65 0.72 0.07, 6.01 0.8
0 88 0.43 0.04, 3.45 0.4
bottle_feeding
No 105 — —
Yes 98 0.99 0.39, 2.59 >0.9
breastfeeding_code
<6 60 — —
>6 30 0.70 0.17, 2.79 0.6
>12 78 1.71 0.58, 5.15 0.3
>24 35 3.68 0.97, 15.2 0.062
diet_combined_foranalysis
No sugar Always 19 — —
No sugar Recently 46 6.53 1.33, 41.6 0.029
Sweet drinks and/or sweets 138 2.42 0.54, 14.1 0.3
mouth_breathing
No 168 — —
Yes 35 0.72 0.26, 2.00 0.5

1 OR = Odds Ratio, CI = Confidence Interval

Characteristic N OR1 95% CI1 p-value
parental_brushing
Every night 96 — —
Occasionally 63 1.25 0.39, 3.99 0.7
Seldom or never 44 6.25 1.33, 35.1 0.027
dental_hygienist_attendance
Regular twice year 7 — —
>2 43 0.45 0.02, 9.51 0.6
1 65 5.65 0.37, 140 0.2
0 88 2.43 0.14, 62.3 0.6
dentist_attendance
Regular twice year 11 — —
>2 39 0.40 0.01, 16.7 0.7
1 63 0.01 0.00, 0.54 0.055
0 90 0.01 0.00, 0.29 0.027
age 203 0.65 0.43, 0.94 0.032
gender
Boy 108 — —
Girl 95 0.32 0.10, 0.92 0.040
general_health_01
0 193 — —
1 10 11.2 1.00, 152 0.055
toothbrushing_frequency
Twice daily and more 107 — —
In evenings 61 13.1 2.23, 97.1 0.007
In mornings or less than once 27 4.62 0.41, 75.1 0.2
Never 8 0.06 0.00, 3.11 0.14
toothpaste
Fluoride 102 — —
Low Fluoride 91 13.4 1.75, 167 0.022
Fluoride Free 10 0.31 0.01, 14.6 0.6
improvement_hygiene
No 100 — —
Yes 103 13.0 1.54, 188 0.032
sugar_liquid
No 68 — —
Yes 135 1.90 0.53, 7.31 0.3
sugar_solid
No 43 — —
Yes 160 0.41 0.09, 1.70 0.2
plaque
No 105 — —
Yes 98 16.4 4.90, 69.3 <0.001
bottle_feeding
No 105 — —
Yes 98 0.96 0.30, 3.08 >0.9
breastfeeding_code
<6 60 — —
>6 30 1.14 0.13, 8.56 >0.9
>12 78 1.67 0.39, 7.38 0.5
>24 35 6.98 1.26, 43.1 0.030
mouth_breathing
No 168 — —
Yes 35 0.89 0.19, 3.88 0.9
toothbrushing_frequency * improvement_hygiene 203
In evenings * Yes 36 0.15 0.01, 1.33 0.095
In mornings or less than once * Yes 13 13.8 0.30, 1,031 0.2
Never * Yes 6 0.00 0.00, 1,099,462,362,098,545,590,272 >0.9
toothpaste * improvement_hygiene 203
Low Fluoride * Yes 30 0.01 0.00, 0.16 0.002
Fluoride Free * Yes 4 0.00 >0.9

1 OR = Odds Ratio, CI = Confidence Interval

Check model

m1

## We fitted a logistic model (estimated using ML) to predict ecc2 with age, gender, hygiene_combined_foranalysis, parental_brushing, plaque, dental_hygienist_attendance, bottle_feeding, breastfeeding_code, diet_combined_foranalysis and mouth_breathing (formula: ecc2 ~ age + gender + hygiene_combined_foranalysis + parental_brushing + plaque + dental_hygienist_attendance + bottle_feeding + breastfeeding_code + diet_combined_foranalysis + mouth_breathing). The model's explanatory power is substantial (Tjur's R2 = 0.45). The model's intercept, corresponding to age = 0, gender = Boy, hygiene_combined_foranalysis = Twice daily and more Fluoride Always, parental_brushing = Every night, plaque = No, dental_hygienist_attendance = Regular twice year, bottle_feeding = No, breastfeeding_code = <6, diet_combined_foranalysis = No sugar Always and mouth_breathing = No, is at -2.71 (95% CI [-6.70, 0.81], p = 0.144). Within this model:
## 
##   - The effect of age is statistically significant and negative (beta = -0.47, 95% CI [-0.81, -0.16], p = 0.005; Std. beta = -0.66, 95% CI [-1.15, -0.22])
##   - The effect of gender [Girl] is statistically non-significant and negative (beta = -0.51, 95% CI [-1.35, 0.32], p = 0.229; Std. beta = -0.51, 95% CI [-1.35, 0.32])
##   - The effect of hygiene combined foranalysis [Twice daily and more Fluoride Recently] is statistically significant and positive (beta = 2.28, 95% CI [0.48, 5.25], p = 0.039; Std. beta = 2.28, 95% CI [0.48, 5.25])
##   - The effect of hygiene combined foranalysis [Twice daily and more Fluoride Free] is statistically non-significant and negative (beta = -0.65, 95% CI [-4.49, 3.12], p = 0.719; Std. beta = -0.65, 95% CI [-4.49, 3.12])
##   - The effect of hygiene combined foranalysis [In evenings Fluoride Always] is statistically non-significant and positive (beta = 2.47, 95% CI [-0.81, 6.06], p = 0.129; Std. beta = 2.47, 95% CI [-0.81, 6.06])
##   - The effect of hygiene combined foranalysis [In evenings Fluoride Recently] is statistically significant and positive (beta = 2.74, 95% CI [0.72, 5.81], p = 0.022; Std. beta = 2.74, 95% CI [0.72, 5.81])
##   - The effect of hygiene combined foranalysis [In evenings Low Fluoride] is statistically significant and positive (beta = 2.80, 95% CI [0.82, 5.85], p = 0.017; Std. beta = 2.80, 95% CI [0.82, 5.85])
##   - The effect of hygiene combined foranalysis [Less than once per day and/or no fluoride] is statistically significant and positive (beta = 2.48, 95% CI [0.49, 5.52], p = 0.035; Std. beta = 2.48, 95% CI [0.49, 5.52])
##   - The effect of parental brushing [Occasionally] is statistically non-significant and positive (beta = 0.32, 95% CI [-0.57, 1.20], p = 0.475; Std. beta = 0.32, 95% CI [-0.57, 1.20])
##   - The effect of parental brushing [Seldom or never] is statistically significant and positive (beta = 2.23, 95% CI [1.03, 3.61], p < .001; Std. beta = 2.23, 95% CI [1.03, 3.61])
##   - The effect of plaque [Yes] is statistically significant and positive (beta = 1.69, 95% CI [0.86, 2.58], p < .001; Std. beta = 1.69, 95% CI [0.86, 2.58])
##   - The effect of dental hygienist attendance [>2] is statistically non-significant and negative (beta = -0.62, 95% CI [-2.99, 1.56], p = 0.583; Std. beta = -0.62, 95% CI [-2.99, 1.56])
##   - The effect of dental hygienist attendance [1] is statistically non-significant and negative (beta = -0.33, 95% CI [-2.64, 1.79], p = 0.765; Std. beta = -0.33, 95% CI [-2.64, 1.79])
##   - The effect of dental hygienist attendance [0] is statistically non-significant and negative (beta = -0.85, 95% CI [-3.16, 1.24], p = 0.435; Std. beta = -0.85, 95% CI [-3.16, 1.24])
##   - The effect of bottle feeding [Yes] is statistically non-significant and negative (beta = -7.73e-03, 95% CI [-0.95, 0.95], p = 0.987; Std. beta = -7.73e-03, 95% CI [-0.95, 0.95])
##   - The effect of breastfeeding code [>6] is statistically non-significant and negative (beta = -0.35, 95% CI [-1.79, 1.03], p = 0.621; Std. beta = -0.35, 95% CI [-1.79, 1.03])
##   - The effect of breastfeeding code [>12] is statistically non-significant and positive (beta = 0.53, 95% CI [-0.54, 1.64], p = 0.334; Std. beta = 0.53, 95% CI [-0.54, 1.64])
##   - The effect of breastfeeding code [>24] is statistically non-significant and positive (beta = 1.30, 95% CI [-0.03, 2.72], p = 0.062; Std. beta = 1.30, 95% CI [-0.03, 2.72])
##   - The effect of diet combined foranalysis [No sugar Recently] is statistically significant and positive (beta = 1.88, 95% CI [0.29, 3.73], p = 0.029; Std. beta = 1.88, 95% CI [0.29, 3.73])
##   - The effect of diet combined foranalysis [Sweet drinks and/or sweets] is statistically non-significant and positive (beta = 0.88, 95% CI [-0.63, 2.65], p = 0.280; Std. beta = 0.88, 95% CI [-0.63, 2.65])
##   - The effect of mouth breathing [Yes] is statistically non-significant and negative (beta = -0.33, 95% CI [-1.36, 0.70], p = 0.527; Std. beta = -0.33, 95% CI [-1.36, 0.70])
## 
## Standardized parameters were obtained by fitting the model on a standardized version of the dataset. 95% Confidence Intervals (CIs) and p-values were computed using