Jan 8, 2021. This Markdown refers to the article entitled “Resilience and vulnerability in adolescents with primary headaches: a cross-sectional population-based study”, currently under review in Headache. All data and codes are available at https://osf.io/wmxc9/

If you want to reproduce all results from the beginning, you will have to download the original data (OSF -> Original data), change the file name in read_excel command, and run all chunks with eval = FALSE. Please keep in mind that this data was an excel file in which anyone could find names and other identifications. Therefore, to preserve participants anonymity, I removed all identification variables before uploading it to open science framework website. No other changes were carried out.

If you just want to reproduce the results presented in the manuscript, please download and load the the importable data (CSV or Rdata- line 77) and run the chunks without eval = false. The analystical procedures start at line ~740.

All chunks are named to make the sections clear.

If you have any comments or questions, please reach me out at Thank you

1 Get R data (processed)

Load this file and then go to line 740.

2 Data processing

If someone wants to get the raw data, i.e. the original dataset, please run the following code and all the other chunks in which “eval = false” is present.

3 get raw data

4 Data cleaning

Data cleaning is the first step of data analysis. The function clean_names from janitor package will be used.

How many participants do not have filled cases in mastery

## # A tibble: 20 x 3
##    feature num_missing pct_missing
##    <fct>         <int>       <dbl>
##  1 m1                3     0.00885
##  2 m2                2     0.00590
##  3 m3                9     0.0265 
##  4 m4                7     0.0206 
##  5 m5                6     0.0177 
##  6 m6                4     0.0118 
##  7 m7                6     0.0177 
##  8 m8                4     0.0118 
##  9 m9                4     0.0118 
## 10 m10               4     0.0118 
## 11 m11               5     0.0147 
## 12 m12               3     0.00885
## 13 m13               3     0.00885
## 14 m14               7     0.0206 
## 15 m15               3     0.00885
## 16 m16               5     0.0147 
## 17 m17               1     0.00295
## 18 m18               3     0.00885
## 19 m19               2     0.00590
## 20 m20               3     0.00885
## # A tibble: 9 x 2
##     age sumNA
##   <dbl> <dbl>
## 1    10     0
## 2    11    14
## 3    12    10
## 4    13    14
## 5    14    11
## 6    15     9
## 7    16    14
## 8    17    12
## 9    18     0

5 Mastery

Mastery is formed of all 20 items that, in sequence, give the opportunity to have results of optimism,self-efficacy, and adaptability. Based on the DSM-5 criterion for learning disabilities, if the standardized result of a raw score was less or equal than -1.5 standard deviation (using all group as reference), the participant was allocated in the cattegory of being ‘at risk’.

Following this criterion, 23 (7%) children were classified in ‘at risk’ group.

mastery_prob n prop
0 316 0.93
1 23 0.07

5.1 Optimism

I’m assuming optimis is composed of items 1-4 and 18-20, as exposed in excel spreadsheet. The standardized score was computed based on their sum.

In this sample, 22 (6%) of the children were classified in ‘at risk’ group.

optimism_prob n prop
0 317 0.94
1 22 0.06

5.2 Self-efficacy

I’m assuming Self-efficacy is composed of 10 items (5-14), as exposed in excel spreadsheet. The standardized score was computed based on their sum.

In this sample, 20 (6%) of the children were classified in ‘at risk’ group.

self_prob n prop
0 319 0.94
1 20 0.06

5.3 Adaptability

I’m assuming adaptability is composed of 3 items (15-17), as exposed in excel spreadsheet. This (sub)scale is applied to certain age-interval. The standardized score was computed based on their sum.

In this sample, 12 (4%) of the children were classified in ‘at risk’ group.

adaptability_prob n prop
0 128 0.38
1 12 0.04
NA 199 0.59

The cronbach’s alpha was computed for this subscale just to certify some of its properties

## 
## Reliability analysis   
## Call: alpha(x = .)
## 
##   raw_alpha std.alpha G6(smc) average_r S/N   ase mean   sd median_r
##       0.56      0.56    0.48       0.3 1.3 0.042  2.8 0.81     0.24
## 
##  lower alpha upper     95% confidence boundaries
## 0.48 0.56 0.64 
## 
##  Reliability if an item is dropped:
##     raw_alpha std.alpha G6(smc) average_r  S/N alpha se var.r med.r
## m15      0.63      0.63    0.46      0.46 1.70    0.040    NA  0.46
## m16      0.33      0.33    0.20      0.20 0.49    0.073    NA  0.20
## m17      0.38      0.38    0.24      0.24 0.62    0.067    NA  0.24
## 
##  Item statistics 
##       n raw.r std.r r.cor r.drop mean  sd
## m15 336  0.66  0.66  0.33   0.25  2.8 1.1
## m16 334  0.78  0.78  0.61   0.45  2.7 1.1
## m17 338  0.75  0.76  0.58   0.42  2.9 1.1
## 
## Non missing response frequency for each item
##        0    1    2    3    4 miss
## m15 0.05 0.05 0.30 0.25 0.35 0.01
## m16 0.03 0.09 0.34 0.21 0.33 0.01
## m17 0.03 0.07 0.28 0.23 0.39 0.00

6 Relatedness

Relatedness is composed of 24 items that form 4 subscales.

In this study, 24 (7%) children were classied ‘at risk’.

relatedness_prob n prop
0 315 0.93
1 24 0.07

6.1 Trust

Trust is composed of 7 items (6-10 and 23-24).

In this research, 28 (8%) children were classified ‘at risk’

trust_prob n prop
0 311 0.92
1 28 0.08

6.2 Support

Support is composed of 6 items (5, and 18-22).

In this study, 31 (9%) children were classied in ‘at risk’ group.

support_prob n prop
0 308 0.91
1 31 0.09

6.3 Comfort

The Comfort subscale is composed of 4 items (1-4)

In this study, 23 (7%) children were classified at risk.

comfortt_prob n prop
0 316 0.93
1 23 0.07

6.4 Tolerance

The Tolerance subscale is composed of 7 items(11-17). This scale is not possible for certain age intervals.

In this study, 17 (5%) children were classified at risk.

tolerance_prob n prop
0 265 0.78
1 17 0.05
NA 57 0.17

7 Reactivity

Reactivity is composed of 20 items that form 3 subscales. The higher the score in this scale, the greater the risk of developmental problems.

In this study, 32 (9%) children were classified at risk.

reactivity_prob n prop
0 307 0.91
1 32 0.09

7.1 Sensitivity

Sensitivy subscale is composed of 6 items.

In this study, 27 (8%) children were classified at risk.

sensitivity_prob n prop
0 312 0.92
1 27 0.08

7.2 Recovery

Recovery subscale is composed of 4 items (10-13.)

In this study, 30 (9%) children were classified at risk.

recovery_prob n prop
0 309 0.91
1 30 0.09

7.3 Impairment

Impairment subscale is composed of 10 items (7-9 and 14-20)

In this study, 28 (8%) children were classified at risk.

impairment_prob n prop
0 311 0.92
1 28 0.08

8 Total ressources

Total ressources are formed of the average of the raw scores of Mastery and Relatedness. In this scale, the lower the result, the greater the risk.

The standardized score of Total Ressources was computed. If a participant has 0 as his/her result, it means an average score. Results above than 0 indicates a protective environment and results below 0 indicates a risk factor.

In this study, 28 (8%) children were classified at risk.

resources_prob n prop
0 311 0.92
1 28 0.08

9 Vulnerability

This scale is composed of the difference between Reactivity and Resources. To achieve these results, one needs to subtract the resource T-score from the emotional Reactivity t-score. In this study, we used the raw score.

Positive results reveal risk outcomes. It means the child has more reactivity behaviors than resources to deal with them. On the other hand, negative results means child has more resources than emotional reactivity.

In this study, 18 (5%) children were classified at risk.

vulnerability_prob n prop
0 321 0.95
1 18 0.05

The distribution of the raw results of vulnerability is exposed below.

! Done with raw data

10 Variables adjustments

11 Variables

! Done with everything (last check on october 9, 2020)

12 Abstract

Background: The scarcity of studies on the role of resilience resources (RRs) and vulnerability risk (VR) in children and adolescents with primary headache hampers the development of a risk-resilience model for pediatric headaches. Objective: To examine the extent to which headache frequency and diagnosis are associated with RRs and VR and explore possible predictors of low RRs and high VR in a cross-sectional population-based study in adolescents. Methods: This is a cross-sectional population study conducted in a small city in Brazil (Delfinópolis). Consents and analyzable data were obtained from 339/378 adolescents (89.7%). RRs and VR were assessed using the validated Brazilian version of the Resiliency Scales for Children and Adolescents (RSCA), completed by the adolescents. Parents filled a structured questionnaire assessing sociodemographic and headache characteristics, as well as the Brazilian-validated version of the Strengths and Difficulties Questionnaire (SDQ) added to the impact supplement to evaluate the adolescent’s psychosocial adjustment skills. Teachers completed a structured questionnaire about the students’ school performance. Results: A higher frequency of headache was associated with lower RRs (F3,335 = 2.99, p = 0.031) and higher VR (F3,335 = 4.05, p = 0.007). Headache diagnosis did not significantly influence the risk of having lower RRs or higher VR. In the exploratory analysis, females (OR 3.07 (95%CI: 1.16 to 9.34)) and individuals with psychosocial adjustment problems (OR 7.53 (95%CI: 2.51 to 22.36)) were predictors of low RRs, and prenatal exposure to tobacco (OR 5.61 (95%CI: 1.57 to 20.86)) was a predictor of high VR in adolescents with primary headache. Conclusions: The risk of low RRs and high VR was associated with a higher headache frequency, but not with headache diagnosis. These findings may contribute to the development of a risk-resilience model of headaches in pediatric population and help identify novel targets and develop effective resources for successful interventions.

13 Missing data and outliers

## # A tibble: 5 x 3
##   feature             num_missing pct_missing
##   <fct>                     <int>       <dbl>
## 1 mastery_total                 0           0
## 2 relatedness_total             0           0
## 3 reactivity_total              0           0
## 4 ressources_total              0           0
## 5 vulnerability_total           0           0

14 Methods

14.1 Participants

14.2 Table 1

## Frequencies  
## base_uso$age_group  
## Type: Factor  
## 
##                  Freq   % Valid   % Valid Cum.   % Total   % Total Cum.
## -------------- ------ --------- -------------- --------- --------------
##        [10,12]     99     29.20          29.20     29.20          29.20
##        (12,15]    160     47.20          76.40     47.20          76.40
##       (15,Inf]     80     23.60         100.00     23.60         100.00
##           <NA>      0                               0.00         100.00
##          Total    339    100.00         100.00    100.00         100.00
## 
## base_uso$sex  
## Type: Factor  
## 
##                Freq   % Valid   % Valid Cum.   % Total   % Total Cum.
## ------------ ------ --------- -------------- --------- --------------
##       Female    181     53.39          53.39     53.39          53.39
##         Male    158     46.61         100.00     46.61         100.00
##         <NA>      0                               0.00         100.00
##        Total    339    100.00         100.00    100.00         100.00
## 
## base_uso$race  
## Type: Factor  
## 
##               Freq   % Valid   % Valid Cum.   % Total   % Total Cum.
## ----------- ------ --------- -------------- --------- --------------
##       white    239     72.64          72.64     70.50          70.50
##       Other     90     27.36         100.00     26.55          97.05
##        <NA>     10                               2.95         100.00
##       Total    339    100.00         100.00    100.00         100.00
## 
## base_uso$classe_economica  
## Type: Character  
## 
##               Freq   % Valid   % Valid Cum.   % Total   % Total Cum.
## ----------- ------ --------- -------------- --------- --------------
##          AB     93     27.43          27.43     27.43          27.43
##           C    203     59.88          87.32     59.88          87.32
##          DE     43     12.68         100.00     12.68         100.00
##        <NA>      0                               0.00         100.00
##       Total    339    100.00         100.00    100.00         100.00

14.3 Instruments / Measures

This research relied on the results obtained by several scales developed to access and measure different domains of children’s resilience. All scales had its psychometric properties previously studied and this research accessed the realibility of the data gathered through the Cronbach’s alpha coefficient. The following information presents the measures.

The Sense of Mastery scale (α = .87) is composed of 20 items and aims to access three different domains, i.e., Optimism, SeItslf-Efficacy, and Adaptability. Its manual guide defines optimis as a positive attitude(s) about the world/life in general and about individual’s life specifically, currenly, and in the future. Self-efficacy is conceived as the sense that one can master his or her environment. Adaptability is the ability to learn from one’s mistakes and to accept feedback from others. In the present research, results were roughly normally distributed (Mean = 50, Median = 50).

The Sense of Relatedness scale (α = .90) is composed of 24 items and access fours domains, such as sense of trust, perceived access to comfort with others, and tolerance. The Sense of Turst is conceptualized as the extent to which others are perceived as reliable and the extent to which one can be authentic in relationship with others. Results were slightly left skewed (Mean = 63, Median = 65).

The Emotional Reactivity Scale (α = .90) is composed of 20 items and access two domains, that is sensitivity and recovery. In addition to that, this scale also includes a measure of the participant’s overall resources (named as Resource Index Score) and Vulnerability (named as Vulnerability Index Score. The Resource Index Score is computed by the standardized average of the sense of Mastery T score and Sense of Relatedness T score. Similarly, the Vulnerability Index Score is derived by the standardized difference between the Resource Index and the Emotional Reactivity T score. Results were slightly right skewed (Mean = 31, Median = 29).

Attention-deficit/hyperactivity disorder (ADHD) and mental health were evaluated to assess their role as mediators of RRs, VR, and primary headaches. ADHD symptoms were assessed using the Brazilian-validated version of the Multimodality Treatment Study–Swanson, Nolan, and Pelham–version IV (MTA-SNAP-IV) scale,22 in accordance with the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5).23. The Cronbach’s alpha of this scale was 0.91  (95% CI = [0.9 0.92]).
## 
## Reliability analysis   
## Call: alpha(x = .)
## 
##   raw_alpha std.alpha G6(smc) average_r S/N   ase mean   sd median_r
##       0.91      0.91    0.93      0.37  10 0.007 0.63 0.52     0.35
## 
##  lower alpha upper     95% confidence boundaries
## 0.9 0.91 0.92 
## 
##  Reliability if an item is dropped:
##        raw_alpha std.alpha G6(smc) average_r  S/N alpha se var.r med.r
## snap1       0.91      0.91    0.93      0.37 10.2   0.0072 0.012  0.36
## snap2       0.91      0.91    0.92      0.36  9.7   0.0075 0.012  0.35
## snap3       0.91      0.91    0.92      0.37 10.0   0.0073 0.012  0.35
## snap4       0.91      0.91    0.92      0.37  9.8   0.0074 0.012  0.35
## snap5       0.90      0.91    0.92      0.36  9.6   0.0075 0.011  0.34
## snap6       0.90      0.91    0.92      0.36  9.6   0.0076 0.011  0.34
## snap7       0.91      0.91    0.92      0.37  9.8   0.0074 0.013  0.35
## snap8       0.90      0.90    0.92      0.36  9.4   0.0077 0.011  0.34
## snap9       0.90      0.91    0.92      0.36  9.7   0.0075 0.011  0.35
## snap10      0.91      0.91    0.92      0.37 10.1   0.0072 0.013  0.36
## snap11      0.91      0.91    0.92      0.36  9.8   0.0074 0.013  0.35
## snap12      0.91      0.91    0.92      0.37 10.0   0.0073 0.012  0.36
## snap13      0.91      0.91    0.93      0.38 10.2   0.0072 0.012  0.36
## snap14      0.91      0.91    0.92      0.37  9.9   0.0073 0.012  0.35
## snap15      0.91      0.91    0.92      0.38 10.2   0.0071 0.011  0.36
## snap16      0.90      0.91    0.92      0.36  9.7   0.0075 0.012  0.34
## snap17      0.90      0.91    0.92      0.36  9.7   0.0075 0.012  0.35
## snap18      0.90      0.91    0.92      0.36  9.7   0.0075 0.012  0.35
## 
##  Item statistics 
##          n raw.r std.r r.cor r.drop mean   sd
## snap1  334  0.55  0.55  0.50   0.48 0.90 0.90
## snap2  336  0.66  0.67  0.65   0.61 0.61 0.77
## snap3  336  0.61  0.60  0.57   0.54 0.65 0.85
## snap4  332  0.63  0.63  0.61   0.57 0.52 0.79
## snap5  328  0.71  0.71  0.70   0.66 0.56 0.75
## snap6  329  0.70  0.70  0.69   0.65 0.79 0.91
## snap7  334  0.64  0.64  0.61   0.58 0.60 0.87
## snap8  328  0.75  0.75  0.74   0.71 0.87 0.85
## snap9  330  0.68  0.68  0.66   0.62 0.73 0.88
## snap10 330  0.57  0.57  0.53   0.50 0.66 0.90
## snap11 329  0.65  0.66  0.63   0.60 0.55 0.80
## snap12 329  0.56  0.59  0.56   0.52 0.22 0.56
## snap13 333  0.52  0.53  0.49   0.46 0.35 0.67
## snap14 333  0.60  0.61  0.59   0.54 0.47 0.80
## snap15 331  0.53  0.52  0.49   0.46 0.80 0.95
## snap16 330  0.68  0.68  0.66   0.63 0.66 0.84
## snap17 328  0.68  0.67  0.66   0.62 0.75 0.90
## snap18 333  0.68  0.67  0.65   0.62 0.71 0.91
## 
## Non missing response frequency for each item
##           0    1    2    3 10 miss
## snap1  0.31 0.55 0.10 0.04  0 0.01
## snap2  0.54 0.35 0.08 0.03  0 0.01
## snap3  0.55 0.31 0.09 0.05  0 0.01
## snap4  0.63 0.27 0.07 0.04  0 0.02
## snap5  0.57 0.32 0.08 0.03  0 0.03
## snap6  0.46 0.37 0.09 0.08  0 0.03
## snap7  0.60 0.26 0.08 0.06  0 0.01
## snap8  0.37 0.46 0.11 0.06  0 0.03
## snap9  0.51 0.32 0.12 0.05  0 0.03
## snap10 0.56 0.27 0.11 0.06  0 0.03
## snap11 0.61 0.27 0.09 0.03  0 0.03
## snap12 0.84 0.12 0.03 0.01  0 0.03
## snap13 0.74 0.18 0.06 0.02  0 0.02
## snap14 0.69 0.20 0.08 0.04  0 0.02
## snap15 0.49 0.31 0.11 0.09  0 0.02
## snap16 0.53 0.32 0.11 0.05  0 0.03
## snap17 0.50 0.31 0.12 0.06  0 0.03
## snap18 0.53 0.29 0.11 0.07  0 0.02

The parents also completed the validated Brazilian version of the Strengths and Difficulties Questionnaire (SDQ) added to the impact supplement.24, 25 The SDQ is a 25-item instrument that was developed to evaluate psychosocial adjustment from the perspective of parents and/or teachers. The SDQ consists of five scales, each with five items that assess emotional symptoms, conduct problems, hyperactivity/inattention, peer problems, and prosocial behavior problems. The parents completed the parental version of the SDQ and the impact supplement that was related to any adjustment symptoms in terms of chronicity, resultant distress, social impairment, and burden to others, a compulsory criterion for an ADHD diagnosis according to the DSM-5.23 The Cronbach’s alpha of this questionnaire was 0.8 (95% CI = [0.77,0.83]).

## 
## Reliability analysis   
## Call: alpha(x = ., check.keys = T)
## 
##   raw_alpha std.alpha G6(smc) average_r S/N   ase mean   sd median_r
##        0.8       0.8    0.84      0.14 4.1 0.015 0.76 0.29     0.14
## 
##  lower alpha upper     95% confidence boundaries
## 0.77 0.8 0.83 
## 
##  Reliability if an item is dropped:
##        raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## sdq_1-      0.79      0.79    0.83      0.14 3.8    0.016 0.012  0.13
## sdq2        0.79      0.80    0.83      0.14 3.9    0.016 0.012  0.14
## sdq3        0.80      0.80    0.83      0.14 4.0    0.015 0.012  0.14
## sdq4-       0.79      0.80    0.83      0.14 3.9    0.016 0.013  0.13
## sdq5        0.79      0.79    0.82      0.14 3.8    0.017 0.012  0.13
## sdq6        0.80      0.80    0.83      0.14 4.1    0.015 0.013  0.14
## sdq7        0.79      0.80    0.83      0.14 3.9    0.016 0.011  0.13
## sdq8        0.81      0.81    0.84      0.15 4.2    0.015 0.011  0.15
## sdq9-       0.80      0.80    0.83      0.14 4.0    0.015 0.012  0.14
## sdq10       0.79      0.80    0.83      0.14 3.9    0.016 0.012  0.13
## sdq11       0.80      0.80    0.83      0.14 4.0    0.015 0.013  0.14
## sdq12       0.79      0.79    0.82      0.14 3.8    0.016 0.012  0.13
## sdq13       0.79      0.79    0.82      0.14 3.8    0.016 0.012  0.13
## sdq14       0.80      0.80    0.83      0.14 3.9    0.016 0.013  0.14
## sdq15       0.79      0.79    0.82      0.14 3.8    0.017 0.012  0.13
## sdq16       0.80      0.80    0.83      0.14 4.0    0.016 0.013  0.15
## sdq17-      0.80      0.80    0.83      0.14 4.0    0.016 0.013  0.14
## sdq18       0.79      0.79    0.82      0.13 3.7    0.016 0.012  0.13
## sdq19       0.79      0.79    0.83      0.14 3.9    0.016 0.013  0.13
## sdq20-      0.80      0.80    0.83      0.14 4.0    0.016 0.012  0.14
## sdq21       0.79      0.79    0.83      0.14 3.9    0.016 0.012  0.13
## sdq22       0.80      0.81    0.84      0.15 4.2    0.015 0.012  0.15
## sdq23       0.81      0.81    0.84      0.15 4.2    0.015 0.012  0.15
## sdq24       0.79      0.80    0.83      0.14 3.9    0.016 0.013  0.14
## sdq25       0.79      0.79    0.82      0.14 3.8    0.016 0.012  0.13
## 
##  Item statistics 
##          n raw.r std.r r.cor r.drop  mean   sd
## sdq_1- 329  0.44  0.49  0.47  0.388 1.228 0.47
## sdq2   333  0.43  0.42  0.38  0.349 0.363 0.66
## sdq3   328  0.39  0.35  0.30  0.282 0.854 0.86
## sdq4-  329  0.43  0.45  0.41  0.357 1.422 0.67
## sdq5   332  0.58  0.55  0.53  0.500 0.666 0.84
## sdq6   326  0.32  0.31  0.25  0.223 0.475 0.73
## sdq7   331  0.42  0.43  0.41  0.342 0.559 0.66
## sdq8   323  0.21  0.18  0.12  0.098 0.904 0.81
## sdq9-  325  0.31  0.36  0.32  0.237 1.422 0.63
## sdq10  329  0.48  0.46  0.43  0.388 0.565 0.81
## sdq11  326  0.30  0.36  0.31  0.246 0.150 0.44
## sdq12  330  0.55  0.54  0.53  0.480 0.327 0.63
## sdq13  331  0.55  0.54  0.52  0.471 0.574 0.76
## sdq14  326  0.38  0.43  0.39  0.314 0.227 0.51
## sdq15  328  0.59  0.56  0.55  0.502 0.954 0.85
## sdq16  329  0.39  0.36  0.32  0.294 1.304 0.76
## sdq17- 330  0.34  0.37  0.32  0.265 1.276 0.55
## sdq18  327  0.59  0.58  0.58  0.514 0.468 0.69
## sdq19  328  0.49  0.48  0.44  0.405 0.518 0.74
## sdq20- 329  0.37  0.38  0.34  0.287 1.547 0.69
## sdq21  329  0.46  0.47  0.44  0.377 0.684 0.69
## sdq22  328  0.17  0.23  0.17  0.126 0.052 0.28
## sdq23  329  0.24  0.23  0.16  0.140 1.128 0.81
## sdq24  328  0.46  0.43  0.39  0.355 0.723 0.83
## sdq25  333  0.50  0.52  0.50  0.435 0.613 0.70
## 
## Non missing response frequency for each item
##          0    1    2 3 miss
## sdq_1 0.02 0.18 0.80 0 0.03
## sdq2  0.74 0.16 0.10 0 0.02
## sdq3  0.45 0.24 0.30 0 0.03
## sdq4  0.10 0.23 0.67 0 0.03
## sdq5  0.58 0.19 0.23 0 0.02
## sdq6  0.67 0.19 0.14 0 0.04
## sdq7  0.54 0.37 0.10 0 0.02
## sdq8  0.38 0.33 0.28 0 0.05
## sdq9  0.07 0.27 0.65 0 0.04
## sdq10 0.64 0.15 0.21 0 0.03
## sdq11 0.88 0.09 0.03 0 0.04
## sdq12 0.76 0.15 0.09 0 0.03
## sdq13 0.60 0.24 0.17 0 0.02
## sdq14 0.81 0.15 0.04 0 0.04
## sdq15 0.38 0.28 0.34 0 0.03
## sdq16 0.19 0.33 0.49 0 0.03
## sdq17 0.05 0.17 0.78 0 0.03
## sdq18 0.65 0.24 0.12 0 0.04
## sdq19 0.63 0.23 0.15 0 0.03
## sdq20 0.11 0.32 0.57 0 0.03
## sdq21 0.45 0.42 0.13 0 0.03
## sdq22 0.96 0.02 0.02 0 0.03
## sdq23 0.27 0.33 0.40 0 0.03
## sdq24 0.52 0.23 0.25 0 0.03
## sdq25 0.51 0.36 0.13 0 0.02

Information obtained from teachers. Before conducting the interviews with the parents, the teachers completed the MTA-SNAP-IV and were also asked to provide structured information about the students’ school performance, with measurements of overall achievement for the school year that was derived from competencies in language, mathematics, science, and social studies. The Cronbach’s alpha of this questionnaire was 0.97 (95% CI = [0.96,0.97]). Adolescents were ranked as being below expectations (i.e., failed to achieve a minimal number of established milestones for the year), matching expectations, or exceeding expectations (i.e., achieved milestones that are only expected to be achieved in the following school year) for the grade in accordance with education board standards.

## 
## Reliability analysis   
## Call: alpha(x = .)
## 
##   raw_alpha std.alpha G6(smc) average_r S/N    ase mean   sd median_r
##       0.97      0.97    0.98      0.63  30 0.0026 0.64 0.71     0.61
## 
##  lower alpha upper     95% confidence boundaries
## 0.96 0.97 0.97 
## 
##  Reliability if an item is dropped:
##         raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## snapp1       0.96      0.97    0.98      0.62  28   0.0028 0.013  0.61
## snapp2       0.96      0.97    0.98      0.62  28   0.0028 0.013  0.61
## snapp3       0.96      0.97    0.98      0.62  28   0.0029 0.015  0.60
## snapp4       0.96      0.96    0.98      0.62  28   0.0029 0.013  0.61
## snapp5       0.96      0.97    0.98      0.62  28   0.0029 0.013  0.61
## snapp6       0.96      0.97    0.98      0.62  28   0.0028 0.013  0.61
## snapp7       0.97      0.97    0.98      0.62  28   0.0028 0.015  0.60
## snapp8       0.96      0.97    0.98      0.62  28   0.0028 0.014  0.61
## snapp9       0.96      0.97    0.98      0.62  28   0.0029 0.014  0.61
## snapp10      0.97      0.97    0.98      0.64  30   0.0027 0.014  0.62
## snapp11      0.96      0.97    0.98      0.62  28   0.0029 0.015  0.60
## snapp12      0.97      0.97    0.98      0.63  29   0.0027 0.014  0.62
## snapp13      0.97      0.97    0.98      0.63  29   0.0028 0.015  0.62
## snapp14      0.97      0.97    0.98      0.63  28   0.0028 0.015  0.61
## snapp15      0.97      0.97    0.98      0.64  30   0.0026 0.014  0.62
## snapp16      0.97      0.97    0.98      0.63  29   0.0027 0.014  0.62
## snapp17      0.96      0.97    0.98      0.62  28   0.0028 0.014  0.61
## snapp18      0.96      0.97    0.98      0.62  28   0.0028 0.015  0.61
## 
##  Item statistics 
##           n raw.r std.r r.cor r.drop mean   sd
## snapp1  339  0.83  0.83  0.82   0.81 0.94 0.92
## snapp2  339  0.83  0.83  0.82   0.81 0.87 0.92
## snapp3  338  0.86  0.86  0.85   0.84 0.64 0.86
## snapp4  338  0.87  0.87  0.87   0.86 0.78 0.97
## snapp5  338  0.84  0.84  0.84   0.82 0.75 0.96
## snapp6  339  0.83  0.82  0.82   0.80 0.76 0.96
## snapp7  339  0.80  0.81  0.80   0.78 0.45 0.81
## snapp8  338  0.82  0.81  0.80   0.79 1.10 1.01
## snapp9  337  0.86  0.85  0.85   0.84 0.78 0.93
## snapp10 338  0.70  0.70  0.68   0.66 0.41 0.96
## snapp11 338  0.84  0.85  0.84   0.82 0.52 0.88
## snapp12 336  0.73  0.75  0.73   0.70 0.25 0.67
## snapp13 339  0.76  0.77  0.76   0.73 0.33 0.73
## snapp14 338  0.79  0.80  0.79   0.76 0.40 0.79
## snapp15 338  0.70  0.69  0.67   0.66 0.97 1.05
## snapp16 337  0.76  0.77  0.76   0.73 0.44 0.83
## snapp17 339  0.80  0.81  0.81   0.78 0.50 0.85
## snapp18 339  0.81  0.82  0.82   0.79 0.55 0.90
## 
## Non missing response frequency for each item
##            0    1    2    3 10 miss
## snapp1  0.38 0.36 0.19 0.07  0 0.00
## snapp2  0.43 0.33 0.17 0.06  0 0.00
## snapp3  0.56 0.29 0.09 0.06  0 0.00
## snapp4  0.52 0.25 0.14 0.08  0 0.00
## snapp5  0.53 0.26 0.12 0.08  0 0.00
## snapp6  0.52 0.29 0.11 0.09  0 0.00
## snapp7  0.69 0.22 0.03 0.06  0 0.00
## snapp8  0.34 0.35 0.19 0.13  0 0.00
## snapp9  0.48 0.34 0.10 0.08  0 0.01
## snapp10 0.76 0.15 0.01 0.07  0 0.00
## snapp11 0.67 0.20 0.05 0.07  0 0.00
## snapp12 0.85 0.09 0.03 0.03  0 0.01
## snapp13 0.78 0.15 0.02 0.04  0 0.00
## snapp14 0.74 0.17 0.04 0.05  0 0.00
## snapp15 0.43 0.30 0.14 0.13  0 0.00
## snapp16 0.72 0.18 0.05 0.06  0 0.01
## snapp17 0.68 0.20 0.05 0.06  0 0.00
## snapp18 0.66 0.21 0.06 0.07  0 0.00

14.4 Table 3

The table below summarises the main statistics. Double checked on november 6 2020. In the manuscript, this represents the table 3.

cefaleia_mes No headache Low frequency Intermediate High frequency
Count 40 254 28 17
mastery_total_mean 52.0 50.6 45.9 43.9
optimism_mean 18.6 18.1 16.4 15.1
self_efficacy_mean 25.1 24.1 22.0 21.1
adaptability_numeric_mean 8.0 8.4 7.3 7.9
relatedness_total_mean 67.4 63.7 63.1 58.9
trust_mean 19.1 18.6 17.9 16.6
support_mean 18.5 17.4 16.8 17.5
comfort_mean 10.9 10.4 10.8 9.2
tolerance_numeric_mean 19.1 17.0 17.7 15.2
reactivity_total_mean 28.5 30.4 36.1 38.4
sensitivity_mean 10.3 11.1 13.1 13.9
recovery_mean 5.5 6.2 7.8 7.0
impairment_mean 12.7 13.1 15.2 17.5
ressources_total_mean 59.7 57.2 54.5 51.4
vulnerability_total_mean -31.2 -26.8 -18.4 -13.1
mastery_total_sd 11.9 11.9 11.8 11.4
optimism_sd 4.9 4.9 4.6 5.0
self_efficacy_sd 6.8 6.8 6.7 5.6
adaptability_numeric_sd 2.4 2.3 2.1 2.1
relatedness_total_sd 15.7 15.0 14.5 11.4
trust_sd 5.2 4.9 5.3 4.8
support_sd 4.6 4.8 4.7 3.5
comfort_sd 3.1 3.4 2.9 4.0
tolerance_numeric_sd 4.9 4.5 4.9 5.1
reactivity_total_sd 16.8 14.8 16.9 13.6
sensitivity_sd 5.4 4.8 5.1 5.3
recovery_sd 4.6 4.0 3.8 4.1
impairment_sd 8.8 8.6 10.3 7.0
ressources_total_sd 13.1 12.1 12.2 8.0
vulnerability_total_sd 22.4 20.0 20.9 17.7

15 Results

We accessed the relationship between headache and the results obtained by each scales descriptively by using graphs and frequencies. Bar graphs are easy to understan, and very useful for showing the pattern of the results.

15.1 Table 2

## $data
##           Outcome
## Predictor    1  0 Total
##   [10,12]   89 10    99
##   (12,15]  141 19   160
##   (15,Inf]  69 11    80
##   Total    299 40   339
## 
## $measure
##           odds ratio with 95% C.I.
## Predictor  estimate     lower    upper
##   [10,12]  1.000000        NA       NA
##   (12,15]  1.190430 0.5359415 2.799646
##   (15,Inf] 1.414157 0.5586810 3.621071
## 
## $p.value
##           two-sided
## Predictor  midp.exact fisher.exact chi.square
##   [10,12]          NA           NA         NA
##   (12,15]   0.6740582    0.8396325  0.6599642
##   (15,Inf]  0.4617109    0.4899402  0.4506905
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor   1  0 Total
##    Female 172  9   181
##    Male   127 31   158
##    Total  299 40   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate    lower    upper
##    Female 1.000000       NA       NA
##    Male   4.590844 2.183984 10.63592
## 
## $p.value
##          two-sided
## Predictor   midp.exact fisher.exact   chi.square
##    Female           NA           NA           NA
##    Male   2.917951e-05 3.356408e-05 3.040773e-05
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor   1  0 Total
##     1     211 28   239
##     2      80 10    90
##     3       8  2    10
##     Total 299 40   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##         1    1.000    NA    NA
##         2    0.951 0.419  2.00
##         3    1.976 0.261  8.62
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##         1         NA           NA         NA
##         2      0.898        1.000      0.878
##         3      0.450        0.344      0.430
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor   1  0 Total
##     AB     77 16    93
##     C     184 19   203
##     DE     38  5    43
##     Total 299 40   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##        AB    1.000    NA    NA
##        C     0.498 0.242  1.03
##        DE    0.647 0.195  1.81
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##        AB         NA           NA         NA
##        C      0.0613       0.0792     0.0523
##        DE     0.4223       0.4566     0.4027
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"

15.2 Low frequency

## $data
##           Outcome
## Predictor    1  0 Total
##   [10,12]   79 20    99
##   (12,15]  117 43   160
##   (15,Inf]  58 22    80
##   Total    254 85   339
## 
## $measure
##           odds ratio with 95% C.I.
## Predictor  estimate lower upper
##   [10,12]      1.00    NA    NA
##   (12,15]      1.44 0.797  2.69
##   (15,Inf]     1.49 0.743  3.02
## 
## $p.value
##           two-sided
## Predictor  midp.exact fisher.exact chi.square
##   [10,12]          NA           NA         NA
##   (12,15]       0.228        0.237      0.224
##   (15,Inf]      0.260        0.289      0.252
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor   1  0 Total
##    Female 141 40   181
##    Male   113 45   158
##    Total  254 85   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##    Female      1.0    NA    NA
##    Male        1.4 0.856   2.3
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##    Female         NA           NA         NA
##    Male         0.18        0.209      0.176
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor   1  0 Total
##     1     177 62   239
##     2      72 18    90
##     3       5  5    10
##     Total 254 85   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##         1    1.000    NA    NA
##         2    0.718 0.387  1.28
##         3    2.840 0.741 10.89
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##         1         NA           NA         NA
##         2      0.266        0.314     0.2628
##         3      0.124        0.139     0.0928
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor   1  0 Total
##     AB     67 26    93
##     C     157 46   203
##     DE     30 13    43
##     Total 254 85   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##        AB    1.000    NA    NA
##        C     0.754 0.432  1.33
##        DE    1.120 0.494  2.46
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##        AB         NA           NA         NA
##        C       0.329        0.381      0.324
##        DE      0.782        0.839      0.785
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"

15.3 Intermediate frequency

## $data
##           Outcome
## Predictor   1   0 Total
##   [10,12]   7  92    99
##   (12,15]  13 147   160
##   (15,Inf]  8  72    80
##   Total    28 311   339
## 
## $measure
##           odds ratio with 95% C.I.
## Predictor  estimate lower upper
##   [10,12]     1.000    NA    NA
##   (12,15]     0.869 0.312  2.23
##   (15,Inf]    0.688 0.227  2.04
## 
## $p.value
##           two-sided
## Predictor  midp.exact fisher.exact chi.square
##   [10,12]          NA           NA         NA
##   (12,15]       0.775        0.815      0.757
##   (15,Inf]      0.496        0.590      0.482
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor  1   0 Total
##    Female 19 162   181
##    Male    9 149   158
##    Total  28 311   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##    Female     1.00    NA    NA
##    Male       1.92 0.859  4.63
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##    Female         NA           NA         NA
##    Male        0.114        0.118      0.109
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor  1   0 Total
##     1     20 219   239
##     2      5  85    90
##     3      3   7    10
##     Total 28 311   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate  lower upper
##         1     1.00     NA    NA
##         2     1.52 0.5873  4.77
##         3     0.21 0.0521  1.09
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##         1         NA           NA         NA
##         2      0.409       0.4884     0.3907
##         3      0.062       0.0539     0.0206
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor  1   0 Total
##     AB     7  86    93
##     C     17 186   203
##     DE     4  39    43
##     Total 28 311   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##        AB    1.000    NA    NA
##        C     0.902 0.333  2.19
##        DE    0.783 0.217  3.26
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##        AB         NA           NA         NA
##        C       0.826        1.000      0.804
##        DE      0.719        0.742      0.724
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"

15.4 High frequency

## $data
##           Outcome
## Predictor   1   0 Total
##   [10,12]   3  96    99
##   (12,15]  11 149   160
##   (15,Inf]  3  77    80
##   Total    17 322   339
## 
## $measure
##           odds ratio with 95% C.I.
## Predictor  estimate lower upper
##   [10,12]     1.000    NA    NA
##   (12,15]     0.440 0.093  1.48
##   (15,Inf]    0.803 0.135  4.79
## 
## $p.value
##           two-sided
## Predictor  midp.exact fisher.exact chi.square
##   [10,12]          NA           NA         NA
##   (12,15]       0.195         0.26      0.184
##   (15,Inf]      0.800         1.00      0.790
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## $data
##          Outcome
## Predictor  1   0 Total
##    Female 12 169   181
##    Male    5 153   158
##    Total  17 322   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##    Female     1.00    NA    NA
##    Male       2.13 0.761  6.98
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##    Female         NA           NA         NA
##    Male        0.154        0.212      0.145
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"
## 
##  
##    Cell Contents
## |-------------------------|
## |                       N |
## | Chi-square contribution |
## |           N / Row Total |
## |           N / Col Total |
## |         N / Table Total |
## |-------------------------|
## 
##  
## Total Observations in Table:  339 
## 
##  
##                                            | .$high_frequency 
## .$cor_1_branca_2_nao_branca_3_nao_informou |         0 |         1 | Row Total | 
## -------------------------------------------|-----------|-----------|-----------|
##                                          1 |       225 |        14 |       239 | 
##                                            |     0.018 |     0.339 |           | 
##                                            |     0.941 |     0.059 |     0.705 | 
##                                            |     0.699 |     0.824 |           | 
##                                            |     0.664 |     0.041 |           | 
## -------------------------------------------|-----------|-----------|-----------|
##                                          2 |        87 |         3 |        90 | 
##                                            |     0.027 |     0.507 |           | 
##                                            |     0.967 |     0.033 |     0.265 | 
##                                            |     0.270 |     0.176 |           | 
##                                            |     0.257 |     0.009 |           | 
## -------------------------------------------|-----------|-----------|-----------|
##                                          3 |        10 |         0 |        10 | 
##                                            |     0.026 |     0.501 |           | 
##                                            |     1.000 |     0.000 |     0.029 | 
##                                            |     0.031 |     0.000 |           | 
##                                            |     0.029 |     0.000 |           | 
## -------------------------------------------|-----------|-----------|-----------|
##                               Column Total |       322 |        17 |       339 | 
##                                            |     0.950 |     0.050 |           | 
## -------------------------------------------|-----------|-----------|-----------|
## 
##  
## Statistics for All Table Factors
## 
## 
## Pearson's Chi-squared test 
## ------------------------------------------------------------
## Chi^2 =  1.42     d.f. =  2     p =  0.492 
## 
## 
## 
## [1] 1.8
## [1] Inf
## $data
##          Outcome
## Predictor  1   0 Total
##     AB     3  90    93
##     C     10 193   203
##     DE     4  39    43
##     Total 17 322   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate  lower upper
##        AB    1.000     NA    NA
##        C     0.667 0.1397  2.28
##        DE    0.332 0.0588  1.66
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##        AB         NA           NA         NA
##        C       0.541        0.761      0.508
##        DE      0.174        0.207      0.136
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "median-unbiased estimate & mid-p exact CI"

16 Mastery (table 4)

To check the RSCA indexes, scales, and subscales as a function of headache frequency, a Robust one-way ANOVA was carried out using the raw score results as the dependent variable. The effect of headache on the Mastery results was significant (F(3, 335) = 3.13, p = 0.03). Post-hoc comparison revealed significant differences between the headache-free group and the intermediate (Δ: -6.096, CI 95% [-11.96, -0.97]) and high frequency group (Δ: -8.39, CI 95% [-15.17, -1.23]).

## 
##  Breusch Pagan Test for Heteroskedasticity
##  -----------------------------------------
##  Ho: the variance is constant            
##  Ha: the variance is not constant        
## 
##                   Data                    
##  -----------------------------------------
##  Response : mastery_total 
##  Variables: fitted values of mastery_total 
## 
##       Test Summary        
##  -------------------------
##  DF            =    1 
##  Chi2          =    0.1495 
##  Prob > Chi2   =    0.6990
## -----------------------------------------------
##        Test             Statistic       pvalue  
## -----------------------------------------------
## Shapiro-Wilk              0.9899         0.0189 
## Kolmogorov-Smirnov        0.0514         0.3329 
## Cramer-von Mises         26.8886         0.0000 
## Anderson-Darling          0.5751         0.1344 
## -----------------------------------------------
Df F Pr(>F)
(Intercept) 1 761.45 0.00
factor(cefaleia_mes) 3 3.13 0.03
Residuals 335 NA NA
Re-checked on Janury 8, 2021.
estimate original boot_bias boot_se boot_med boot_skew boot_kurtosis x2_5_percent x97_5_percent r sig sd_aprox epm_apro t_stats p_val
1 52.7 0.0 1.7 52.8 -0.1 0.1 49.4 56.1 1000 p < 0.05 54.4 1.7 30.6 0.0
2 -1.9 0.0 1.9 -1.9 0.0 0.2 -5.6 1.8 1000 ns 59.5 1.9 -1.0 0.3
3 -6.5 0.0 2.7 -6.5 -0.1 0.0 -11.9 -1.2 1000 p < 0.05 86.7 2.7 -2.4 0.0
4 -8.4 -0.2 3.5 -8.5 0.0 -0.3 -15.0 -1.4 1000 p < 0.05 109.6 3.5 -2.4 0.0

16.1 Graphical analysis

The following images report the confidence interval and the variables distribution. These results provide evidence that after the bootstrap, data were normally distributed.

16.2 ANOVA with Post hoc

## 
## 
## ANOVA results using mastery_total as the dependent variable
##  
## 
##             Predictor        SS  df        MS      F    p partial_eta2
##           (Intercept) 108264.02   1 108264.02 767.90 .000             
##  factor(cefaleia_mes)   1367.01   3    455.67   3.23 .023          .03
##                 Error  47230.99 335    140.99                         
##  CI_90_partial_eta2
##                    
##          [.00, .06]
##                    
## 
## Note: Values in square brackets indicate the bounds of the 90% confidence interval for partial eta-squared
## $emmeans
##  cefaleia_mes   emmean    SE  df lower.CL upper.CL
##  No headache      52.0 1.877 335     47.3     56.7
##  Low frequency    50.6 0.745 335     48.8     52.5
##  Intermediate     45.9 2.244 335     40.3     51.5
##  High frequency   43.9 2.880 335     36.7     51.1
## 
## Confidence level used: 0.95 
## Conf-level adjustment: mvt method for 4 estimates 
## 
## $contrasts
##  contrast                     estimate   SE  df t.ratio p.value
##  Low frequency - No headache     -1.38 2.02 335 -0.685  0.8320 
##  Intermediate - No headache      -6.10 2.93 335 -2.084  0.0970 
##  High frequency - No headache    -8.14 3.44 335 -2.369  0.0490 
## 
## P value adjustment: mvt method for 3 tests
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Dunnett Contrasts
## 
## 
## Fit: lm(formula = mastery_total ~ cefaleia_mes, data = base_uso)
## 
## Linear Hypotheses:
##                                   Estimate Std. Error t value Pr(>|t|)  
## Low frequency - No headache == 0     -1.38       2.02   -0.68    0.832  
## Intermediate - No headache == 0      -6.10       2.93   -2.08    0.097 .
## High frequency - No headache == 0    -8.14       3.44   -2.37    0.049 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)

16.3 Categorical outcomes

Based on DSM-5 recommendation, we used results below of 1.5 standard deviation to assign children to a ‘at risk’ group, as previously cited.

x
Logistic regression predicting mastery_prob
OR(95%CI) P(Wald’s test) P(LR-test)
cefaleia_mes: ref.=No headache 0.11
Low frequency 2.45 (0.31,19.06) 0.393
Intermediate 6.5 (0.69,61.64) 0.103
High frequency 8.36 (0.8,87.11) 0.076
x
Log-likelihood = -81.0689

No. of observations = 339 AIC value = 170.1378 |

The table below reports the estimate values in the log-odds form, the standard error of each estimate, and the odds ratio. The interpretation of the estimates considers the log odds scale. As an example, the expected change in log odds is 0.57 for those with a low-frequency headache when compared to those with no headache. The Odds-Ratio is the ratio of the odds for the groups, is always non-negative, and between 0 and \(infinity\). its interpretation is recommended, once its results are more straightforward and intuitive. The odds ratio of having a low sense of mastery among those with low-frequency headache is 1.77 compared to those with no headache, i.e, the odds of low sense of mastery in headache-free participants is estimated to be 1.77 times the odds of low sense of mastery in children experiencing low-frequency headache (77.1% higher = (1-1.771*100)). However, this result is not significant. Significant results are marked by asterisks.

16.4 Relative Risk

Relative risks can be estimated by OR applying this equation:

\[\hat{RR} = \frac{OR}{1-Risk_{control}+Risk_{control}*OR}\]

are computed below from the contingency table between the factor and the outcome:

The following plot shows the Odds-ratio results. The overall effect estimate and its 95% confidence intervals are plotted, and the vertical line right over the number 1 means equal chances.

16.5 Mastery - Optmism

D f F Pr (>F)
(Intercept) 1 594.93 0.00
factor(cefaleia_mes) 3 3.52 0.02
Residuals 335 NA NA
estimate original boot_bias boot_se boot_med boot_skew boot_kurtosis x2_5_percent x97_5_percent r sig sd_aprox epm_apro t_stats p_val
1 18.94 -0.02 0.73 18.93 -0.14 0.00 17.52 20.40 1000 p < 0.05 23.2 0.73 25.81 0.00
2 -0.73 0.02 0.80 -0.69 0.11 -0.08 -2.31 0.82 1000 ns 25.3 0.80 -0.91 0.36
3 -2.35 0.02 1.09 -2.31 -0.03 -0.10 -4.49 -0.23 1000 p < 0.05 34.4 1.09 -2.16 0.03
4 -3.94 0.06 1.44 -3.92 0.08 -0.06 -6.83 -1.17 1000 p < 0.05 45.6 1.44 -2.73 0.01

16.6 ANOVA with Post hoc

## $emmeans
##  cefaleia_mes   emmean    SE  df lower.CL upper.CL
##  No headache      18.6 0.768 335     16.7     20.6
##  Low frequency    18.1 0.305 335     17.3     18.9
##  Intermediate     16.4 0.917 335     14.1     18.7
##  High frequency   15.1 1.177 335     12.1     18.0
## 
## Confidence level used: 0.95 
## Conf-level adjustment: mvt method for 4 estimates 
## 
## $contrasts
##  contrast                     estimate    SE  df t.ratio p.value
##  Low frequency - No headache     -0.52 0.826 335 -0.633  0.8610 
##  Intermediate - No headache      -2.23 1.196 335 -1.866  0.1550 
##  High frequency - No headache    -3.57 1.405 335 -2.537  0.0310 
## 
## P value adjustment: mvt method for 3 tests
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Dunnett Contrasts
## 
## 
## Fit: lm(formula = optimism ~ cefaleia_mes, data = base_uso)
## 
## Linear Hypotheses:
##                                   Estimate Std. Error t value Pr(>|t|)  
## Low frequency - No headache == 0    -0.523      0.826   -0.63    0.861  
## Intermediate - No headache == 0     -2.232      1.196   -1.87    0.155  
## High frequency - No headache == 0   -3.566      1.405   -2.54    0.031 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)

16.7 Mastery - Self-efficacy

D f F Pr (>F)
(Intercept) 1 537.27 0.00
factor(cefaleia_mes) 3 2.05 0.11
Residuals 335 NA NA

16.8 Mastery - adaptability_numeric

D f F Pr (>F)
(Intercept) 1 188.64 0.00
factor(cefaleia_mes) 3 1.01 0.39
Residuals 136 NA NA

17 Relatedness (SR) (table 4)

Contrasting to this result, no significant effect was found on the results of the Relatedness scale (F(3,335) = 2.04, p = 0.11).

D f F Pr (>F)
(Intercept) 1 862.54 0.00
cefaleia_mes 3 2.04 0.11
Residuals 335 NA NA

17.1 Categorical outcomes

The sense of Relatedness scale is composed of 24 items and includes four components aspects that contribute to a sense of relatedness, such as the sense of trust, perceived access to support, comfort with others and tolerance. children with scores equal to or lower than 45 in the T-score scale were considered at risk and, therefore, assigned a value of 1.

x
Logistic regression predicting relatedness_prob
OR(95%CI) P(Wald’s test) P(LR-test)
cefaleia_mes: ref.=No headache 0.52
Low frequency 3.15 (0.41,24.23) 0.27
Intermediate 4.68 (0.46,47.54) 0.192
High frequency 2.44 (0.14,41.4) 0.538
x
Log-likelihood = -85.5494

No. of observations = 339 AIC value = 179.0989 |

17.2 Relative Risk

The graph below illustrated the Odds-ratio of having low sense of relatedness. The higher the odds, the more likely the child is to suffer from a low sense of relatedness.

17.3 Relatedness - sense of trust

D f F Pr (>F)
(Intercept) 1 593.88 0.00
factor(cefaleia_mes) 3 1.46 0.23
Residuals 335 NA NA

17.4 Relatedness - sense of support

D f F Pr (>F)
(Intercept) 1 677.40 0.00
factor(cefaleia_mes) 3 1.18 0.32
Residuals 335 NA NA

17.5 Relatedness - sense of comfort

D f F Pr (>F)
(Intercept) 1 400.91 0.00
factor(cefaleia_mes) 3 1.14 0.33
Residuals 335 NA NA

17.6 Relatedness - tolerance

D f F Pr (>F)
(Intercept) 1 627.21 0
factor(cefaleia_mes) 3 4.68 0
Residuals 278 NA NA

18 Emotional Reactivity (table 4)

The effect of headache on the Reactivity results was significant (F(3, 335) = 3.04, p =.03). The higher frequency headache group had higher results on the Reactivity scale when compared to group headache-free (Δ: 11.07, CI 95% [2.43, 20.12]).

D f F Pr (>F)
(Intercept) 1 134.46 0.00
cefaleia_mes 3 3.04 0.03
Residuals 335 NA NA
estimate original boot_bias boot_se boot_med boot_skew boot_kurtosis x2_5_percent x97_5_percent r sig sd_aprox epm_apro t_stats p_val
1 27.7 -0.1 2.8 27.6 0.0 -0.3 22.3 33.4 1000 p < 0.05 89.4 2.8 9.8 0.0
2 1.7 0.1 3.0 1.8 0.0 -0.3 -4.2 7.5 1000 ns 94.6 3.0 0.6 0.6
3 6.5 0.0 4.0 6.5 -0.1 -0.1 -1.5 14.4 1000 ns 128.0 4.0 1.6 0.1
4 11.1 -0.1 4.5 10.9 0.0 0.7 2.4 20.0 1000 p < 0.05 141.8 4.5 2.5 0.0

18.1 ANOVA with Post hoc

## $emmeans
##  cefaleia_mes   emmean   SE  df lower.CL upper.CL
##  No headache      28.5 2.40 335     22.5     34.5
##  Low frequency    30.4 0.95 335     28.0     32.8
##  Intermediate     36.1 2.87 335     28.9     43.3
##  High frequency   38.4 3.68 335     29.1     47.6
## 
## Confidence level used: 0.95 
## Conf-level adjustment: mvt method for 4 estimates 
## 
## $contrasts
##  contrast                     estimate   SE  df t.ratio p.value
##  Low frequency - No headache      1.89 2.58 335 0.730   0.8050 
##  Intermediate - No headache       7.57 3.74 335 2.024   0.1110 
##  High frequency - No headache     9.85 4.40 335 2.242   0.0670 
## 
## P value adjustment: mvt method for 3 tests
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Dunnett Contrasts
## 
## 
## Fit: lm(formula = optimism ~ cefaleia_mes, data = base_uso)
## 
## Linear Hypotheses:
##                                   Estimate Std. Error t value Pr(>|t|)  
## Low frequency - No headache == 0    -0.523      0.826   -0.63    0.861  
## Intermediate - No headache == 0     -2.232      1.196   -1.87    0.155  
## High frequency - No headache == 0   -3.566      1.405   -2.54    0.031 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)

18.2 Categorical outcomes

x
Logistic regression predicting reactivity_prob
OR(95%CI) P(Wald’s test) P(LR-test)
cefaleia_mes: ref.=No headache 0.799
Low frequency 1.23 (0.35,4.3) 0.748
Intermediate 2.06 (0.42,10.01) 0.372
High frequency 1.64 (0.25,10.85) 0.605
x
Log-likelihood = -105.4643

No. of observations = 339 AIC value = 218.9285 |

18.3 Relative Risk

18.4 Reactivity - sensitivity

D f F Pr (>F)
(Intercept) 1 158.55 0.00
factor(cefaleia_mes) 3 2.64 0.05
Residuals 335 NA NA
estimate original boot_bias boot_se boot_med boot_skew boot_kurtosis x2_5_percent x97_5_percent sig
(Intercept) 10.42 -0.04 1.03 10.38 -0.07 -0.24 8.45 12.49 p < 0.05
factor(cefaleia_mes)Low frequency 0.48 0.06 1.08 0.52 0.11 -0.27 -1.70 2.54 ns
factor(cefaleia_mes)Intermediate 2.19 0.05 1.45 2.27 0.01 -0.29 -0.70 4.98 ns
factor(cefaleia_mes)High frequency 3.39 0.04 1.76 3.43 0.09 0.21 -0.11 6.81 ns

18.5 Reactivity - recovery

D f F Pr (>F)
(Intercept) 1 50.96 0.00
factor(cefaleia_mes) 3 2.15 0.09
Residuals 335 NA NA

18.6 Reactivity - impairment

D f F Pr (>F)
(Intercept) 1 76.35 0.00
factor(cefaleia_mes) 3 1.98 0.12
Residuals 335 NA NA

19 Resources Index (RI) (Table 4)

The effect of headache on the Resources scale was significant F(3, 335) = 2.99, p = 0.03). The high frequency group had significant results (Δ: -9.44, CI 95% [-15.02, -4.12]) when compared to headache-free group.

D f F Pr (>F)
(Intercept) 1 1064.55 0.00
cefaleia_mes 3 2.99 0.03
Residuals 335 NA NA
estimate original boot_bias boot_se boot_med boot_skew boot_kurtosis x2_5_percent x97_5_percent r sig sd_aprox epm_apro t_stats p_val
1 60.79 -0.02 1.97 60.77 -0.11 -0.10 56.95 64.67 1000 p < 0.05 62.3 1.97 30.86 0.00
2 -3.21 0.03 2.10 -3.18 0.11 0.17 -7.35 0.87 1000 ns 66.3 2.10 -1.53 0.13
3 -5.73 -0.02 3.03 -5.69 -0.12 0.19 -11.65 0.23 1000 ns 95.8 3.03 -1.89 0.06
4 -9.44 0.10 2.75 -9.33 0.11 -0.08 -14.94 -4.15 1000 p < 0.05 87.1 2.75 -3.43 0.00

19.1 ANOVA with Post hoc

## $emmeans
##  cefaleia_mes   emmean    SE  df lower.CL upper.CL
##  No headache      59.7 1.904 335     55.0     64.5
##  Low frequency    57.2 0.756 335     55.3     59.1
##  Intermediate     54.5 2.275 335     48.8     60.2
##  High frequency   51.4 2.920 335     44.1     58.7
## 
## Confidence level used: 0.95 
## Conf-level adjustment: mvt method for 4 estimates 
## 
## $contrasts
##  contrast                     estimate   SE  df t.ratio p.value
##  Low frequency - No headache     -2.56 2.05 335 -1.249  0.4540 
##  Intermediate - No headache      -5.23 2.97 335 -1.761  0.1910 
##  High frequency - No headache    -8.31 3.49 335 -2.385  0.0470 
## 
## P value adjustment: mvt method for 3 tests
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Dunnett Contrasts
## 
## 
## Fit: lm(formula = optimism ~ cefaleia_mes, data = base_uso)
## 
## Linear Hypotheses:
##                                   Estimate Std. Error t value Pr(>|t|)  
## Low frequency - No headache == 0    -0.523      0.826   -0.63    0.861  
## Intermediate - No headache == 0     -2.232      1.196   -1.87    0.155  
## High frequency - No headache == 0   -3.566      1.405   -2.54    0.031 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)

19.2 Categorical outcomes

The Resource Index is the standardized average of the Sense of Mastery and Sense of Relatedness scales.

x
Logistic regression predicting resources_prob
OR(95%CI) P(Wald’s test) P(LR-test)
cefaleia_mes: ref.=No headache 0.413
Low frequency 3.7 (0.48,28.23) 0.207
Intermediate 4.68 (0.46,47.54) 0.192
High frequency 5.2 (0.44,61.67) 0.191
x
Log-likelihood = -95.2047

No. of observations = 339 AIC value = 198.4095 |

19.3 Relative Risk

20 vulnerability index (VI) (table 4)

In the same direction of the results of Ressources, the main effect of headache on Vulnerability results was significant F(3, 335) = 4.05, p < 0.01) and the comparison revealed that intermediate (Δ: 11.04, CI 95% [0.12, 21.51]) and high fequency group (Δ: 18.34, CI 95% [7.42, 29.71]) had higher results than the disease-free group.

D f F Pr (>F)
(Intercept) 1 87.40 0.00
cefaleia_mes 3 4.05 0.01
Residuals 335 NA NA
estimate original boot_bias boot_se boot_med boot_skew boot_kurtosis x2_5_percent x97_5_percent r sig sd_aprox epm_apro t_stats p_val
1 -30.85 -0.16 3.73 -31.03 -0.12 -0.10 -38.00 -23.4 1000 p < 0.05 118 3.73 -8.27 0.00
2 3.99 0.13 4.00 4.09 0.14 -0.24 -3.98 11.7 1000 ns 126 4.00 1.00 0.32
3 11.04 0.02 5.45 11.02 0.01 -0.15 0.35 21.7 1000 p < 0.05 172 5.45 2.03 0.04
4 18.34 -0.18 5.74 18.35 -0.10 0.44 7.27 29.8 1000 p < 0.05 181 5.74 3.20 0.00

20.1 ANOVA with Post hoc

## $emmeans
##  cefaleia_mes   emmean   SE  df lower.CL upper.CL
##  No headache     -31.2 3.21 335    -39.3   -23.19
##  Low frequency   -26.8 1.27 335    -30.0   -23.59
##  Intermediate    -18.4 3.83 335    -28.0    -8.83
##  High frequency  -13.1 4.92 335    -25.4    -0.74
## 
## Confidence level used: 0.95 
## Conf-level adjustment: mvt method for 4 estimates 
## 
## $contrasts
##  contrast                     estimate   SE  df t.ratio p.value
##  Low frequency - No headache      4.44 3.45 335 1.288   0.4290 
##  Intermediate - No headache      12.80 5.00 335 2.560   0.0300 
##  High frequency - No headache    18.17 5.87 335 3.093   0.0060 
## 
## P value adjustment: mvt method for 3 tests
## 
##   Simultaneous Tests for General Linear Hypotheses
## 
## Multiple Comparisons of Means: Dunnett Contrasts
## 
## 
## Fit: lm(formula = optimism ~ cefaleia_mes, data = base_uso)
## 
## Linear Hypotheses:
##                                   Estimate Std. Error t value Pr(>|t|)  
## Low frequency - No headache == 0    -0.523      0.826   -0.63    0.861  
## Intermediate - No headache == 0     -2.232      1.196   -1.87    0.155  
## High frequency - No headache == 0   -3.566      1.405   -2.54    0.031 *
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## (Adjusted p values reported -- single-step method)

20.2 Categorical outcomes

The Vulnerability Index scrore is the standardized difference between the Emotional Reactivy T-score and the Resource Index score. It quantifies children’s personal vulnerability as the relative discrepancy between their combined self-perceived resources (the Resource Index) and their fragility as described by emotional reactivity (the Emotional Reactitivy Scale).

x
Logistic regression predicting vulnerability_prob
OR(95%CI) P(Wald’s test) P(LR-test)
cefaleia_mes: ref.=No headache 0.139
Low frequency 1.77 (0.22,14.06) 0.591
Intermediate 6.5 (0.69,61.64) 0.103
High frequency 5.2 (0.44,61.67) 0.191
x
Log-likelihood = -67.6092

No. of observations = 339 AIC value = 143.2185 |

20.3 Relative Risk

21 Migraine

First, I’ll add a specific variable to the dataset.

D f F Pr (>F)
(Intercept) 1 752.80 0.00
factor(migraine_status) 6 1.68 0.12
Residuals 332 NA NA
D f F Pr (>F)
(Intercept) 1 858.07 0.00
factor(migraine_status) 6 1.92 0.08
Residuals 332 NA NA
D f F Pr (>F)
(Intercept) 1 130.27 0.00
factor(migraine_status) 6 0.46 0.84
Residuals 332 NA NA
D f F Pr (>F)
(Intercept) 1 1075.44 0.00
factor(migraine_status) 6 2.25 0.04
Residuals 332 NA NA
D f F Pr (>F)
(Intercept) 1 84.18 0.00
factor(migraine_status) 6 1.25 0.28
Residuals 332 NA NA

22 Logistic regression

Another aim of this study is to check which variables are related to children at great psychological risk. These children were assigned to specific groups. The first one was composed of children that experienced at least one episode of headache in their lifetime and previously classified at low resources risk group (n =27). The second group included children with at last one episode of headache during his/her life and previously classified at high vulnerability group (n =17).

logistic regression was conducted including all established risk factors and subject variables (Migraine, age, race, sex, socioeconomic status, problems at sleeping, prematurity, use of tobacco and alcohol during pregnancy, low birth weight during delivery, low psychological strengths, and having ADHD). The variance inflation factor was calculated to determine the degree of multicollinearity present in the data results.

Checking the independence of the groups ? (Mcnemar)

risk_cefaleia_resources/risk_cefaleia_vulnerability 0 1 Total
0 96% (300) 4% (12) 100% (312)
1 81% (22) 19% (5) 100% (27)
Total 95% (322) 5% (17) 100% (339) 
## Confusion Matrix and Statistics
## 
##           Reference
## Prediction   0   1
##          0 300  12
##          1  22   5
##                                        
##                Accuracy : 0.9          
##                  95% CI : (0.863, 0.93)
##     No Information Rate : 0.95         
##     P-Value [Acc > NIR] : 1.000        
##                                        
##                   Kappa : 0.177        
##                                        
##  Mcnemar's Test P-Value : 0.123        
##                                        
##             Sensitivity : 0.2941       
##             Specificity : 0.9317       
##          Pos Pred Value : 0.1852       
##          Neg Pred Value : 0.9615       
##              Prevalence : 0.0501       
##          Detection Rate : 0.0147       
##    Detection Prevalence : 0.0796       
##       Balanced Accuracy : 0.6129       
##                                        
##        'Positive' Class : 1            
## 
## Data Frame Summary  
## base_uso  
## Dimensions: 339 x 10  
## Duplicates: 142  
## 
## -------------------------------------------------------------------------------------------------------------
## No   Variable     Stats / Values            Freqs (% of Valid)   Graph                   Valid      Missing  
## ---- ------------ ------------------------- -------------------- ----------------------- ---------- ---------
## 1    age          Mean (sd) : 13.9 (2.1)    10 :  7 ( 2.1%)                              339        0        
##      [numeric]    min < med < max:          11 : 50 (14.8%)      II                      (100%)     (0%)     
##                   10 < 14 < 18              12 : 42 (12.4%)      II                                          
##                   IQR (CV) : 3 (0.1)        13 : 52 (15.3%)      III                                         
##                                             14 : 48 (14.2%)      II                                          
##                                             15 : 60 (17.7%)      III                                         
##                                             16 : 33 ( 9.7%)      I                                           
##                                             17 : 39 (11.5%)      II                                          
##                                             18 :  8 ( 2.4%)                                                  
## 
## 2    race         1. white                  239 (72.6%)          IIIIIIIIIIIIII          329        10       
##      [factor]     2. Other                   90 (27.4%)          IIIII                   (97.05%)   (2.95%)  
## 
## 3    sex          1. Female                 181 (53.4%)          IIIIIIIIII              339        0        
##      [factor]     2. Male                   158 (46.6%)          IIIIIIIII               (100%)     (0%)     
## 
## 4    ses          1. AB                      93 (27.4%)          IIIII                   339        0        
##      [factor]     2. C                      203 (59.9%)          IIIIIIIIIII             (100%)     (0%)     
##                   3. DE                      43 (12.7%)          II                                          
## 
## 5    sleeping     1. no                     321 (96.7%)          IIIIIIIIIIIIIIIIIII     332        7        
##      [factor]     2. yes                     11 ( 3.3%)                                  (97.94%)   (2.06%)  
## 
## 6    premature    1. no                     293 (88.0%)          IIIIIIIIIIIIIIIII       333        6        
##      [factor]     2. yes                     40 (12.0%)          II                      (98.23%)   (1.77%)  
## 
## 7    smoking      1. no                     263 (78.0%)          IIIIIIIIIIIIIII         337        2        
##      [factor]     2. yes                     74 (22.0%)          IIII                    (99.41%)   (0.59%)  
## 
## 8    alcohol      1. no                     301 (89.6%)          IIIIIIIIIIIIIIIII       336        3        
##      [factor]     2. yes                     35 (10.4%)          II                      (99.12%)   (0.88%)  
## 
## 9    sdq_risk     1. no                     306 (90.3%)          IIIIIIIIIIIIIIIIII      339        0        
##      [factor]     2. yes                     33 ( 9.7%)          I                       (100%)     (0%)     
## 
## 10   adhd         1. no                     329 (97.0%)          IIIIIIIIIIIIIIIIIII     339        0        
##      [factor]     2. yes                     10 ( 2.9%)                                  (100%)     (0%)     
## -------------------------------------------------------------------------------------------------------------

22.1 Statistical model for Table 5

Descriptive analysis for all predictors included

## Data Frame Summary  
## base_uso  
## Dimensions: 339 x 10  
## Duplicates: 142  
## 
## -------------------------------------------------------------------------------------------------------------
## No   Variable     Stats / Values            Freqs (% of Valid)   Graph                   Valid      Missing  
## ---- ------------ ------------------------- -------------------- ----------------------- ---------- ---------
## 1    age          Mean (sd) : 13.9 (2.1)    10 :  7 ( 2.1%)                              339        0        
##      [numeric]    min < med < max:          11 : 50 (14.8%)      II                      (100%)     (0%)     
##                   10 < 14 < 18              12 : 42 (12.4%)      II                                          
##                   IQR (CV) : 3 (0.1)        13 : 52 (15.3%)      III                                         
##                                             14 : 48 (14.2%)      II                                          
##                                             15 : 60 (17.7%)      III                                         
##                                             16 : 33 ( 9.7%)      I                                           
##                                             17 : 39 (11.5%)      II                                          
##                                             18 :  8 ( 2.4%)                                                  
## 
## 2    race         1. white                  239 (72.6%)          IIIIIIIIIIIIII          329        10       
##      [factor]     2. Other                   90 (27.4%)          IIIII                   (97.05%)   (2.95%)  
## 
## 3    sex          1. Female                 181 (53.4%)          IIIIIIIIII              339        0        
##      [factor]     2. Male                   158 (46.6%)          IIIIIIIII               (100%)     (0%)     
## 
## 4    ses          1. AB                      93 (27.4%)          IIIII                   339        0        
##      [factor]     2. C                      203 (59.9%)          IIIIIIIIIII             (100%)     (0%)     
##                   3. DE                      43 (12.7%)          II                                          
## 
## 5    sleeping     1. no                     321 (96.7%)          IIIIIIIIIIIIIIIIIII     332        7        
##      [factor]     2. yes                     11 ( 3.3%)                                  (97.94%)   (2.06%)  
## 
## 6    premature    1. no                     293 (88.0%)          IIIIIIIIIIIIIIIII       333        6        
##      [factor]     2. yes                     40 (12.0%)          II                      (98.23%)   (1.77%)  
## 
## 7    smoking      1. no                     263 (78.0%)          IIIIIIIIIIIIIII         337        2        
##      [factor]     2. yes                     74 (22.0%)          IIII                    (99.41%)   (0.59%)  
## 
## 8    alcohol      1. no                     301 (89.6%)          IIIIIIIIIIIIIIIII       336        3        
##      [factor]     2. yes                     35 (10.4%)          II                      (99.12%)   (0.88%)  
## 
## 9    sdq_risk     1. no                     306 (90.3%)          IIIIIIIIIIIIIIIIII      339        0        
##      [factor]     2. yes                     33 ( 9.7%)          I                       (100%)     (0%)     
## 
## 10   adhd         1. no                     329 (97.0%)          IIIIIIIIIIIIIIIIIII     339        0        
##      [factor]     2. yes                     10 ( 2.9%)                                  (100%)     (0%)     
## -------------------------------------------------------------------------------------------------------------

The VIF was then computed for each model.

##                            GVIF Df GVIF^(1/(2*Df))
## age                        1.03  1            1.02
## race                       1.11  1            1.05
## relevel(sex, ref = "Male") 1.08  1            1.04
## ses                        1.29  2            1.07
## sleeping                   1.17  1            1.08
## premature                  1.02  1            1.01
## smoking                    1.36  1            1.17
## alcohol                    1.15  1            1.07
## sdq_risk                   1.16  1            1.08
##                            GVIF Df GVIF^(1/(2*Df))
## age                        1.06  1            1.03
## race                       1.11  1            1.06
## relevel(sex, ref = "Male") 1.03  1            1.02
## ses                        1.47  2            1.10
## sleeping                   1.27  1            1.13
## premature                  1.03  1            1.02
## smoking                    1.45  1            1.21
## alcohol                    1.11  1            1.06
## sdq_risk                   1.10  1            1.05
## 
## Call:
## glm(formula = risk_cefaleia_resources ~ age + race + relevel(sex, 
##     ref = "Male") + ses + sleeping + premature + smoking + alcohol + 
##     sdq_risk, family = "binomial", data = base_uso)
## 
## Deviance Residuals: 
##    Min      1Q  Median      3Q     Max  
## -1.327  -0.404  -0.282  -0.208   2.897  
## 
## Coefficients:
##                                  Estimate Std. Error z value Pr(>|z|)    
## (Intercept)                        -0.581      1.685   -0.34  0.73045    
## age                                -0.197      0.121   -1.63  0.10381    
## raceOther                           0.126      0.543    0.23  0.81661    
## relevel(sex, ref = "Male")Female    1.122      0.523    2.15  0.03194 *  
## sesC                               -0.680      0.527   -1.29  0.19738    
## sesDE                              -0.396      0.791   -0.50  0.61657    
## sleepingyes                         0.166      0.996    0.17  0.86794    
## prematureyes                        0.464      0.623    0.75  0.45582    
## smokingyes                         -0.131      0.648   -0.20  0.83951    
## alcoholyes                         -0.046      0.754   -0.06  0.95133    
## sdq_riskyes                         2.019      0.550    3.67  0.00024 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 169.54  on 313  degrees of freedom
## Residual deviance: 145.60  on 303  degrees of freedom
##   (25 observations deleted due to missingness)
## AIC: 167.6
## 
## Number of Fisher Scoring iterations: 6
## 
## Call:
## glm(formula = risk_cefaleia_vulnerability ~ age + race + relevel(sex, 
##     ref = "Male") + ses + sleeping + premature + smoking + alcohol + 
##     sdq_risk, family = "binomial", data = base_uso)
## 
## Deviance Residuals: 
##    Min      1Q  Median      3Q     Max  
## -1.133  -0.342  -0.255  -0.160   2.941  
## 
## Coefficients:
##                                  Estimate Std. Error z value Pr(>|z|)   
## (Intercept)                       -2.7223     1.8668   -1.46   0.1448   
## age                               -0.0430     0.1347   -0.32   0.7495   
## raceOther                         -0.7166     0.7249   -0.99   0.3229   
## relevel(sex, ref = "Male")Female   1.0047     0.6104    1.65   0.0998 . 
## sesC                              -1.0719     0.6603   -1.62   0.1045   
## sesDE                             -1.0482     0.9424   -1.11   0.2661   
## sleepingyes                        1.2047     0.9814    1.23   0.2196   
## prematureyes                       0.0366     0.8146    0.04   0.9642   
## smokingyes                         1.7254     0.6475    2.66   0.0077 **
## alcoholyes                        -0.4620     0.8413   -0.55   0.5829   
## sdq_riskyes                        0.0157     0.8529    0.02   0.9853   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 126.43  on 313  degrees of freedom
## Residual deviance: 112.00  on 303  degrees of freedom
##   (25 observations deleted due to missingness)
## AIC: 134
## 
## Number of Fisher Scoring iterations: 6

some manual computations were then performed, just to ensure the reliability of the results

## 
## Call:
## glm(formula = risk_cefaleia_vulnerability ~ sdq_risk, family = binomial, 
##     data = base_uso)
## 
## Deviance Residuals: 
##    Min      1Q  Median      3Q     Max  
## -0.354  -0.317  -0.317  -0.317   2.456  
## 
## Coefficients:
##             Estimate Std. Error z value Pr(>|z|)    
## (Intercept)   -2.965      0.265  -11.20   <2e-16 ***
## sdq_riskyes    0.224      0.776    0.29     0.77    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 134.89  on 338  degrees of freedom
## Residual deviance: 134.81  on 337  degrees of freedom
## AIC: 138.8
## 
## Number of Fisher Scoring iterations: 5
## $data
##          Outcome
## Predictor yes  no Total
##     1       2  15    17
##     0      31 291   322
##     Total  33 306   339
## 
## $measure
##          odds ratio with 95% C.I.
## Predictor estimate lower upper
##         1     1.00    NA    NA
##         0     1.25 0.273  5.73
## 
## $p.value
##          two-sided
## Predictor midp.exact fisher.exact chi.square
##         1         NA           NA         NA
##         0      0.731        0.676      0.772
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "Unconditional MLE & normal approximation (Wald) CI"

22.2 Table 5

  risk cefaleia resources risk cefaleia
vulnerability
Predictors Odds Ratios CI Statistic p Odds Ratios CI Statistic p
(Intercept) 0.560 0.020 – 15.565 -0.345 0.730 0.066 0.001 – 2.388 -1.458 0.145
age 0.821 0.641 – 1.035 -1.627 0.104 0.958 0.731 – 1.248 -0.319 0.749
race [Other] 1.134 0.363 – 3.162 0.232 0.817 0.488 0.095 – 1.771 -0.989 0.323
relevel(sex, ref =
“Male”) [relevel(sex, ref
= “Male”)Female] 		3.071 1.164 – 9.341 2.145 0.032 2.731 0.888 – 10.334 1.646 0.100
ses [C] 0.507 0.177 – 1.438 -1.289 0.197 0.342 0.088 – 1.231 -1.623 0.105
ses [DE] 0.673 0.121 – 2.906 -0.501 0.617 0.351 0.042 – 1.931 -1.112 0.266
sleeping [yes] 1.180 0.128 – 7.050 0.166 0.868 3.336 0.378 – 20.187 1.228 0.220
premature [yes] 1.591 0.409 – 4.970 0.746 0.456 1.037 0.151 – 4.287 0.045 0.964
smoking [yes] 0.877 0.222 – 2.934 -0.203 0.840 5.615 1.572 – 20.875 2.665 0.008
alcohol [yes] 0.955 0.181 – 3.745 -0.061 0.951 0.630 0.088 – 2.759 -0.549 0.583
sdq_risk [yes] 7.532 2.514 – 22.364 3.670 <0.001 1.016 0.137 – 4.495 0.018 0.985
Observations 314 314
R2 Tjur 0.101 0.051

23 Post hoc Sensitivity analysis

Check distribution of the results

## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  base_uso$cefaleia_lifetime and base_uso$sexo
## X-squared = 16, df = 1, p-value = 6e-05
## # A tibble: 4 x 3
##   cefaleia_lifetime sex        n
##               <dbl> <fct>  <int>
## 1                 0 Female     9
## 2                 0 Male      31
## 3                 1 Female   172
## 4                 1 Male     127

## Response z_resources :
## 
## Call:
## lm(formula = z_resources ~ age + race + relevel(sex, ref = "Male") + 
##     ses + sleeping + premature + smoking + alcohol + sdq_risk, 
##     data = base_uso)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -3.227 -0.587  0.028  0.681  3.027 
## 
## Coefficients:
##                                  Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                       0.04140    0.39399    0.11    0.916    
## age                               0.00489    0.02755    0.18    0.859    
## raceOther                         0.06061    0.12633    0.48    0.632    
## relevel(sex, ref = "Male")Female -0.02438    0.11198   -0.22    0.828    
## sesC                              0.01326    0.13150    0.10    0.920    
## sesDE                            -0.06567    0.19768   -0.33    0.740    
## sleepingyes                      -0.56626    0.30831   -1.84    0.067 .  
## prematureyes                     -0.02673    0.17277   -0.15    0.877    
## smokingyes                        0.07950    0.15077    0.53    0.598    
## alcoholyes                       -0.13179    0.19322   -0.68    0.496    
## sdq_riskyes                      -0.84860    0.19175   -4.43  1.3e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.978 on 303 degrees of freedom
##   (25 observations deleted due to missingness)
## Multiple R-squared:  0.0818, Adjusted R-squared:  0.0515 
## F-statistic:  2.7 on 10 and 303 DF,  p-value: 0.00348
## 
## 
## Response cefaleia_lifetime :
## 
## Call:
## lm(formula = cefaleia_lifetime ~ age + race + relevel(sex, ref = "Male") + 
##     ses + sleeping + premature + smoking + alcohol + sdq_risk, 
##     data = base_uso)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -0.9692  0.0203  0.0953  0.1510  0.3359 
## 
## Coefficients:
##                                   Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                       0.764740   0.124378    6.15  2.5e-09 ***
## age                              -0.001712   0.008696   -0.20  0.84404    
## raceOther                        -0.012181   0.039882   -0.31  0.76025    
## relevel(sex, ref = "Male")Female  0.128475   0.035350    3.63  0.00033 ***
## sesC                              0.107018   0.041513    2.58  0.01041 *  
## sesDE                             0.095518   0.062405    1.53  0.12691    
## sleepingyes                       0.116064   0.097329    1.19  0.23400    
## prematureyes                      0.058610   0.054541    1.07  0.28341    
## smokingyes                        0.014968   0.047596    0.31  0.75338    
## alcoholyes                       -0.061100   0.060997   -1.00  0.31730    
## sdq_riskyes                       0.000559   0.060532    0.01  0.99264    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.309 on 303 degrees of freedom
##   (25 observations deleted due to missingness)
## Multiple R-squared:  0.0715, Adjusted R-squared:  0.0409 
## F-statistic: 2.33 on 10 and 303 DF,  p-value: 0.0116
## 
## Type II MANOVA Tests: Pillai test statistic
##                            Df test stat approx F num Df den Df  Pr(>F)    
## age                         1    0.0002     0.03      2    302  0.9680    
## race                        1    0.0010     0.15      2    302  0.8599    
## relevel(sex, ref = "Male")  1    0.0418     6.58      2    302  0.0016 ** 
## ses                         2    0.0226     1.74      4    606  0.1406    
## sleeping                    1    0.0146     2.24      2    302  0.1084    
## premature                   1    0.0038     0.58      2    302  0.5615    
## smoking                     1    0.0013     0.20      2    302  0.8176    
## alcohol                     1    0.0052     0.79      2    302  0.4558    
## sdq_risk                    1    0.0610     9.81      2    302 7.4e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## Response z_vulnerability :
## 
## Call:
## lm(formula = z_vulnerability ~ age + race + relevel(sex, ref = "Male") + 
##     ses + sleeping + premature + smoking + alcohol + sdq_risk, 
##     data = base_uso)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -2.4180 -0.6514 -0.0209  0.7108  2.5867 
## 
## Coefficients:
##                                  Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                      -0.15601    0.38990   -0.40     0.69    
## age                               0.00869    0.02726    0.32     0.75    
## raceOther                        -0.08078    0.12502   -0.65     0.52    
## relevel(sex, ref = "Male")Female  0.05631    0.11081    0.51     0.61    
## sesC                             -0.16915    0.13013   -1.30     0.19    
## sesDE                            -0.12351    0.19563   -0.63     0.53    
## sleepingyes                       0.43878    0.30511    1.44     0.15    
## prematureyes                     -0.05906    0.17097   -0.35     0.73    
## smokingyes                        0.12272    0.14920    0.82     0.41    
## alcoholyes                        0.20814    0.19121    1.09     0.28    
## sdq_riskyes                       0.85658    0.18975    4.51  9.1e-06 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.968 on 303 degrees of freedom
##   (25 observations deleted due to missingness)
## Multiple R-squared:  0.0942, Adjusted R-squared:  0.0643 
## F-statistic: 3.15 on 10 and 303 DF,  p-value: 0.000748
## 
## 
## Response cefaleia_lifetime :
## 
## Call:
## lm(formula = cefaleia_lifetime ~ age + race + relevel(sex, ref = "Male") + 
##     ses + sleeping + premature + smoking + alcohol + sdq_risk, 
##     data = base_uso)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -0.9692  0.0203  0.0953  0.1510  0.3359 
## 
## Coefficients:
##                                   Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                       0.764740   0.124378    6.15  2.5e-09 ***
## age                              -0.001712   0.008696   -0.20  0.84404    
## raceOther                        -0.012181   0.039882   -0.31  0.76025    
## relevel(sex, ref = "Male")Female  0.128475   0.035350    3.63  0.00033 ***
## sesC                              0.107018   0.041513    2.58  0.01041 *  
## sesDE                             0.095518   0.062405    1.53  0.12691    
## sleepingyes                       0.116064   0.097329    1.19  0.23400    
## prematureyes                      0.058610   0.054541    1.07  0.28341    
## smokingyes                        0.014968   0.047596    0.31  0.75338    
## alcoholyes                       -0.061100   0.060997   -1.00  0.31730    
## sdq_riskyes                       0.000559   0.060532    0.01  0.99264    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.309 on 303 degrees of freedom
##   (25 observations deleted due to missingness)
## Multiple R-squared:  0.0715, Adjusted R-squared:  0.0409 
## F-statistic: 2.33 on 10 and 303 DF,  p-value: 0.0116
## 
## Type II MANOVA Tests: Pillai test statistic
##                            Df test stat approx F num Df den Df Pr(>F)    
## age                         1    0.0005     0.08      2    302 0.9262    
## race                        1    0.0016     0.24      2    302 0.7884    
## relevel(sex, ref = "Male")  1    0.0419     6.60      2    302 0.0016 ** 
## ses                         2    0.0294     2.26      4    606 0.0615 .  
## sleeping                    1    0.0104     1.59      2    302 0.2056    
## premature                   1    0.0045     0.68      2    302 0.5091    
## smoking                     1    0.0024     0.37      2    302 0.6944    
## alcohol                     1    0.0079     1.21      2    302 0.3009    
## sdq_risk                    1    0.0635    10.25      2    302  5e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
  • End of manuscript -