Search is not available for this dataset
text
stringlengths 1
134k
| id
stringlengths 11
16
| article_id
stringlengths 8
10
| section_title
stringlengths 1
1.35k
| educational_score
float64 0.52
5.16
| domain
stringclasses 3
values | document_type
stringclasses 4
values | domain_scores
listlengths 3
3
| document_type_scores
listlengths 4
4
| language
stringclasses 38
values | language_score
float64 0
1
|
|---|---|---|---|---|---|---|---|---|---|---|
Further, Brazilian students were more likely to have a diagnosed mental disorder than German students ( P < 0.001). The most commonly diagnosed mental disorders reported by Brazilian students were anxiety ( n = 568, 35.2%), depression ( n = 433, 26.8%), other mental disorders ( n = 157, 9.7%) and attention-deficit hyperactivity disorder ( n = 118, 7.3%). German students reported depression ( n = 612, 31.1%), anxiety ( n = 437, 22.2%), other mental disorders ( n = 333, 16.8%) and eating disorders ( n = 215, 10.9%) as the most commonly diagnosed mental disorders. Finally, Brazilian students were more likely to be fully vaccinated against COVID-19 than German students ( P < 0.001).
|
39494847_p22
|
39494847
|
Results
| 3.601707 |
biomedical
|
Study
|
[
0.9969573020935059,
0.0004264063318260014,
0.0026163123548030853
] |
[
0.999184787273407,
0.0005412483587861061,
0.00023597292602062225,
0.00003798290345002897
] |
en
| 0.999994 |
There was a statistically significant difference between Brazilian and German students, with the former presenting higher levels of depressive symptoms than the latter [ MD = 4.943, 95% CI Bca (4.595, 5.296)] ( Table 2 ). The covariates age and gender had significant effects on the PHQ-9 sum scores model . Further, significantly more Brazilian students reported clinically relevant depressive symptoms [95% CI BCa (−0.260, −0.219)], and they had a higher prevalence of suicidal thoughts [95% CI BCa (−0.170, −0.122)] compared with German students ( Table 2 ). Table 2 Means, standard deviations and confidence intervals for mental health measures and social and emotional characteristics of Brazilian and German samples during COVID-19 Variable Brazilian students German students Test P -value η 2 partial/ϕ M a (s.d.) 95% CI M a (s.d.) 95% CI Depressive symptoms (PHQ-9) 12.85 (7.40) 12.55–13.15 8.35 (5.52) 8.20–8.49 F = 1002.293 <0.001 η 2 = 0.112 Alcohol use (AUDIT-C) 2.80 (2.52) b 2.70–2.91 2.91 (2.31) 2.27–2.35 F = 2.912 0.088 η 2 = 0.000 Drug or substance consumption 2.13 (1.28) 1.99, 2.28 1.74 (1.03) 1.67, 1.81 F = 18.940 <0.001 η 2 = 0.017 Loneliness (UCLA-3) 5.25 (2.35) 5.16–5.34 5.75 (1.86) 5.70–5.80 F = 37.848 <0.001 η 2 = 0.005 Self-efficacy (GSE) 2.82 (0.59) 2.80–2.85 2.77 (0.46) 2.76–2.78 F = 0.077 0.782 η 2 = 0.000 Perceived stress (PSS-4) 7.81 (2.52) c 7.70–7.92 7.23 (3.20) 7.15–7.31 F = 60.924 <0.001 η 2 = 0.008 Social support (ESSI) 18.04 (5.15) 17.93–18.25 20.71 (4.10) 20.60–20.83 F = 539.367 <0.001 η 2 = 0.064 Resilience (BRS) 2.75 (0.87) d 2.72–2.79 3.04 (0.77) 3.02–3.06 F (1, 864) = 270.633 <0.001 η 2 = 0.033 Attitude towards vaccination 4.92 (0.44) 4.90 (4.93) 4.63 (0.78) 4.61–4.65 F = 270.435 <0.001 η 2 = 0.033 n (%) n (%) Suicidal thoughts e 808 (33.2) NA 1075 (19.6) NA χ 2 = 169.880 <0.001 ϕ = 0.147 Clinically relevant depressive symptoms f 1488 (61.1) NA 1936 (35.4) NA χ 2 = 453.399 <0.001 ϕ = 0.239 PHQ-9, Patient Health Questionnaire-9; AUDIT-C, Alcohol Hazardous Use Subscale of the Alcohol Use Disorders Identification Test; ESSI, ENRICHD Social Support Instrument; GSE, General Self-Efficacy Scale; PSS-4, Perceived Stress Scale; UCLA, Three-Item Loneliness Scale. NA, not applicable. a. Scores adjusted for age and gender; bootstrapping sample. b. Reduced sample size of n = 2421 participants owing to missing data. c. Reduced sample size of n = 1709 participants owing to missing data. d. Reduced sample size of n = 2394 participants owing to missing data. e. Suicidal thoughts based on item 9 score ≥1 on PHQ-9. f. Clinically relevant depressive symptoms based on having a score of 10 or more on the PHQ-9.
|
39494847_p23
|
39494847
|
Depressive symptoms
| 4.223538 |
biomedical
|
Study
|
[
0.9991772770881653,
0.00039347793790511787,
0.00042922384454868734
] |
[
0.9992843270301819,
0.0002045000292127952,
0.0004511613806243986,
0.00006005321483826265
] |
en
| 0.999996 |
Regarding hazardous alcohol use (hazardous subscale of AUDIT-C), there was no statistically significant difference between Brazilian and German students ( P = 0.088). The covariate gender presented a significant effect on the model ( P < 0.001; η 2 partial = 0.022), whereas age did not ( P = 0.053).
|
39494847_p24
|
39494847
|
Alcohol and drug or substance consumption
| 2.628692 |
biomedical
|
Study
|
[
0.9695422053337097,
0.0009402713039889932,
0.02951761707663536
] |
[
0.9985664486885071,
0.0011046187719330192,
0.0002671578258741647,
0.00006184187805047259
] |
en
| 0.999998 |
Concerning drug or substance consumption, n = 2074 (85.1%) of Brazilian participants reported having never consumed any drug or substance, as did n = 4720 (86.2%) of German participants, with no statistically significant difference found between the two samples [χ 2 = 1.748, P = 0.186]. Among those who reported any drug consumption, Brazilian students had a higher frequency of drug or substance consumption than German participants [ MD = 0.323, 95% CI BCa (0.154, 0.507)]. Both covariates, age and gender, had significant effects on the model (both P < 0.05, η 2 partial = 0.016 and η 2 partial = 0.007, respectively).
|
39494847_p25
|
39494847
|
Alcohol and drug or substance consumption
| 4.025131 |
biomedical
|
Study
|
[
0.9980819225311279,
0.00036529035423882306,
0.0015527845825999975
] |
[
0.9996312856674194,
0.00015711340529378504,
0.00018047445337288082,
0.00003118037784588523
] |
en
| 0.999997 |
German students presented significantly higher levels of loneliness (UCLA-3) compared with Brazilian students [ MD = 0.315, 95% CI Bca (0.203, 0.423)]. The covariate age had a significant effect on the model , whereas gender did not ( P = 0.508). Self-efficacy (GSE) sum scores did not differ between Brazilian and German students ( P = 0.782). Age and gender had significant effects on the model . Brazilian students had significantly higher sum scores for perceived stress (PSS-4) than German students [ MD = 0.692, 95% CI Bca (0.532, 0.851)]. Age and gender had significant effects on the model . Social support (ESSI) sum scores were significantly higher among German students compared with Brazilian students [ MD = 2.611, 95% CI Bca (2.391, 2.832)]. Gender had a significant effect on the model , but age did not ( P = 0.711). Finally, resilience (BRS) sum scores were significantly higher among German students compared with Brazilian students [ MD = 0.336, 95% CI Bca (0.297, 0.373)]. Both covariates, age and gender, had significant effects on the model .
|
39494847_p26
|
39494847
|
Social and emotional aspects
| 3.990587 |
biomedical
|
Study
|
[
0.9612981677055359,
0.0006034913822077215,
0.03809831663966179
] |
[
0.9993656277656555,
0.000293590099317953,
0.0003104435745626688,
0.00003033396387763787
] |
en
| 0.999996 |
Brazilian students presented statistically significantly more favourable attitudes towards vaccination in general compared with German students [ MD = 0.290, 95% CI BCa (0.261, 0.318)]. Neither of the covariates (age or gender) had a significant effect on the model (both P > 0.05).
|
39494847_p27
|
39494847
|
Attitude towards vaccination
| 2.79236 |
biomedical
|
Study
|
[
0.9920318126678467,
0.0008291830308735371,
0.007139001972973347
] |
[
0.9987568855285645,
0.0008476726943627,
0.00033159981830976903,
0.00006380544073181227
] |
en
| 0.999996 |
This study investigated differences between university students from Brazil and Germany regarding mental health outcomes, social and emotional aspects, and attitudes towards vaccination. As hypothesised, Brazilian students reported more depressive symptoms and suicidal thoughts than German students. Despite the between-sample differences in sociodemographic characteristics, we considered our Brazilian and German groups to be representative of their respective university student populations, with differences related to each country's geopolitical and cultural characteristics. The sample sizes were substantial, and the results for each country were comparable with findings of similar studies conducted in both countries. 16 , 17 However, the diversity within each country should be considered when evaluating the results and the extent to which they can be generalised, as in both samples the universities were located in one region/federal state.
|
39494847_p28
|
39494847
|
Discussion
| 3.880174 |
biomedical
|
Study
|
[
0.9949491024017334,
0.00051214947598055,
0.004538773559033871
] |
[
0.9996210336685181,
0.00014263042248785496,
0.00020201411098241806,
0.000034330529160797596
] |
en
| 0.999998 |
In Brazil, a Latin American middle-income country, there was a compartmentalised response to the COVID-19 crisis, and each subnational government managed its own health policies, with an important role for governors, with little or no support from the federal government – which eventually produced horizontal and vertical competition for scarce supplies in the fight against COVID-19. 31 , 32 Therefore, social distancing measures, use of masks and testing could vary significantly across the country.
|
39494847_p29
|
39494847
|
Discussion
| 1.730165 |
other
|
Other
|
[
0.27811387181282043,
0.0021516643464565277,
0.7197344303131104
] |
[
0.023385951295495033,
0.9752156734466553,
0.0010518706403672695,
0.0003465599729679525
] |
en
| 0.999997 |
Germany, a European high-income country, implemented a mitigation strategy to curb the spread of COVID-19. Lockdowns and strict rules regarding social distancing were imposed nationwide at different points in time. 11 , 16 Although these restrictions were alternated with relaxation of control measures, face masks were mandatory at different time points across German regions. 33 Germany was also among the countries that implemented mandatory COVID-19 certification for proof of vaccination status (at least two shots) to enter restaurants, stores, nightclubs and gyms, among other public services, and also to travel internationally. 34
|
39494847_p30
|
39494847
|
Discussion
| 1.727437 |
other
|
Other
|
[
0.34343481063842773,
0.002939238678663969,
0.6536259651184082
] |
[
0.027249876409769058,
0.9713366627693176,
0.0009232982411049306,
0.0004901390639133751
] |
en
| 0.999995 |
In both Germany and Brazil, similar to other countries, misinformation regarding COVID-19 was an important issue that may have negatively affected prevention measures. 7 , 35 On the other hand, trustworthy sources of information may positively affect how people respond psychologically and behaviourally to crises. 36 Accordingly, an immediate and consistent response can be considered a powerful public health prevention tool for health crises such as the COVID-19 pandemic worldwide.
|
39494847_p31
|
39494847
|
Discussion
| 1.981025 |
biomedical
|
Other
|
[
0.7746743559837341,
0.0026670400984585285,
0.22265860438346863
] |
[
0.03156416490674019,
0.9600914716720581,
0.00784598384052515,
0.0004983915132470429
] |
en
| 0.999997 |
Considering the sample as a whole, n = 3424 (43.3%) participants reported clinically relevant depressive symptoms. Although these results do not represent clinical diagnoses, as PHQ-9 is a self-report instrument, they corroborate previous literature indicating that university students worldwide are at increased risk of developing depression. 1 – 4 , 15 , 37
|
39494847_p32
|
39494847
|
Depressive symptoms
| 2.665636 |
biomedical
|
Study
|
[
0.9971880316734314,
0.0008934072684496641,
0.0019185625715181231
] |
[
0.9980746507644653,
0.0015197317115962505,
0.0003020340809598565,
0.00010355051199439913
] |
en
| 0.999997 |
Brazilian students reported higher levels of depressive symptoms and had a greater prevalence of clinically relevant depressive symptoms and suicidal thoughts than German students. The effect sizes observed between the two groups indicated a higher risk among Brazilian students of developing depression compared with German students. The prevalence of clinically relevant depressive symptoms among Brazilian students in our sample was higher compared with those reported in France (16.1%), 38 the USA (48.14%), 39 , 40 China (22.0%) and Ethiopia (46.3%) 41 and in the Brazilian general population (41.9%). 42 The prevalence of clinically relevant depressive symptoms among German students was similar to the pooled prevalence (34%) of depressive symptoms among university students during the pandemic worldwide. 1 , 2 Before the pandemic, prevalences of 6.1 to 65.5% for depressive symptoms and 3.9 to 49.1% for suicidal ideation among university students internationally were reported. 3 An increased risk of suicide among university students during the COVID-19 pandemic has been reported in both Brazil 43 and Germany, 16 as well as higher numbers of students reporting suicidal thoughts in both countries. Our results may be related to the higher likelihood of Brazilian students having received a diagnosis of any mental disorder (36.5%) compared with German students (22.6%), as this has been found to be a predictor of depressive symptoms among university students. 16 , 17
|
39494847_p33
|
39494847
|
Depressive symptoms
| 4.089781 |
biomedical
|
Study
|
[
0.9991484880447388,
0.00035118040977977216,
0.0005003175465390086
] |
[
0.9995385408401489,
0.00012894369137939066,
0.0002900516556110233,
0.000042475272493902594
] |
en
| 0.999997 |
By contrast, although German students had a higher prevalence of diagnosed depression disorder (31.1%) compared with Brazilian students (26.8%), their levels of self-reported depressive symptoms and suicidal thoughts were lower than those of their Brazilian counterparts. Considering that 47.7% ( n = 591) of the German students who disclosed having been diagnosed with any mental disorder were receiving any form of treatment (i.e. psychotherapy and/or medication), 16 compared with 63.6% ( n = 564) of the Brazilians, 17 we hypothesise that this could be pandemic-related, as previous research has indicated that countries where governments implemented stringent policies promptly (which was the case in Germany) had lower prevalence of clinically significant depressive symptoms. 10 As we assessed lifetime diagnoses, some participants may have no longer been in need of treatment by the time the survey was conducted. In addition, the high number of deaths by COVID-19 in Brazil may indicate that Brazilian students experienced grief more frequently; this could be related to the prevalence of depressive symptoms and suicidal thoughts in this sample. On the other hand, bereavement-related emotions are often mistaken for symptoms of major depressive disorder. 44 Longitudinal studies could help us to understand this phenomenon better.
|
39494847_p34
|
39494847
|
Depressive symptoms
| 4.080789 |
biomedical
|
Study
|
[
0.9988755583763123,
0.0003852269728668034,
0.0007391975377686322
] |
[
0.999491810798645,
0.0001665826712269336,
0.0003022706077899784,
0.00003930164530174807
] |
en
| 0.999997 |
Although levels of hazardous alcohol use did not differ between the two groups, a higher frequency of drug or substance consumption was observed among Brazilian students compared with German students. Changes in alcohol consumption among university students during the COVID-19 pandemic have been reported, with studies finding that it increased or decreased, although some studies reported no changes. 15 , 45 – 47 There have also been heterogeneous results regarding drug or substance consumption during COVID-19. Increases in drug or substance use during the COVID-19 pandemic have been reported worldwide, 48 as have decreases or no changes. 49 Returning to live with family was associated with less alcohol consumption. 45 Both alcohol and drug or substance use are associated with poorer mental health outcomes, such as depressive symptoms, and social and emotional aspects, such as boredom. 48 , 49 Nevertheless, attributing causes to the differences between the two samples would require more comprehensive measures and longitudinal studies.
|
39494847_p35
|
39494847
|
Alcohol and drug or substance consumption
| 3.927403 |
biomedical
|
Study
|
[
0.9961218237876892,
0.0003011802036780864,
0.003576976712793112
] |
[
0.9894546270370483,
0.0004493323212955147,
0.010027207434177399,
0.00006882124580442905
] |
en
| 0.999998 |
Although German students experienced higher levels of perceived loneliness than Brazilian students, they also reported higher levels of perceived social support. The higher levels of loneliness among German students could be related to their sociodemographic characteristics, as they were more likely to be single, living alone and not a parent compared with Brazilian students. These higher levels of loneliness could also be related to the more frequent and extended periods of lockdowns in Germany compared with Brazil. Although loneliness has been identified as a predictor of suicidal ideation and behavior 50 and associated with more depressive symptoms, 51 we believe that social support worked as a protective factor against the impact of loneliness on mental health outcomes 52 among German students.
|
39494847_p36
|
39494847
|
Social and emotional aspects
| 2.618652 |
biomedical
|
Study
|
[
0.8827752470970154,
0.0007640182739123702,
0.11646070331335068
] |
[
0.9951687455177307,
0.004348208196461201,
0.00041959184454753995,
0.00006338902312563732
] |
en
| 0.999995 |
Similarly, the level of perceived loneliness among Brazilians could be understood in light of the sociodemographic characteristics of that sample (i.e. living with others, having a relationship, being a parent) and the more relaxed social restriction measures applied in Brazil compared with Germany. However, availability of social support indicates the extent to which an individual perceives that they can rely on their social relationships and feels valued and cared about. 27 This means that one may feel connected but not supported. Therefore, the lower levels of perceived social support among Brazilians despite the lower levels of loneliness compared with Germans could be related to the higher levels of depressive symptoms 51 in the Brazilian sample (e.g. helplessness). Moreover, it could reflect aspects of the culture and the psychosocial implications of socioeconomic inequality, which increased during the COVID-19 pandemic, affecting interpersonal relations, domestic and academic life, and work–life balance. 9 , 37 , 42 Finally, these results illustrate the relevance of assessing both loneliness and social support constructs. 52
|
39494847_p37
|
39494847
|
Social and emotional aspects
| 3.724073 |
biomedical
|
Study
|
[
0.9108433127403259,
0.000541000859811902,
0.08861581981182098
] |
[
0.9981270432472229,
0.0010607318254187703,
0.000767820980399847,
0.000044391326810000464
] |
en
| 0.999996 |
In both samples, perceived social support was found to be a joint protective factor for resilience and self-efficacy. 16 , 17 Although self-efficacy did not differ between the two samples, German students reported higher resilience levels compared with Brazilian students. As social support and resilience have been identified as protective factors against depressive symptoms 12 and suicidal thoughts, 5 these results could also reflect the differences observed in depressive symptoms reported above, as less perceived satisfactory support is associated with more suicidal thoughts among university students. 5
|
39494847_p38
|
39494847
|
Social and emotional aspects
| 3.227645 |
biomedical
|
Study
|
[
0.9690306782722473,
0.0006253584870137274,
0.03034389205276966
] |
[
0.998354971408844,
0.0008460916578769684,
0.0007527688867412508,
0.0000461734744021669
] |
en
| 0.999997 |
The higher levels of perceived stress among Brazilian students compared with German students may be related to their government's response to the COVID-19 crisis. The denial of the seriousness of the pandemic by the Brazilian Federal Government and its late response to the COVID-19 crisis led Brazil to become one of the countries with the highest number of deaths from COVID-19 worldwide. 31 , 32 Moreover, Brazil's political instability, social inequality and insecurity may affect students directly. For some Brazilian students, the shift to online teaching became a barrier to study continuation owing to lack of or poor equipment or internet access. German students also experienced negative effects of the pandemic and faced changes in their study and living conditions that included a shift to remote teaching, lack of social contact with other students and financial loss. 12 , 15 They were more likely to report being worried and thinking about the coronavirus than the German general population 35 and showed an increased prevalence of suicidal ideation compared with 2020 and 2021, 35 indicating long-term effects of the pandemic.
|
39494847_p39
|
39494847
|
Social and emotional aspects
| 3.018635 |
biomedical
|
Study
|
[
0.5479498505592346,
0.0007624790887348354,
0.4512876570224762
] |
[
0.991317629814148,
0.0069933063350617886,
0.0015874162781983614,
0.0001016395035549067
] |
en
| 0.999998 |
A more favourable opinion towards vaccinations, in general, was found among Brazilian students compared with German students. This is in line with the higher likelihood of Brazilian participants to be fully vaccinated against COVID-19 (i.e. 97.3%) compared with their German counterparts [93.2%; χ 2 = 55.075, P < 0.001, ϕ = 0.083]. Unlike the Brazilian university where the survey took place, in German universities, a full vaccination certificate was not required to access the university; this may have influenced the number of fully vaccinated students. Moreover, the survey was conducted earlier in Brazil than in Germany, and the percentages of the populations that were fully vaccinated were 74.35% in Brazil and 75.97% in Germany. 11
|
39494847_p40
|
39494847
|
Attitude towards vaccination
| 3.736282 |
biomedical
|
Study
|
[
0.9949305653572083,
0.0004149079613853246,
0.004654486197978258
] |
[
0.9994633793830872,
0.0003019179857801646,
0.00020335879526101053,
0.000031390325602842495
] |
en
| 0.999996 |
Brazil has a structured immunisation strategy within its public universal healthcare system 53 and rapidly increased vaccination against COVID-19 as soon as the vaccine was authorised in the country. 14 In Germany, vaccination is mostly provided by physicians in private practices. When a vaccine is officially recommended in the national immunisation schedule, the costs of childhood and adult vaccination are fully reimbursed by health insurance funds. Vaccination hesitancy varies across countries, 14 and more favourable opinions towards vaccinations in general have been reported to predict vaccination intention. 35 From a public health perspective, this implies a need to design vaccination campaigns that address contextual barriers and negative beliefs toward vaccination.
|
39494847_p41
|
39494847
|
Attitude towards vaccination
| 3.003021 |
biomedical
|
Other
|
[
0.9750391840934753,
0.0022235880605876446,
0.022737309336662292
] |
[
0.179176464676857,
0.775689423084259,
0.0443553626537323,
0.0007787939393892884
] |
en
| 0.999997 |
This study had several strengths. First, it had a large sample size that we considered to be representative of the respective countries. Second, we used identical measures for the data collection in both countries, ensuring the comparability of results. Third, we used validated, internationally comparable and widely used questionnaires. The study also had some limitations. First, owing to our exploratory approach and the nature of the data, we did not consider further hypotheses regarding different cultural aspects that might have influenced sociodemographic differences. Specifically, there was a statistically significant difference in age between the two samples, with German students being younger than Brazilian students. By contrast, Brazilian students were more likely to be doing a bachelor degree. This difference might be related to later access to university in Brazil compared with Germany, probably owing to educational differences. We addressed this limitation by controlling for age as a covariate in all analyses. Second, we used one item for screening illicit drug use, which was rephrased from the AUDIT-C questionnaire, and analysed the results descriptively. We acknowledge the limitation of not using a standardised measure for illicit substance use and suggest this could be included in future studies. Third, we assessed attitudes toward vaccination with a single item, which was analysed descriptively to characterise the samples according to the COVID-19 context in each country; future research could survey vaccination attitudes in more detail and evaluate how sociodemographic characteristics (including country of origin) and psychosocial and emotional aspects could affect attitudes towards vaccination. Further, the two time points of data collection were not identical, although they were very close; this should be considered when analysing the results. Nevertheless, regardless of time point, Brazil and Germany had different government measures regarding social distancing. Our results shed light on possible long-lasting effects of the COVID-19 pandemic on university students’ mental health and indicate a need to longitudinally assess students’ mental health status and provide mental healthcare tailored to them.
|
39494847_p42
|
39494847
|
Strengths and limitations
| 4.149447 |
biomedical
|
Study
|
[
0.9977572560310364,
0.000636701937764883,
0.0016060348134487867
] |
[
0.9994853734970093,
0.00014847192505840212,
0.00031490190303884447,
0.00005121172216604464
] |
en
| 0.999997 |
This study illustrates the negative impact of the COVID-19 pandemic on the mental health of university students in Brazil and Germany and highlights difficulties in coping, high levels of social and emotional distress and clinically relevant symptoms among students, in line with international findings. 1 – 8 , 12 , 15 – 17 , 35 , 37 – 39 , 41 , 43 , 46 Although the results were obtained using self-report screening instruments and not clinical assessment, they strongly indicate a need for mental health promotion strategies to prevent maladaptive coping mechanisms such as alcohol and drug use 15 and prevent mental disorders from developing and becoming chronic. 10 Although fear, grief and anxiety are expected responses to such a health crisis, 36 they also indicate that regardless of the country, mental healthcare should be widely and easily accessible to university students, together with sanitary measures to control a pandemic. The differences between the samples could indicate country/cultural differences that could be further investigated and should be considered when managing a global health emergency such as the COVID-19 pandemic. Finally, the study highlights that university students are at increased risk of developing mental disorders in times of crises, underlining a critical call to action for universities to proactively enhance support systems (for instance, through online support groups 54 ). Thus, it seems likely that providing mental health promotion programmes and implementing accessible, low-threshold (online or in-person) mental health support targeting university students may help them to cope.
|
39494847_p43
|
39494847
|
Future implications
| 4.165109 |
biomedical
|
Study
|
[
0.9986404776573181,
0.0006342646083794534,
0.0007252360228449106
] |
[
0.9984403252601624,
0.00024067929189186543,
0.0012444342719390988,
0.00007462468784069642
] |
en
| 0.999998 |
This research was completed in Scotland, which has a publicly funded healthcare system, managed by local Health Boards, who oversee general practice, hospitals, community services, therapies, diagnostics and allied health services. General practitioners serve as gatekeepers to mental health, psychiatry specialists and community paediatrics. Long-term ADHD management can occur in primary care under ‘shared care’ arrangements, but practice variations exist in prescribing and mental health/ADHD services. Recruitment challenges persist, particularly for psychiatrists and mental health nurses and in community paediatrics. ADHD assessments should follow the National Institute for Health and Care Excellence ADHD guidelines, 8 Scottish Intercollegiate Guidelines Network ‘Management of attention deficit and hyperkinetic disorders in children and young people’, 19 and the recent update on ADHD by the Royal College of Psychiatrists (Scotland). 20 Recommendations include considering differential diagnoses, conducting full assessments before diagnosis, psychosocial strategies as first-line supports for children and young people, giving environmental modification advice before prescribing for adults, and establishing local practice protocols between mental health teams and general practitioners. Prescription decisions should consider various factors, with agreements on processes and responsibilities for addressing concerns. Medications typically used to treat ADHD in Scotland include methylphenidate preparations, dexamfetamine preparations, atomoxetine, guanfacine and modafinil.
|
39113462_p0
|
39113462
|
Study background
| 3.834262 |
biomedical
|
Other
|
[
0.9832584857940674,
0.014333994127810001,
0.002407468855381012
] |
[
0.08582822233438492,
0.7676107883453369,
0.14343449473381042,
0.0031265367288142443
] |
en
| 0.999997 |
The analyses of medication use in Scotland presented here are part of a comprehensive programme aiming to act as a catalyst for developing improved services. The Scottish Government partnered with the authors of the current work, the National Autism Implementation Team (NAIT) and the Royal College of Psychiatrists in 2020 to establish comprehensive pathways for adult ADHD care as part of a national effort to enhance neurodevelopmental assessment, diagnosis and supports across the lifespan. 21 – 25 The NAIT team is composed of neurodivergent and neurotypical individuals, as well as professionals from various domains, including psychiatry, education, speech and language therapy, and occupational therapy. The team collaborates with academics and patients to champion evidence-informed practices and neuro-affirming work, with a focus on multi-professional and cross-sector work, environmental modifications and partnership with neurodivergent individuals. 21 – 25
|
39113462_p1
|
39113462
|
Study background
| 2.253651 |
biomedical
|
Other
|
[
0.9724332690238953,
0.007409326732158661,
0.02015739493072033
] |
[
0.02056761458516121,
0.9765810966491699,
0.0019827375654131174,
0.00086854153778404
] |
en
| 0.999997 |
This study aimed to analyse prescribing patterns for medication typically used to treat ADHD in Scotland for adolescents and adults. Although recent analysis from Public Health Scotland 26 offered initial insights into decreasing prescription rates after 15 years of age and significant regional variation in medication usage in Scotland, our study aims to further explore and substantiate these findings by using a comprehensive population data-set.
|
39113462_p2
|
39113462
|
Aims
| 3.387872 |
biomedical
|
Study
|
[
0.9969218373298645,
0.0017537157982587814,
0.0013243359280750155
] |
[
0.9989888072013855,
0.0006650759605690837,
0.00019699029508046806,
0.00014913278573658317
] |
en
| 0.999997 |
The study used information from the Prescribing Information System (PIS). 26 The PIS includes all National Health Service (NHS) prescribing relating to all medicines, and their costs, that are prescribed and dispensed in the community in Scotland. General practitioners provide most prescriptions, with other authorised prescribers such as nurses also contributing. Prescriptions written in hospitals that are dispensed in the community are also included. Prescriptions dispensed within hospital settings or hospital-based clinics are not included. Private prescriptions and direct supplies are also not included. Data include a unique identifier, the Community Health Index (CHI) number, prescriber and dispenser details for community prescribing, costs and drug information. Data on practices (e.g. NHS Boards), and prescribable items (e.g. manufacturer, formulation code, strength) are also included. NHS Scotland uses extensive computerised record keeping, with around 100 million data items loaded into the PIS per annum. 26 The PIS data are refreshed monthly, and therefore provide a very robust indication of routine prescribing practice in the community.
|
39113462_p3
|
39113462
|
Data-set
| 3.39238 |
biomedical
|
Study
|
[
0.9769725799560547,
0.013404067605733871,
0.009623277932405472
] |
[
0.9940342307090759,
0.004903885535895824,
0.0005743831279687583,
0.00048755822353996336
] |
en
| 0.999997 |
An extract from the PIS was made in 2021 for 2019 and 2010 data. Both extracts identified people prescribed medication from British National Formulary Section 4.4 ‘CNS stimulants and drugs used for ADHD’ (including methylphenidate preparations (immediate release and sustained release), dexamfetamine preparations (immediate release and sustained release), atomoxetine and guanfacine), but excluding modafinil because of its non-ADHD indications. The extracts included items prescribed in Scotland, with all available prescriber types included. Information is for items that have been prescribed and subsequently dispensed. The extracts reported data by unique patient identified (UPI) (the number of cases prescribed medication typically used to treat ADHD), defined daily dose (DDD) (assumed average maintenance dose per day used on its main indication in adults) and gross ingredient cost (GIC) (a measure to show the cost of items). Changes in DDD, GIC and UPI from 2010 to 2019 were analysed.
|
39113462_p4
|
39113462
|
Data processes and analysis
| 3.989018 |
biomedical
|
Study
|
[
0.9985626339912415,
0.0010697742691263556,
0.00036758751957677305
] |
[
0.9987685084342957,
0.0007531464216299355,
0.00036253462894819677,
0.00011576881661312655
] |
en
| 0.999998 |
The 2019 extract additionally included age and NHS Scotland Health Board locality. Age was calculated on 30 September 2019 and all data by age is a unique count, based on a unique identifier (CHI). The following age categories were chosen: children and young people aged 10–19 years and adults aged 20–59 years. These age categories allowed comparison between paediatric and adult services. Data for people over 60 years of age were excluded because of the limited numbers recorded. The number of cases prescribed medication typically used to treat ADHD (UPIs) was compared with the National Records of Scotland mid-year population estimate by age group for NHS Board. 27 The ratio of number of cases and corresponding mid-year population estimate was calculated (cases per population – treated prevalence). We set the expected prevalence of ADHD in adolescents as between 5 and 7%, and the expected prevalence of ADHD in adults as between 2 and 4%. 1 – 3 The number of cases identified was then compared to expected prevalence and the treated proportion of expected prevalence calculated for age groups 10–19 years and 20–59 years.
|
39113462_p5
|
39113462
|
Data processes and analysis
| 4.045541 |
biomedical
|
Study
|
[
0.9989506006240845,
0.0006945269415155053,
0.0003548930981196463
] |
[
0.9994294047355652,
0.0002722545468714088,
0.0002308900438947603,
0.00006747170846210793
] |
en
| 0.999996 |
This analysis is based on an anonymised, aggregated data-set containing routinely collected data, and therefore no ethical approvals have been sought.
|
39113462_p6
|
39113462
|
Ethics
| 1.582726 |
biomedical
|
Study
|
[
0.9413741827011108,
0.0022576856426894665,
0.05636819452047348
] |
[
0.6854149103164673,
0.3104894161224365,
0.0027721747756004333,
0.0013234831858426332
] |
en
| 0.999996 |
An analysis of UPI, DDD and GIC ( Table 1 ) showed prescribing patterns of medication for people with ADHD between 2010 and 2019 at NHS Board (geographical) level. Overall, there was a large increase in the volume of people receiving medication for ADHD between 2010 and 2019. For Scotland as a whole, this increase was +233.2% for the number of people dispensed a prescription (UPI), +234.9% for the DDD and +216.6% for changes in expenditure on these medicines (GIC). Table 1 Increases in people prescribed medication, by NHS Board , for individuals aged ≥10 years Health Board area UPI a (% increase) DDD b (% increase) GIC c (% increase) Ayrshire and Arran 173.1 168.6 156.9 Borders 160.1 183.7 151.8 Dumfries and Galloway 158.1 203.5 178.9 Fife 183.8 177.4 145.2 Forth Valley 189.2 180.3 170.6 Grampian 217.1 214.7 232.8 Greater Glasgow and Clyde 260.4 269.0 224.2 Highland 201.7 219.4 213.4 Lanarkshire 301.0 349.6 314.7 Lothian 293.8 294.7 302.0 Orkney 136.8 142.9 133.9 Shetland 200.0 194.5 186.3 Tayside 293.1 261.7 244.1 Western Isles 277.8 354.6 586.6 Mean (s.d.) 217.6 (56.69) 229.6 (66.56) 231.5 (116.33) Scotland 233.2 234.9 216.6 UPI, unique patient identified; DDD, defined daily dose; GIC, gross ingredient cost. a. UPI represents the number of cases prescribed medication typically used to treat attention-deficit hyperactivity disorder. b. DDD is the assumed average maintenance dose per day used on its main indication. c. GIC is a measure to show the cost of items reimbursed before deduction of any dispenser discount.
|
39113462_p7
|
39113462
|
Changes in prescribing between 2010 and 2019
| 4.125841 |
biomedical
|
Study
|
[
0.997509241104126,
0.0020047728903591633,
0.00048603612231090665
] |
[
0.998993456363678,
0.000455827743280679,
0.0004260138375684619,
0.00012468770728446543
] |
en
| 0.999996 |
A marked variation in increase between NHS Board level was also observed from these data, for each of the parameters. Between Scotland's 14 Health Board areas, increases ranged from +136.8% to +301% for UPI (mean increase +217.6%, s.d. 56.69). Increases across Health Boards for DDD ranged from +142.9% to +354.6% (mean increase +229.6%, s.d. 66.56). The increase for GIC across health boards ranged from +133.9% to +586.6% (mean increase +231.5%, s.d. 116.33).
|
39113462_p8
|
39113462
|
Changes in prescribing between 2010 and 2019
| 2.962419 |
biomedical
|
Study
|
[
0.9622082114219666,
0.0050241220742464066,
0.032767657190561295
] |
[
0.9940926432609558,
0.005332304630428553,
0.00038741540629416704,
0.00018768162408377975
] |
en
| 0.999996 |
The 2019 extract was 15 516 people aged 10–59 years. The CHI capture rates were 94.27%. In the extract, 92.5% had a valid CHI, and 7.5% had missing data and so were excluded from the analysis, leaving a usable sample of 12 804 .
|
39113462_p9
|
39113462
|
ADHD prescribing in 2019
| 1.931687 |
biomedical
|
Study
|
[
0.976207971572876,
0.002096752868965268,
0.021695176139473915
] |
[
0.9734418392181396,
0.02569342590868473,
0.00035628757905215025,
0.0005084945587441325
] |
en
| 0.999996 |
Using 2019 mid-year population estimates for Scotland 27 1.35% (95% CI 1.33–1.39) of the entire adolescent (10–19 years) Scottish population were receiving medication typically used to treat ADHD on the NHS ( Table 2 ). There was variance geographically across Scotland: between 0.26 and 2.92% ( Table 2 ) of local adolescent populations were receiving medication typically used to treat ADHD in 2019. Variance in the treated proportion of expected prevalence was between 3.62 and 58.36% of adolescents with likely ADHD at the 5% level ( Table 2 ), and between 2.59 and 41.69% at the 7% level ( Table 2 ). For Scotland as a whole, at the 5% level, 27.07% of children with likely ADHD were receiving medication typically used to treat ADHD, and at the 7% level, 19.34% of likely ADHD cases were receiving medication typically used to treat ADHD. Table 2 Cases per population (treated prevalence) and comparison of expected prevalence with the number of treated cases, by NHS Board Population a UPI b Cases per population (treated prevalence) 95% CI lower limit 95% CI upper limit Expected prevalence c Treated proportion of expected prevalence d 5% 7% 5% 7% Health Board area n n % n n % % Ayrshire and Arran 39 233 313 0.80 0.71 0.89 1962 2746 15.96 11.40 Borders 12 054 251 2.08 1.82 2.34 603 844 41.65 29.75 Dumfries and Galloway 15 140 252 1.66 1.46 1.87 757 1060 33.29 23.78 Fife 41 162 840 2.04 1.90 2.18 2058 2881 40.81 29.15 Forth Valley 34 444 467 1.36 1.23 1.48 1722 2411 27.12 19.37 Grampian 60 770 1066 1.75 1.65 1.86 3039 4254 35.08 25.06 Greater Glasgow and Clyde 121 249 1303 1.07 1.02 1.13 6062 8487 21.49 15.35 Highland 34 047 369 1.08 0.97 1.19 1702 2383 21.68 15.48 Lanarkshire 74 460 392 0.53 0.47 0.58 3723 5212 10.53 7.52 Lothian 92 151 1236 1.34 1.27 1.42 4608 6451 26.83 19.16 Orkney 2273 <10 e <0.44 − − 114 159 <8.77 <6.29 Shetland 2577 13 0.50 0.23 0.78 129 180 10.09 7.21 Tayside 44 309 1293 2.92 2.76 3.08 2215 3102 58.36 41.69 Western Isles 2763 <10 e <0.36 − − 138 193 <7.25 <5.18 Scotland 576 632 7815 1.35 1.33 1.39 28 832 40 364 27.07 19.34 UPI, unique patient identified; ADHD, attention-deficit hyperactivity disorder. a. National Records of Scotland mid-year population estimate by age group, for NHS Board, 2019. b. UPI: the numbers of cases prescribed medication typically used to treat ADHD. c. Expected prevalence of ADHD, as applied to the population. d. Expected prevalence compared with the number of people identified as being treated for ADHD (UPI). e. Records of fewer than ten cases have been hidden to manage disclosure risk.
|
39113462_p10
|
39113462
|
Adolescent
| 4.192236 |
biomedical
|
Study
|
[
0.9990350008010864,
0.0005971460486762226,
0.0003679012006614357
] |
[
0.9992156028747559,
0.00035650981590151787,
0.0003568496322259307,
0.00007097611523931846
] |
en
| 0.999997 |
The same approach was applied to the adult cohort (20–59 years) ( Table 3 ). Using 2019 mid-year population estimates, 27 0.17% (95% CI 0.17–0.18) of the entire Scottish adult population were receiving medication typically used to treat ADHD on the NHS in 2019. There was variance geographically. Across different areas of Scotland, between 0 and 0.28% ( Table 3 ) of the local adult population were receiving medication typically used to treat ADHD in 2019. Variance in the treated proportion of expected prevalence was between 0 and 14.19% of adults with likely ADHD at the 2% level ( Table 2 ), and between 0 and 7.10% at the 5% level ( Table 2 ). For Scotland as a whole, at the 2% level, 8.54% of adults with likely ADHD were receiving medication typically used to treat ADHD, and at the 4% level, 4.27% of likely ADHD cases were receiving medication typically used to treat ADHD. Table 3 Cases per population (treated prevalence) and comparison of expected prevalence with the number of treated cases, by NHS Board Population a UPI b Cases per population (treated prevalence) 95% CI lower limit 95% CI upper limit Expected prevalence c Treated proportion of expected prevalence d 2% 4% 2% 4% Health Board area n n % n n % % Ayrshire and Arran 182 499 214 0.12 0.12 0.13 3650 7300 5.86 2.93 Borders 54 493 108 0.20 0.20 0.24 1090 2180 9.91 4.95 Dumfries and Galloway 70 172 102 0.15 0.15 0.17 1403 2807 7.27 3.63 Fife 191 299 492 0.26 0.26 0.28 3826 7652 12.86 6.43 Forth Valley 162 270 203 0.13 0.13 0.14 3245 6491 6.26 3.13 Grampian 317 681 454 0.14 0.14 0.16 6354 12 707 7.15 3.57 Greater Glasgow and Clyde 669 668 1070 0.16 0.16 0.17 13 393 26 787 7.99 3.99 Highland 158 096 210 0.13 0.13 0.15 3162 6324 6.64 3.32 Lanarkshire 350 041 284 0.08 0.08 0.09 7001 14 002 4.06 2.03 Lothian 520 525 1251 0.24 0.24 0.25 10 411 20 821 12.02 6.01 Orkney 10 964 0 0 − − 219 439 0 0 Shetland 11 563 0 0 − − 231 463 0 0 Tayside 214 932 610 0.28 0.28 0.31 4299 8597 14.19 7.10 Western Isles 12 596 0 0 − − 252 504 0 0 Scotland 2 926 799 4998 0.17 0.17 0.18 58 536 117 072 8.54 4.27 UPI, unique patient identified; ADHD, attention-deficit hyperactivity disorder. a. National Records of Scotland mid-year population estimate by age group, for NHS Board, 2019. b. UPI: the numbers of cases prescribed medication typically used to treat ADHD. c. Expected prevalence of ADHD, as applied to the population. d. Expected prevalence compared with the number of people identified as being treated for ADHD (UPI).
|
39113462_p11
|
39113462
|
Adult
| 4.163771 |
biomedical
|
Study
|
[
0.9989650249481201,
0.000702793593518436,
0.0003320797404740006
] |
[
0.9993622899055481,
0.00027783075347542763,
0.0002853607293218374,
0.00007459465268766508
] |
en
| 0.999997 |
The main benefit of this analysis lies in its utilisation of contemporary data from a national prescribing database, offering a reasonably accurate snapshot of current practice in a national population. The study revealed significant variation in the proportion of the population receiving active pharmacological treatment for ADHD. The number of people receiving pharmacological therapy was significantly below conservative estimates of those who might benefit from such intervention in the community. Our results illustrate that despite a noticeable increase in ADHD prescribing, levels remain lower than expected. This study is important given the concerns about rising prescriptions and fears of clinician overprescribing. These findings serve as strong evidence to the contrary. Some commentators express particular concern over the substantial rise in adult prescriptions. Alternatively, this is a predictable phenomenon, as increasing numbers of individuals initiated on medication during childhood go on to seek continued treatment as they progress into adulthood. Additionally, more adults are seeking diagnosis as ADHD becomes more accepted and understood, and awareness of missed diagnosis in childhood and how this might present in adulthood increases. This an equity and public health issue, as the perceived issues and nervousness about increases may be used as a reason for capping, limiting or even not treating individuals.
|
39113462_p12
|
39113462
|
Discussion
| 4.145622 |
biomedical
|
Study
|
[
0.9989392161369324,
0.0008271239348687232,
0.0002336971665499732
] |
[
0.9973260164260864,
0.00036889154580421746,
0.0021398793905973434,
0.0001651171623962
] |
en
| 0.999997 |
It remains evident that there was a significant increase in the number of individuals receiving medication for ADHD between 2010 and 2019. This increase was illustrated using the numbers of people dispensed a prescription (UPI, +233.2%), the DDD (+234.9%) and changes in expenditure on these medicines (GIC, +216.6%). However, the analysis also revealed variations in the number of treated cases compared with estimates of prevalence, indicating less use of medication than might be expected, given the likely size of the population who might benefit from pharmacological interventions. In Scotland, with prevalence estimated at 5–7%, between 72.93 and 80.66% of the adolescent population with likely ADHD were not prescribed medication typically used to treat ADHD in 2019. This represents less than a quarter of the estimated eligible cohort who received a prescription in 2019. For adults, the findings were even more pronounced, with between 91.46 and 95.73% of adults with likely ADHD not prescribed medication typically used to treat ADHD in 2019. This means that fewer than one in ten of the estimated eligible adult cohort were identified as having received a prescription in 2019. Variation in the treated proportion of expected prevalence was present across geographies. In some areas of Scotland, no adult was identified as having a prescription for medication typically used to treat ADHD in 2019. No area exceeded 15% in treatment of expected prevalence for adults. All areas of Scotland had some ADHD prescribing for adolescents, and in some areas, treatment approached or exceeded 50% of expected prevalence.
|
39113462_p13
|
39113462
|
Discussion
| 4.104064 |
biomedical
|
Study
|
[
0.9990559220314026,
0.0006420656573027372,
0.0003020042204298079
] |
[
0.9990848302841187,
0.0002522606519050896,
0.0005817859200760722,
0.0000811133868410252
] |
en
| 0.999998 |
Our data is in keeping with other published evidence. Studies are identifying decreased ADHD treatment prevalence alongside increases in diagnosis. Recent publications from the UK 28 – 30 found that for cases aged 10–20 years, 61.6% (95% CI 60.6–62.5%) had a prescription at some point for medication typically used to treat ADHD; however, prescribing prevalence declined between the ages of 16 and 18 years, from 37.8% (95% CI 36.6–38.9) to 23.7% (95% CI 22.7–24.6%). This research also confirmed regional variations in prescribing in the UK. 29 The authors identified patchy distribution and unavailability of dedicated ADHD UK services, 29 as well as the risk of cessation of medication, with vulnerable individuals with ongoing ADHD symptoms facing barriers to re-entering services and accessing support. 28 A recent meta-analysis of international studies 31 found pooled pharmacological treatment rates were 19.1% (95% CI 11.5–29.9) in children and adolescents with ADHD. This research further estimated that 70% of young people with ADHD might benefit from a trial with pharmacological treatment, and that a substantial number of young people with a diagnosis who might benefit from medication were not receiving it. 31
|
39113462_p14
|
39113462
|
Discussion
| 4.128922 |
biomedical
|
Study
|
[
0.9994188547134399,
0.00037573272129520774,
0.00020544754806905985
] |
[
0.9936562180519104,
0.0003085861389990896,
0.005907693412154913,
0.0001274121314054355
] |
en
| 0.999997 |
Explaining the factors influencing the observations made in this study is complex, and there are many potential reasons for the discrepancies observed. Geographic discrepancies in ADHD medication use in Scotland are certainly multifactorial in cause. However, one possible explanation is that variations in service provision and available resources in each region, including access to ADHD diagnostic assessment and transition arrangements, are a contributing factor. All areas in Scotland have a children's ADHD pathway, but an analysis from 2021 found only two regional adult ADHD pathways, and no pathways for adults with ADHD without significant comorbidity. 21 These issues could be contributing to limited medication use resulting from a lack of expertise and structures. In some areas of Scotland, no adult was identified as having a prescription. These results came from the island communities of Orkney, Shetland and the Western Isles. These areas are remote and rural, with smaller populations. However, it is possible that local mental health teams have not routinely considered ADHD, despite agreements with larger areas to access more complex health services, including adult neurodevelopmental assessments. Additionally, the limited access to practitioners experienced in treating adults with ADHD could be a crucial factor in the low – or in these areas, entirely absent – use of pharmacological interventions.
|
39113462_p15
|
39113462
|
Discussion
| 4.119998 |
biomedical
|
Study
|
[
0.998839795589447,
0.0007984109106473625,
0.00036191463004797697
] |
[
0.9985830783843994,
0.0003572510904632509,
0.0009481748566031456,
0.0001115493432735093
] |
en
| 0.999995 |
The evidence from this study tentatively points to an incomplete recognition of ADHD in adult psychiatric practice. The need for frequent professional review and misconceptions and biases about ADHD and stimulant treatments lead to reduced uptake of medication among adults. 32 Coupled with insufficient skills, staff and services, this is a significant challenge. 17 The COVID-19 pandemic has exacerbated the situation, leading to delays and backlogs in non-priority appointments, with long waiting lists. 21 Stakeholders report difficulties in accessing help, often facing a ‘postcode lottery’ of services. 21 People referred for ADHD may experience rejection from mental health services as their condition may not be deemed as serious as more complex cases, which may involve use of the Mental Health Care and Treatment Scotland Act and higher perceived risk. 17 In cases of ADHD and mental illness, services may prioritise mental health issues over ADHD assessment and intervention and overlook underlying ADHD concerns. 33 Transitions between services can also lead to potential discontinuation of medication. 28 Some primary care practitioners may not consider ADHD monitoring or prescribing as their responsibility, and may hold unhelpful attitudes about the validity of ADHD as a diagnosis and the role of medication. 34 The difficulty in accessing support in the UK and Scotland may force individuals to resort to costly private services, creating a two-tier health system that disadvantages lower-income families.
|
39113462_p16
|
39113462
|
Discussion
| 4.089723 |
biomedical
|
Study
|
[
0.9971574544906616,
0.002359302481636405,
0.0004831966070923954
] |
[
0.990718424320221,
0.0008768639527261257,
0.00805441103875637,
0.0003503702755551785
] |
en
| 0.999997 |
A further issue is substance use disorders (SUDs). Co-occurrence of ADHD and SUD is common in clinical settings, and a question is whether adults with both diagnoses should be prescribed stimulants despite misuse potential. 35 There are higher rates of SUD in people with ADHD, and some indication that genetic liability to ADHD is associated with higher risks of SUD. 36 However, evidence would suggest that ADHD pharmacotherapy does not increase the risk of SUD. 35 If managed correctly, stimulants are an effective treatment for ADHD, and reducing ADHD symptoms may go on to improve global functioning, reducing risk for SUD, rather than increasing it. 35 This is a particularly pertinent issue for Scotland, where drug-related deaths are relatively high and represent a significant public health issue. Effective interventions that target a particularly at-risk group (in this case people with ADHD) and may lead to reduced levels of drug-related harm and deaths are therefore important, and should be prioritised.
|
39113462_p17
|
39113462
|
Discussion
| 4.005561 |
biomedical
|
Review
|
[
0.996178150177002,
0.002559791784733534,
0.001262099132873118
] |
[
0.10670100152492523,
0.017699310556054115,
0.8748353123664856,
0.0007643300341442227
] |
en
| 0.999997 |
In future research, priority should be given to collecting a broader range of variables, including linked data on diagnoses received, ethnicity, gender and co-occurrence of ADHD with SUDs. Monitoring and evaluating the development, distribution and effectiveness of specialist ADHD practices would provide insights into their impact on diagnosis rates, psychosocial supports offered and medication. Linked data would strengthen the findings of our research, and would be enhanced particularly by diagnostic information about why medications were given. Future research should also explore the possibility of analysing more refined age categories, particularly focusing on young adults versus individuals aged 30 years and above. It is plausible that further insights on prescribing rates would be derived when age is split in this way. Employing research methods that embrace complexity, such as realist evaluation, theory-based approaches and implementation science methodologies, 37 would provide useful insights. Co-production with individuals who have lived experience, as well as conducting qualitative research to directly understand the challenges perceived by people seeking ADHD diagnosis, would also be beneficial. 38
|
39113462_p18
|
39113462
|
Implications
| 4.060518 |
biomedical
|
Study
|
[
0.9994921684265137,
0.0002450523024890572,
0.0002627007488626987
] |
[
0.9933215975761414,
0.0012958117295056581,
0.005285562481731176,
0.00009695110929897055
] |
en
| 0.999998 |
Results highlight implications for practitioners. It is crucial to recognise that while medication is an important and effective intervention, the emphasis should be on prioritising environmental modifications and psychosocial support as first-line strategies. 8 Considering both medication and environmental modifications as complementary components provides individuals with the most effective and tailored support. Findings from this research also emphasise the need for specialised adult approaches that complement those for young people. Building on the recommendations of the recent Royal College of Psychiatrists (Scotland) ADHD update, 20 development of cohesive ‘neurodevelopmental’ pathways emerges as a logical response to the growing need. Although integrated neurodevelopmental pathways remain rare in adult services, there are examples for children's services. 22 , 24 Developing and implementing such pathways requires coverage across self-help, skilled non-statutory services and primary, secondary and tertiary care, representing comprehensive multi-level work that addresses practitioner and institutional elements. 24 Regarding medication, a key challenge is achieving consensus on prescription practices. Non-medical prescribing roles have the potential to assist, but require advanced practitioner status, proper supervision and funding. 39 Formal cross-sector performance indicators for pathways, such as waiting times, good practice in assessment, post-diagnostic and psychosocial supports offered, would facilitate consistent practice and effective monitoring. Recognising the value of non-pharmacological approaches necessitates investing in allied health professional roles, including advanced practitioner and consultant roles. An important aspect in mental health services is ensuring that care approaches are neuro-affirmative. 25 , 40 Non-pharmacological supports (particularly talking therapies) offered to people with ADHD may not fully consider individual needs, and may cause distress or harm if not adequately adapted for neurodivergent people. 40
|
39113462_p19
|
39113462
|
Implications
| 4.096378 |
biomedical
|
Study
|
[
0.9815601110458374,
0.016970505937933922,
0.001469393610022962
] |
[
0.599243700504303,
0.034599848091602325,
0.3619459867477417,
0.00421050563454628
] |
en
| 0.999997 |
The data utilised in this study were sourced from the PIS, facilitated by colleagues at Public Health Scotland, and linked to a unique patient identifier, the CHI number. A limitation is that our data-set does not capture medications dispensed via hospital services or private prescriptions, which could lead to underestimations of medication use. However, it is important to note that in Scotland, at the time of data collection, there were few private providers, and most prescribing would still be in the community. Our analysis employed broad age bands, which might have obscured more nuanced age-related variations. The lack of linked diagnostic records prevents a direct understanding of the actual number of individuals diagnosed with ADHD across different regions. This limitation necessitates reliance on estimated ADHD prevalence rates derived from external studies. These estimates from external sources do not account for local socioeconomic, cultural or environmental variables that affect ADHD prevalence.
|
39113462_p20
|
39113462
|
Limitations
| 3.718938 |
biomedical
|
Study
|
[
0.998781144618988,
0.000463770586065948,
0.0007551114540547132
] |
[
0.999312162399292,
0.0004322488384786993,
0.00019520087516866624,
0.00006032761302776635
] |
en
| 0.999996 |
Understanding prescription rates is complicated by various interpretation challenges. There are complexities in assuming a straightforward correlation between ADHD prevalence and medication prescription rates. For example, a lower than expected rate of prescriptions being filled does not necessarily imply undertreatment, as some people may opt for non-pharmacological interventions or discontinue medication because of side-effects. This also links to a critique of the idea that higher prescription rates indicate better or more appropriate treatment. Prescription percentages not approaching prevalence of ADHD may be appropriate and consistent with high-quality care. It is very important to reemphasise that prescriptions alone do not sufficiently reflect an index of quality treatment, given the emphasis on behavioural and environmental supports as the primary management strategies for ADHD. The developmental trajectory of ADHD further complicates interpretations. Many individuals with ADHD continue to require support across the lifespan, but some adolescents requiring pharmacological treatments may not need them in adulthood, with a waxing and waning pattern of presentation, suggestive of an episodic treatment trajectory. 41 Given these complexities, our findings should be considered carefully.
|
39113462_p21
|
39113462
|
Limitations
| 4.026337 |
biomedical
|
Study
|
[
0.9993403553962708,
0.00042684428626671433,
0.00023277674335986376
] |
[
0.976155161857605,
0.0012863141018897295,
0.022328438237309456,
0.00023007093113847077
] |
en
| 0.999996 |
In conclusion, this study provides insights into medication use for ADHD across adolescents and adults. The discrepancies observed between estimated prevalence of ADHD and rates of medication prescriptions warrant a cautious interpretation. Based on estimates of the likely ADHD population, it was found that approximately a quarter of adolescents and a tenth of adults were recorded as receiving a prescription through the NHS. Although prescription rates have increased over time, there was no evidence to suggest overtreatment or excessive use of medication. Findings indicate less use of medication in certain segments of the population, particularly in adults, but these numbers should be considered alongside the diversity of management strategies in ADHD. The findings highlight a need to further examine current clinical delivery models, particularly in adult services. Although this study is centred on Scotland, insights and recommendations have broader applicability to other regions facing similar challenges.
|
39113462_p22
|
39113462
|
Limitations
| 4.093385 |
biomedical
|
Study
|
[
0.9989699125289917,
0.000813232094515115,
0.00021685041429009289
] |
[
0.9986553192138672,
0.000261973385931924,
0.0009470487129874527,
0.00013565269182436168
] |
en
| 0.999995 |
For many co-occurring health conditions in autism, gaps in the literature prevent conclusions or interpretation of underlying mechanisms. This study therefore aims to compare the physical health of individuals with and without autism, using a nested cross-sectional design on a large, prospectively recruited e-cohort of individuals across a broad range of ages, to assess whether specific conditions or clusters of conditions are more common in individuals with autism.
|
39439364_p0
|
39439364
|
Aims
| 4.044723 |
biomedical
|
Study
|
[
0.9993213415145874,
0.0003241505182813853,
0.0003545710351318121
] |
[
0.9995303153991699,
0.0001636111264815554,
0.00025598882348276675,
0.000050139802624471486
] |
en
| 0.999997 |
The sample was drawn from the National Centre for Mental Health (NCMH) database. The NCMH is a Welsh Government-funded research centre investigating neurodevelopmental, adult psychiatric and neurodegenerative psychiatric disorders across the lifespan. 30 The cohort in the database consists of individuals aged 4 years or older, who have experienced or are related to someone who has experienced one of these disorders, as well as volunteer control participants who have not experienced any disorder. 30 Participants were recruited systematically to the NCMH through disease registers, clinical note screening and identification by clinical care teams, and non-systematically through media advertisements, posters/leaflets in National Health Service (NHS) waiting rooms, voluntary organisations and contacting of individuals involved in previous studies within the Division of Psychological Medicine and Clinical Neurosciences, Cardiff University, UK. 30 All individuals in this study provided written informed consent after viewing the patient information sheet or, for those under 16 years of age or lacking sufficient mental capacity, assent was provided where possible, and written consent was obtained from a nominated individual, such as the next of kin, a family member or a carer. 30 Participants underwent a standardised interview establishing personal and family history of mental illness and medication profile, and were also given a standardised self-report questionnaire to complete. 30
|
39439364_p1
|
39439364
|
Study population and measures
| 4.012365 |
biomedical
|
Study
|
[
0.9990440011024475,
0.0005622219177894294,
0.00039367875433526933
] |
[
0.9993738532066345,
0.00033377838553860784,
0.00022837288270238787,
0.00006399994163075462
] |
en
| 0.999997 |
The NCMH received a favourable ethical opinion from the Wales Research Ethics Committee 2 on 25 November 2016 . This project utilised data held within the NCMH ethical approval, as part of existing questionnaires. An application for data access was made in December 2021, and approved after internal board review in January 2022. The NCMH database was interrogated and individuals were extracted for inclusion in this study if they had reported having a clinical diagnosis of autism and had filled in any of the medical history section of the questionnaire ( n = 813) . An unmatched control sample of individuals without autism and with no mental health or neurodevelopmental conditions was also obtained from the NCMH database. Any individuals who were completely missing health data were excluded from analysis. Where individuals had only completed part of the questionnaire, they were included in analysis for physical conditions where they had data available.
|
39439364_p2
|
39439364
|
Study population and measures
| 3.04065 |
biomedical
|
Study
|
[
0.9958981871604919,
0.000931676768232137,
0.003170053707435727
] |
[
0.9956191182136536,
0.004017460159957409,
0.00020220466831233352,
0.00016111401782836765
] |
en
| 0.999998 |
Following consent, trained researchers administered a standardised interview assessment to gather data on the participant's diagnostic history at enrolment to NCMH. The NCMH Brief Assessment captures information on lifetime physical health conditions in its medical history section at enrolment to NCMH, and this was used to derive the outcome measure of physical health conditions in the sample. Participants are given a list of conditions and told to indicate whether they have ever been told by a doctor or health professional that they have the condition, allowing for a binary outcome of either presence or absence of the condition. Questions on lifetime diagnosis are asked as ‘Has a doctor or health professional ever given you a diagnosis of [listing a broad range of psychiatric and physical health conditions, one at a time, including autism]?’ Further information regarding diagnosis, symptoms, treatment and outcomes was obtained from clinical records, where appropriate consent had been obtained to do so. After assessment of the data (see Supplementary Material), 28 physical health variables were included for statistical analysis.
|
39439364_p3
|
39439364
|
Study population and measures
| 4.041753 |
biomedical
|
Study
|
[
0.9984601736068726,
0.0010351954260841012,
0.0005046224687248468
] |
[
0.9994296431541443,
0.000343016698025167,
0.00015243956295307726,
0.0000749238024582155
] |
en
| 0.999998 |
Initially, data were plotted to visualise and compare group frequencies, followed by χ 2 -test to assess the association between autism and physical health, with a significance level of 5%. Binomial logistic regression was used to calculate odds ratios for each condition in autism, with a 95% confidence interval. Two logistic regression models were used. The first model looked at the odds of each condition in autism and included age and gender as covariates in the model. Autism diagnosis was the independent variable and physical health diagnoses were the dependent variable. A second multivariable logistic regression model further included smoking status, mood stabiliser use and antipsychotic use, as well as age and gender as covariates, again with autism diagnosis as the independent variable and physical health diagnoses the dependent variable. These covariates were selected as known risk factors for a wide range of diseases (see Supplementary Material for explanation and citations). Although some mood stabilisers are also anticonvulsants, removing mood stabilisers as a control variable when investigating epilepsy had little effect on the odds ratio in a post hoc assessment (see Supplementary Material), so it was deemed most appropriate to continue using the same model for all conditions.
|
39439364_p4
|
39439364
|
Statistical analysis
| 4.083358 |
biomedical
|
Study
|
[
0.9994612336158752,
0.0002942559076473117,
0.0002444868441671133
] |
[
0.9994245767593384,
0.00020020740339532495,
0.0003178555052727461,
0.0000574156001675874
] |
en
| 0.999997 |
Additional subanalyses were conducted to investigate the effects of intellectual disability. Binomial logistic regression models were run in which each physical health condition outcome (dependent variable) was assessed, comparing controls with individuals with autism with ( n = 86) and without ( n = 727) intellectual disability, controlling for age, gender, smoking status and antipsychotic and mood stabiliser use. A direct comparison model between autism with intellectual disability and isolated autism was not utilised because of concerns around collider biases between autism and intellectual disability. In the subanalysis, a further five conditions were identified as having no cases among individuals with concurrent intellectual disability, and therefore were not modelled.
|
39439364_p5
|
39439364
|
Intellectual disability subanalysis
| 4.043984 |
biomedical
|
Study
|
[
0.9994113445281982,
0.000329957198118791,
0.00025860185269266367
] |
[
0.999431312084198,
0.000199201051145792,
0.000313052732963115,
0.00005631786916637793
] |
en
| 0.999996 |
Benjamini–Hochberg correction was carried out to adjust for multiple hypothesis testing and reduce the chance of a type 1 error, with an initial α of 0.05. In the main analysis, 75 tests were run (25 tests per model) and included in the correction, leading to a false discovery rate correction of 0.029. In the subanalysis, 20 tests were run, leading to a false discovery rate correction of 0.029. Data were analysed with IBM SPSS Statistics Version 27 for Windows. 31
|
39439364_p6
|
39439364
|
Intellectual disability subanalysis
| 3.881968 |
biomedical
|
Study
|
[
0.9989460110664368,
0.00020362137001939118,
0.0008504217257723212
] |
[
0.998067319393158,
0.0017094501527026296,
0.00016321451403200626,
0.0000600030216446612
] |
en
| 0.999995 |
Data were collected for 3674 participants. After deletion of duplicates and individuals with no recorded physical health data, a total sample of 3594 individuals remained. A total of 813 participants met the inclusion criteria for the autism group, and these were compared against 2781 controls.
|
39439364_p7
|
39439364
|
Results
| 2.505166 |
biomedical
|
Study
|
[
0.9972905516624451,
0.001106753945350647,
0.0016026610974222422
] |
[
0.9965388774871826,
0.002852797508239746,
0.0003803022555075586,
0.0002280847547808662
] |
en
| 0.999997 |
The mean age in the autism sample was 33.73 years (range 11–97 years, s.d. 13.45). The mean age in the control sample was 49.61 years (range 6–93 years, s.d. 18.74). The autism group was 41.9% female, 53.3% male and 4.4% gender variant/non-conforming/transgender male/transgender female; the control group was 65.8% female, 33.0% male and 0.9% gender variant/non-conforming/transgender male/transgender female. The male:female ratio in the autism group was 1.27:1.
|
39439364_p8
|
39439364
|
Results
| 2.790399 |
biomedical
|
Study
|
[
0.9971043467521667,
0.0005508205504156649,
0.0023448877036571503
] |
[
0.9955776929855347,
0.004118205979466438,
0.00018411864584777504,
0.00012004674499621615
] |
en
| 0.999998 |
On initial analysis with χ 2 -test ( Table 1 ), the autism sample was noted to have a significantly greater prevalence ( P < 0.001) of asthma (33.2 v. 16.3%), epilepsy (8.0 v. 1.8%), head injury (16.4 v. 3.2%), migraine headaches (33.2 v. 15.5%), inflammatory bowel disease (7.2 v. 3.1%), liver disease (1.9 v. 0.4%) and other autoimmune conditions (5.2 v. 2.2%) than the control sample. It was also observed that the control group had a significantly higher prevalence of cancer (6.3 v. 1.4%), heart disease (4.0 v. 1.5%), hypertension (16.7 v. 12.0%) and osteoarthritis (8.9 v. 5.5%) than the autism group. All other individual conditions showed no significant difference between the autism and control groups. Table 1 Frequency of physical health conditions in cases and controls, and results of the χ 2 analysis Physical health condition Frequency, n (%) Pearson χ 2 value P -value Autism Controls Asthma 264 (33.2) 453 (16.3) 108.902 <0.001 * Breast cancer 0 (0.0) 56 (2.0) Cancer (other) 11 (1.4) 175 (6.3) 30.741 <0.001 * COPD 4 (1.0) 29 (1.2) 0.178 0.673 Diabetes type 1 7 (0.9) 20 (0.7) 0.199 0.656 Diabetes type 2 35 (4.4) 128 (4.6) 0.082 0.775 Elevated lipids/cholesterol 72 (9.0) 252 (9.1) 0.003 0.959 Epilepsy/seizure disorder 63 (8.0) 50 (1.8) 76.301 <0.001 * Gastric/duodenal ulcers 28 (3.5) 68 (2.4) 2.833 0.092 Head injury 68 (16.4) 79 (3.2) 125.765 <0.001 * Heart disease 12 (1.5) 110 (4.0) 11.290 <0.001 * Hypertension 95 (12.0) 464 (16.7) 10.595 0.001 * Kidney disease 13 (1.6) 36 (1.3) 0.508 0.476 Liver disease 15 (1.9) 12 (0.4) 17.369 <0.001 * Memory loss (dementia) 10 (1.3) 17 (0.6) 3.511 0.061 Migraine headaches 262 (33.2) 428 (15.5) 123.641 <0.001 * Meningitis 6 (1.4) 21 (0.9) 1.220 0.269 Osteoarthritis 44 (5.5) 247 (8.9) 9.371 0.002 * Osteoporosis 14 (1.8) 54 (1.9) 0.127 0.722 Rheumatoid arthritis 26 (3.3) 73 (2.6) 0.981 0.322 Stroke/haemorrhage 11 (1.4) 39 (1.4) 0.002 0.968 Overactive thyroid/hyperthyroid 6 (0.8) 46 (1.7) 3.458 0.063 Underactive thyroid/hypothyroid 40 (5.0) 147 (5.3) 0.095 0.758 Inflammatory bowel disease 30 (7.2) 75 (3.1) 17.097 <0.001 * Other autoimmune condition 22 (5.2) 54 (2.2) 12.553 <0.001 * Human immunodeficiency virus, Parkinson's disease and multiple sclerosis are not included because of absolute frequency counts <5 allowing back-identification, and were not significantly different. COPD, chronic obstructive pulmonary disease. * Indicates P -values that are significant after Benjamini–Hochberg correction (α 0.029).
|
39439364_p9
|
39439364
|
Results
| 4.157804 |
biomedical
|
Study
|
[
0.9994256496429443,
0.000351259543094784,
0.00022320804418995976
] |
[
0.9991108775138855,
0.00023586931638419628,
0.0005838299985043705,
0.00006938229489605874
] |
en
| 0.999999 |
The initial logistic regression model controlling for age and gender found increased odds of 17 of the 28 physical health conditions in the autism group ( Table 2 ). The largest odds ratios were observed for liver disease (odds ratio 11.55, 95% CI 4.36–30.60, P < 0.001), head injury (odds ratio 5.03, 95% CI 3.30–7.67, P < 0.001), osteoporosis (odds ratio 5.16, 95% CI 2.37–11.25, P < 0.001), kidney disease (odds ratio 4.97, 95% CI 2.13–11.59, P < 0.001) and memory loss/dementia (odds ratio 4.93, 95% CI 1.80–13.51, P = 0.002). All 17 conditions remained significant after Benjamini–Hochberg correction. Table 2 Odds ratios and P -values from binomial logistic regression Physical health condition Model 1 Model 2 Adjusted odds ratio (95% CI) P -value Adjusted odds ratio (95% CI) P -value Asthma 2.36 (1.90–2.92) <0.001 * 2.36 (1.84–3.02) <0.001 * Cancer (other) 0.71 (0.34–1.47) 0.357 0.54 (0.20–1.43) 0.215 COPD 3.31 (0.88–12.48) 0.077 7.42 (1.79–30.77) 0.006 * Diabetes type 1 1.22 (0.43–3.41) 0.710 0.64 (0.14–3.00) 0.573 Diabetes type 2 2.74 (1.67–4.48) <0.001 * 2.32 (1.28–4.21) 0.005 * Elevated lipids/cholesterol 3.89 (2.68–5.66) <0.001 * 2.75 (1.74–4.34) <0.001 * Epilepsy/seizure disorder 3.84 (2.35–6.29) <0.001 * 3.44 (1.95–6.04) <0.001 * Gastric/duodenal ulcers 2.65 (1.47–4.77) 0.001 * 2.37 (1.18–4.76) 0.016 * Head injury 5.03 (3.30–7.67) <0.001 * 3.84 (2.30–6.39) <0.001 * Heart disease 0.91 (0.38–2.13) 0.818 0.57 (0.17–1.94) 0.367 Hypertension 2.38 (1.72–3.28) <0.001 * 2.11 (1.44–3.09) <0.001 * Kidney disease 4.97 (2.13–11.59) <0.001 * 4.96 (1.91–12.85) <0.001 * Liver disease 11.55 (4.36–30.60) <0.001 * 10.96 (3.72–32.25) <0.001 * Memory loss (dementia) 4.93 (1.80–13.51) 0.002 * 3.71 (1.13–12.22) 0.031 Migraine headaches 3.66 (2.92–4.59) <0.001 * 3.45 (2.66–4.48) <0.001 * Meningitis 0.85 (0.26–2.79) 0.783 1.06 (0.28–4.01) 0.928 Multiple sclerosis 0.70 (0.08–6.07) 0.743 0.00 (0.00) 0.991 Osteoarthritis 2.84 (1.81–4.43) <0.001 * 2.59 (1.50–4.46) <0.001 * Osteoporosis 5.16 (2.37–11.25) <0.001 * 4.66 (1.76–12.33) 0.002 * Rheumatoid arthritis 3.49 (1.88–6.47) <0.001 * 4.55 (2.33–8.88) <0.001 * Stroke/haemorrhage 1.93 (0.82–4.56) 0.134 1.22 (0.39–3.83) 0.730 Overactive thyroid/hyperthyroid 1.93 (0.74–5.02) 0.177 2.11 (0.68–6.55) 0.196 Underactive thyroid/hypothyroid 2.70 (1.75–4.17) <0.001 * 2.18 (1.28–3.70) 0.004 * Inflammatory bowel disease 2.97 (1.74–5.06) <0.001 * 2.10 (1.06–4.17) 0.035 Other autoimmune condition 3.79 (2.00–7.17) <0.001 * 3.38 (1.60–7.15) 0.001 * Model 1 controls for age and gender; model 2 controls for age, gender, antipsychotic use, mood stabiliser use and smoking status. COPD, chronic obstructive pulmonary disease. * Indicates P -values that are significant after Benjamini–Hochberg correction (α 0.029).
|
39439364_p10
|
39439364
|
Results
| 4.204603 |
biomedical
|
Study
|
[
0.999380350112915,
0.0003571161942090839,
0.00026246230117976665
] |
[
0.9990008473396301,
0.00027749835862778127,
0.0006456972914747894,
0.00007600881508551538
] |
en
| 0.999997 |
The medication- and smoking-adjusted logistic regression model, which controlled for age, gender, smoking, antipsychotic use and mood stabiliser use, found autism increased the odds for 18 out of 28 physical health conditions (the same conditions as the first model, with the addition of COPD). Only 16 remained significant after Benjamini–Hochberg correction ( Table 2 ). In this model, the greatest increase in odds ratios associated with autism were observed for liver disease (odds ratio 10.96, 95% CI 3.72–32.25, P < 0.001), COPD (odds ratio 7.42, 95% CI 1.79–30.77, P = 0.006), kidney disease (odds ratio 4.96, 95% CI 1.91–12.85, P < 0.001), osteoporosis (odds ratio 4.66, 95% CI 1.76–12.33, P = 0.002) and rheumatoid arthritis (odds ratio 4.55, 95% CI 2.33–8.88, P < 0.001). Neither regression model showed any condition to be significantly lower odds in the autism group.
|
39439364_p11
|
39439364
|
Results
| 4.096306 |
biomedical
|
Study
|
[
0.9994211196899414,
0.0003316624788567424,
0.00024726189440116286
] |
[
0.9993139505386353,
0.00019054303993470967,
0.00042950105853378773,
0.00006604161171708256
] |
en
| 0.999994 |
In the intellectual disability subanalysis logistic regression model, most conditions had similar odds ratios in both the isolated autism sample and the autism and intellectual disability sample ; however, fewer results reached significance in the intellectual disability subgroup. Several conditions demonstrated larger odds ratios in the autism and intellectual disability group . These included head injury (odds ratio 8.11, 95% CI 2.51–26.26), liver disease (odds ratio 22.28, 95% CI 3.75–132.51), osteoporosis (odds ratio 29.54, 95% CI 6.20–140.64) and hyperthyroidism (odds ratio 13.89, 95% CI 2.36–81.99). Table 3 Odds ratios from the subanalysis for isolated autism and autism with intellectual disability Physical health condition Isolated autism Autism and intellectual disability Adjusted odds ratio (95% CI) P -value Adjusted odds ratio (95% CI) P -value Asthma 2.32 (1.80–2.98) <0.001 * 2.94 (1.60–5.42) <0.001 * COPD 7.47 (1.81–30.87) 0.005 * 0.00 (0.00) 0.999 Diabetes type 1 0.59 (0.12–2.91) 0.516 1.07 (0.10–11.43) 0.956 Diabetes type 2 2.27 (1.24–4.15) 0.008 * 3.17 (0.86–11.76) 0.084 Elevated lipids/cholesterol 2.75 (1.73–4.36) <0.001 * 2.73 (0.94–7.90) 0.065 Epilepsy/seizure disorder 3.36 (1.89–5.96) <0.001 * 4.41 (1.41–13.82) 0.011 * Gastric/duodenal ulcers 2.42 (1.20–4.88) 0.013 * 1.45 (0.18–11.75) 0.729 Head injury 3.67 (2.19–6.17) <0.001 * 8.11 (2.51–26.26) <0.001 * Heart disease 0.60 (0.17–2.04) 0.409 0.00 (0.00–0.00) 0.997 Hypertension 2.11 (1.44–3.11) <0.001 * 2.13 (0.81–5.60) 0.124 Kidney disease 4.89 (1.87–12.81) 0.001 * 6.22 (0.69–55.62) 0.102 Liver disease 10.42 (3.48–31.17) <0.001 * 22.28 (3.75–132.51) <0.001 * Migraine headaches 3.52 (2.70–4.58) <0.001 * 2.63 (1.33–5.23) 0.006 * Osteoarthritis 2.60 (1.51–4.50) <0.001 * 2.33 (0.50–10.90) 0.282 Osteoporosis 3.90 (1.41–10.81) 0.009 * 29.54 (6.20–140.64) <0.001 * Rheumatoid arthritis 4.37 (2.21–8.66) <0.001 * 7.84 (1.66–37.04) 0.009 * Stroke/haemorrhage 1.16 (0.36–3.74) 0.800 2.11 (0.23–19.85) 0.512 Overactive thyroid/hyperthyroid 1.62 (0.47–5.56) 0.444 13.89 (2.36–81.99) 0.004 * Underactive thyroid/hypothyroid 2.15 (1.26–3.66) 0.005 * 2.81 (0.77–10.33) 0.120 Inflammatory bowel disease 2.05 (1.03–4.09) 0.042 3.78 (0.70–20.39) 0.122 COPD, chronic obstructive pulmonary disease. * Indicates P -values that are significant after Benjamini–Hochberg correction (α 0.029). Fig. 1 Graph showing the odds ratios from subanalysis for physical conditions in isolated autism or autism with intellectual disability, compared with the control population.
|
39439364_p12
|
39439364
|
Intellectual disability subanalysis
| 4.218241 |
biomedical
|
Study
|
[
0.9993863105773926,
0.0002971879148390144,
0.00031647650757804513
] |
[
0.9989781379699707,
0.0003066911594942212,
0.0006523996707983315,
0.00006278121145442128
] |
en
| 0.999997 |
In this study, we investigated the prevalence of physical health conditions in autism compared with a control sample without autism or mental health conditions. The results suggest that individuals with autism are at an increased risk of a range of physical health conditions compared with individuals without autism, across multiple organ systems, and the risk for some conditions is elevated in individuals with comorbid intellectual disability. This continues into older adulthood, with diagnoses such as osteoporosis and dementia being significantly more common in individuals with autism. Such findings are in keeping with results of previous studies, which also find increased odds of a range of physical health conditions in autism, and here we add to that literature with novel associations with previously unstudied physical health conditions. 6 – 9
|
39439364_p13
|
39439364
|
Discussion
| 4.080713 |
biomedical
|
Study
|
[
0.9994224309921265,
0.0003362740098964423,
0.00024131473037414253
] |
[
0.9991443157196045,
0.00015277629427146167,
0.0006311468314379454,
0.00007176245708251372
] |
en
| 0.999999 |
In accordance with existing research, our study found increased odds of epilepsy in autism, although the adjusted odds ratio of 3.44 was lower than reported in previous studies. 7 , 8 , 11 , 13 This may be accounted for by an unusually high frequency of epilepsy in the control sample (1.8% compared with an estimated 0.4–1.0% in the general population 32 ), likely resulting from the ascertainment and self-selection bias present in the wider NCMH database recruitment. It is difficult to compare to previous literature, as studies are heterogenous in their definition of epilepsy, ranging from strict international criteria to simply more than one seizure, and in this study, the questionnaire reported data on ‘epilepsy/seizure disorder’, which could be interpreted by participants to include non-epileptic seizure disorders. Despite this, our study and findings from existing research support the conclusion that epilepsy is significantly more common in autism. This study also recorded an increased risk of migraine headaches, which mirrors findings by Underwood et al 33 in an earlier, smaller sample of this cohort and similar findings in other studies, 10 , 13 , 29 although not all found a significant association. 7 , 11 , 12 Taken in combination, the results are in keeping with Pan et al 13 and Ward et al, 10 who found increased presence of neurological disorders in autism.
|
39439364_p14
|
39439364
|
Discussion
| 4.128234 |
biomedical
|
Study
|
[
0.9994750618934631,
0.0003024843754246831,
0.00022239591635297984
] |
[
0.9988542795181274,
0.00016043260984588414,
0.0009001462021842599,
0.00008509150939062238
] |
en
| 0.999998 |
Our findings strongly support increased odds of liver and kidney disease in individuals with autism, with a nearly 11-fold and five-fold increased risk, respectively; however, existing research on these diseases is sparse. Renal disease was found to have a slightly increased odds in autism in the study by Croen et al 7 (adjusted odds ratio 1.26, 99% CI 1–1.59), whereas Liu et al found no increased risk, 9 but studies are few. Hepatic disease was found to have a non-significant increased odds in adults with autism in the study by Croen et al (adjusted odds ratio 1.58, 99% CI 0.96–2.60), 7 whereas Shedlock et al found that children with autism were more likely to have non-alcoholic fatty liver disease/steatohepatitis than controls (odds ratio 2.74, 95% CI 2.06–3.65). 20 A recent study by Ward et al found increased odds of hepatic/renal disease in individuals with autism, 10 whereas Schott et al found decreased odds of hepatic (odds ratio 0.87 99% CI 0.83–0.91) and renal (odds ratio 0.82, 99% CI 0.80–0.84) disorders. 12 The current results suggest liver and kidney disease may be much more common in the autistic community, but further investigation is required to examine whether this replicates across the autistic population or could be accounted for by confounders in our sample.
|
39439364_p15
|
39439364
|
Discussion
| 4.090563 |
biomedical
|
Study
|
[
0.9995539784431458,
0.00022565500694327056,
0.00022027136583346874
] |
[
0.9980837106704712,
0.00019498563779052347,
0.0016501630889251828,
0.00007123057002900168
] |
en
| 0.999999 |
This analysis also found a trend toward increased metabolic diseases in autism, with increased odds of type 2 diabetes, elevated lipids and hypertension. This echoes findings by Shedlock et al 20 of increased prevalence of obesity, type 2 diabetes, hypertension, hyperlipidaemia and fatty liver disease in children with autism. Indeed, the findings of increased hypertension and hyperlipidaemia fit with most previous findings on these conditions, 7 , 8 , 11 , 12 , 14 , 20 suggesting they are more common in autism, but it is difficult to make conclusions about type 2 diabetes, given the mixed findings in existing literature. This study lacked obesity data, but obesity rates may also be higher in autism, 9 , 12 , 20 , 34 leading to questions of whether the increased prevalence of these conditions is driven by diet and weight, or if independent mechanisms are at play.
|
39439364_p16
|
39439364
|
Discussion
| 4.070879 |
biomedical
|
Study
|
[
0.99954754114151,
0.00021172570995986462,
0.00024074474640656263
] |
[
0.9986497759819031,
0.00018608063692227006,
0.001101600588299334,
0.00006263405521167442
] |
en
| 0.999998 |
The lack of significant findings for type 1 diabetes contrasted other autoimmune conditions, where we found increased odds in the autistic cohort, including rheumatoid arthritis, ‘other autoimmune disease’ and hypothyroidism (which, although not exclusively autoimmune in aetiology, is often caused by autoimmune disease). Overall, we add to the picture that autoimmune diseases are more prevalent in autism, 7 , 23 , 24 although specifics of individual autoimmune conditions are still unclear.
|
39439364_p17
|
39439364
|
Discussion
| 3.759547 |
biomedical
|
Study
|
[
0.999545156955719,
0.00016856862930580974,
0.0002861567190848291
] |
[
0.9917787909507751,
0.0010279143461957574,
0.0070619601756334305,
0.00013134385517332703
] |
en
| 0.999997 |
Findings of significantly increased odds of osteoarthritis, osteoporosis and nominally significant increased odds of memory loss/dementia were in keeping with Croen et al's findings on dementia, 7 and Hand et al's and Liu et al's findings of an increased risk of a range of age-related health conditions, including osteoarthritis, 9 osteoporosis and cognitive disorders. 8 Some researchers theorise that there may be overlapping pathophysiology between autism and neurodegenerative conditions like dementia and Parkinson's disease, including genetic commonalities, defects in neurotransmitters common to both conditions and changes in beta-amyloid seen in autism that may predispose to Alzheimer's disease. 35 This research is in its infancy and many of the proposed mechanisms are currently theoretical, but it may in time yield concrete evidence of shared predisposing mechanisms. Few studies on physical health consider the older autistic population, despite many conditions increasing in prevalence with age, or only occurring in older individuals. This study had a relatively small sample of people aged >50 years in the autism sample (91 individuals, 14.4% of the sample), but its findings provide support to the small number of existing studies and suggest this is an area requiring substantial further investigation, as greater numbers of older adults with autism are identified, potentially including longitudinal studies into older adulthood.
|
39439364_p18
|
39439364
|
Discussion
| 4.071861 |
biomedical
|
Study
|
[
0.9994919300079346,
0.00024093738466035575,
0.0002671240654308349
] |
[
0.9969022274017334,
0.00023176654940471053,
0.0027767159044742584,
0.00008928003808250651
] |
en
| 0.999997 |
Several findings in this study were novel. The study found increased odds of gastric/duodenal ulcers (adjusted odds ratio 2.37, 95% CI 1.18–4.76), which, to the authors’ knowledge, has not been documented in existing literature. This finding may relate to autistic behaviours, and therefore the potential impact of diet and medications like non-steroidal anti-inflammatory drugs on this relationship would be an interesting area for future exploration. Head injury is also not specifically explored in any existing epidemiological literature, despite being associated with autism. 36 This study found significantly increased odds of head injury in autism (adjusted odds ratio 3.84, 95% CI 2.30–6.39). Exploration of the nature of head injuries and the cause behind them (e.g. behavioural, neurological/motor deficits, self-injury) is warranted.
|
39439364_p19
|
39439364
|
Discussion
| 4.016479 |
biomedical
|
Study
|
[
0.9994282126426697,
0.00029554215143434703,
0.0002761799842119217
] |
[
0.9994871616363525,
0.000182124218554236,
0.0002703179488889873,
0.000060335707530612126
] |
en
| 0.999998 |
For some conditions, no significant association with autism was found. Breast cancer, cancer and heart disease did not show significant differences in odds ratios in our sample. Research on these conditions is mixed, with studies finding higher, 8 , 29 lower 11 , 12 and non-significant 7 odds of cancer compared with controls, and studies finding higher odds 7 – 9 , 21 or no difference 9 , 14 , 22 in heart disease in autism. This study also found no significant difference in stroke between groups, despite findings of increased odds in three previous studies. 7 – 9 Other conditions, including multiple sclerosis, Parkinson's disease and HIV, had low absolute prevalence in the data-set such that they are underpowered to detect any differences between groups, and we cannot comment on their prevalence in autism.
|
39439364_p20
|
39439364
|
Discussion
| 3.952464 |
biomedical
|
Study
|
[
0.9993281364440918,
0.00024644675431773067,
0.0004254548402968794
] |
[
0.999116837978363,
0.0001642151764826849,
0.0006623556837439537,
0.00005657157089444809
] |
en
| 0.999996 |
Although we could not do a direct comparison between individuals with and without intellectual disability, we observed that for osteoporosis, hyperthyroidism, liver disease and head injury, there are larger odds relative to the control group in the group with concurrent intellectual disability compared with individuals with autism without intellectual disability. This suggests that this population may be at increased risk of these conditions; however, confidence intervals for both groups overlapped. Several systematic reviews conclude that epilepsy prevalence in autism is higher in those with concurrent intellectual disability, 15 , 16 and one study in people aged >65 years found increased odds of several conditions, including osteoporosis, epilepsy, gastrointestinal disorders, thyroid disorders and cognitive disorders. 37 Another study, however, found the risk of most physical health conditions, compared with individuals without autism, was similar in adults with autism with and without intellectual disability. 9 This is a relative new area of study, limiting the ability to draw conclusions, especially given the size of our sample; however, these early results suggest that concurrent intellectual disability could confer increased risk for certain physical conditions.
|
39439364_p21
|
39439364
|
Discussion
| 3.981257 |
biomedical
|
Study
|
[
0.999480664730072,
0.000276614649919793,
0.00024265186220873147
] |
[
0.9965261816978455,
0.00024356343783438206,
0.0031403980683535337,
0.00008987293404061347
] |
en
| 0.999997 |
The mechanisms underpinning increased physical health problems in autism are likely multifactorial, involving shared aetiological factors, genetic influences and differences in behaviour and lifestyle. It is possible that barriers to accessing appropriate healthcare, such as communication difficulties, different symptom presentation and sensitivity to examinations or healthcare settings, may limit preventative care, leading to a greater development of chronic disease in this population. Poor diet and exercise, which can be a result of behavioural restriction, food hypersensitivities, social difficulties and comorbid motor/nervous conditions, may be another contributing mechanism. Weir et al 38 found that individuals with autism were less likely to meet recommendations for diet and exercise on most measures, and more likely to be underweight or obese than controls. These may increase the risk of a variety of negative health outcomes, such as cardiovascular disease. For epilepsy, there are many studies that evidence areas of overlap between the two conditions. This includes findings of dysfunction in GABA signalling and circuity, common associated genes and abnormal grey-white matter volumes in both autism and epilepsy. 39 A hypothesised biological mechanism underpinning both conditions may be one of excitation–inhibition imbalance in the neural circuitry, leading to hyperexcitability, with different genetic or neurodevelopmental changes found to be common to both epilepsy and autism converging to cause this imbalance. 39 It is possible that other shared mechanisms may exist, but are yet to be determined.
|
39439364_p22
|
39439364
|
Discussion
| 4.342949 |
biomedical
|
Study
|
[
0.9989820122718811,
0.0005589125212281942,
0.0004591286997310817
] |
[
0.8078935742378235,
0.0009734369232319295,
0.19070979952812195,
0.0004230891936458647
] |
en
| 0.999998 |
This was an analysis of existing data, collected over several years, using iteratively updated questionnaires. Several physical health outcomes, including COPD, head injury, HIV, meningitis, inflammatory bowel disease and ‘other autoimmune disease’ were not present in the earliest versions of the NCMH questionnaire, meaning that these had more missing data (20.2–20.7%) compared with other variables (0.6–1.3%). Furthermore, missing data in these variables was overrepresented in autism (approximately 50% missing data, compared with approximately 12% in the control sample). Thus, the findings on these conditions are less reliable and must be taken with more caution.
|
39439364_p23
|
39439364
|
Limitations
| 3.863197 |
biomedical
|
Study
|
[
0.9992626309394836,
0.00026394688757136464,
0.0004734439717140049
] |
[
0.9994625449180603,
0.0002804965479299426,
0.00020953678176738322,
0.00004735158654511906
] |
en
| 0.999997 |
Additional factors limited the generalisability and interpretation of our findings. Rates of physical health conditions in control individuals in the NCMH database may differ from the wider general population, as evidenced by the elevated rate of epilepsy in controls here. This may be caused by ascertainment bias through recruitment from healthcare environments such as NHS waiting rooms, and self-selection bias for individuals with health concerns. Second, some lifestyle factors that could affect physical health in autism, like body mass index and alcohol use, could not be controlled for. Meta-analytic evidence suggests that rates of obesity are significantly higher in autism, 34 and thus obesity, which is a risk factor for a variety of conditions, could be a confounder of the relationship found between autism and physical health. Other variables, like smoking, had differing responses between questionnaire versions, with some versions quantifying smoking and others only asking about smoking as a binary yes or no. For consistency, this study recorded smoking as a binary yes/no, taken from ‘lifetime ever smoked’, and this may miss the full effect of smoking on physical health.
|
39439364_p24
|
39439364
|
Limitations
| 4.057164 |
biomedical
|
Study
|
[
0.9994001388549805,
0.000297721941024065,
0.0003021541051566601
] |
[
0.9993242025375366,
0.00016898279136512429,
0.0004508273268584162,
0.000056046734243864194
] |
en
| 0.999997 |
Furthermore, our control sample was defined by lack of mental health conditions, meaning it is not fully reflective of the general population and, despite knowledge that there is a high burden of co-occurring mental health conditions (e.g. attention-deficit hyperactivity disorder, anxiety disorders, depressive disorders) among the autistic community, 4 our study could not control for the effects of these on physical health outcomes. In addition, clinical notes were not available to confirm diagnoses, so the study relies on self-report of clinician diagnosis, which reduces the reliability of our estimates.
|
39439364_p25
|
39439364
|
Limitations
| 3.364176 |
biomedical
|
Study
|
[
0.9988366961479187,
0.0004133335896767676,
0.0007500292849726975
] |
[
0.999031662940979,
0.00055748934391886,
0.00033955214894376695,
0.00007127347635105252
] |
en
| 0.999998 |
Our autism group had a much lower mean age and only a limited number of individuals over 50 years of age, which may explain the higher rates of cancer, heart disease, hypertension and osteoarthritis in the control group found in the initial χ 2 analysis. Although statistical modelling allowed us to control for the effects of age, the lack of older individuals with autism for comparison may mean we are missing the full picture around the occurrence of conditions that are more common with age.
|
39439364_p26
|
39439364
|
Limitations
| 2.759479 |
biomedical
|
Study
|
[
0.9982261061668396,
0.0005807264824397862,
0.0011931482004001737
] |
[
0.9957422614097595,
0.003690938698127866,
0.0003886134654749185,
0.00017809183918870986
] |
en
| 0.999997 |
Although the overall sample size in our study was not small, it was still small enough that rarer conditions with lower prevalences had too few cases for comparison, and our sample size was smaller than many comparable studies looking at a range of comorbidities. 6 – 9 , 11 , 12 Autistic and control samples were not matched on variables such as age and gender, in an attempt to maximise sample size for testing. This introduced additional issues, as controlling for variables such as age and gender is not optimal, and likely introduce biases. Furthermore, the high percentage of females in the control group (65.8%), as well as the higher percentage of females and lower percentage of individuals with intellectual disability in the autism group than would typically be expected, does raise questions over the representativeness of our sample, and again point toward ascertainment, response and self-selection biases. 40
|
39439364_p27
|
39439364
|
Limitations
| 4.007608 |
biomedical
|
Study
|
[
0.9993423819541931,
0.00027636546292342246,
0.0003811705100815743
] |
[
0.9994673132896423,
0.0002068490575766191,
0.00027598830638453364,
0.000049934460548684
] |
en
| 0.999997 |
Subanalysis was limited by the small sample of autistic individuals with intellectual disability. Our sample included only 86 in this group, with some conditions including fewer than 30 recorded responses. Several conditions had no cases or too few cases for meaningful analysis in this population. It is also likely that there is a selection bias away from individuals with intellectual disabilities in the wider NCMH sample, particularly those that are more severe, as several of the recruitment methods and the use of written questionnaires may better suit those without intellectual disability. To combat this, the NCMH has been undertaking targeted recruitment within the intellectual disability community by using appropriately adapted measures. Ultimately, concerns around size and representativeness of our sample for subanalysis limits any conclusions we can make, and emphasises that the experiences of individuals with intellectual disability are underresearched and underrepresented in the literature.
|
39439364_p28
|
39439364
|
Limitations
| 2.938204 |
biomedical
|
Study
|
[
0.9894587993621826,
0.0005948619800619781,
0.009946386329829693
] |
[
0.9912837743759155,
0.008106663823127747,
0.00048661112668924034,
0.00012292190513107926
] |
en
| 0.999998 |
This study draws attention to the increased physical health burden experienced by the autistic community, and adds to a growing field of research in this area with novel associations. This increased health burden is likely to have wide-ranging effects on quality of life and mental health, and may contribute to a risk of premature mortality in autism. 41 It is vital that research is carried out in all demographics of the autistic community, as some groups, such as older individuals (>50 years) or those with concurrent intellectual disability, are harder to sample and at risk of involuntary exclusion while also being more at risk. Further investigations are required to begin to understand the mechanisms behind the higher rates of physical health conditions observed in individuals with autism, alongside any mediating factors, with the aim of developing intervention and prevention strategies. Furthermore, there may be a role for enhanced health screening in individuals with autism, and future research should identify the value and focus of such screening. It is hoped that it will increase healthcare professionals’ awareness of physical health in autism, and encourage clinicians to have a lower threshold for considering physical illness when individuals present with changes in behaviour, as well as physical signs.
|
39439364_p29
|
39439364
|
Clinical implications and future directions
| 4.121413 |
biomedical
|
Study
|
[
0.9992341995239258,
0.0005158180720172822,
0.0002500877599231899
] |
[
0.995980978012085,
0.00039964658208191395,
0.003492877585813403,
0.00012639390479307622
] |
en
| 0.999997 |
Psychological well-being (PWB) is a multidimensional construct that encompasses various aspects of positive functioning, including self-acceptance, personal growth, purpose in life, environmental mastery, autonomy, and positive relationships with others. In recent years, there has been growing interest in understanding how different life roles and occupational status influence women's PWB, particularly in the context of evolving societal norms and expectations .
|
PMC11698264_p0
|
PMC11698264
|
Introduction
| 1.618078 |
other
|
Other
|
[
0.03281766548752785,
0.0012970197713002563,
0.96588534116745
] |
[
0.01022619754076004,
0.9633996486663818,
0.025673259049654007,
0.0007009018445387483
] |
en
| 0.999996 |
The comparison between housewives and working women's PWB has emerged as a significant area of research, especially in developing nations like India where traditional and modern roles often intersect . Working women often navigate multiple roles, balancing professional responsibilities with domestic duties, which can impact their PWB in complex ways . Studies have shown that employment can provide women with financial independence, social connections, and a sense of achievement, potentially enhancing their PWB . However, the relationship between employment status and PWB is not straightforward, as working women may experience additional stressors related to role conflicts and work-life balance .
|
PMC11698264_p1
|
PMC11698264
|
Introduction
| 1.286046 |
other
|
Other
|
[
0.01484827883541584,
0.0005013607442378998,
0.9846504330635071
] |
[
0.1267891675233841,
0.8561656475067139,
0.015789777040481567,
0.0012555262073874474
] |
en
| 0.999999 |
Conversely, housewives, who focus primarily on domestic responsibilities, may experience different challenges affecting their PWB . Recent research has highlighted that both groups face unique stressors that can influence their mental health and overall life satisfaction . A study conducted in North Bihar, India, found significant differences in PWB between housewives and working women, suggesting that occupational status may play a crucial role in determining mental health outcomes .
|
PMC11698264_p2
|
PMC11698264
|
Introduction
| 2.013676 |
biomedical
|
Study
|
[
0.7371508479118347,
0.0014963564462959766,
0.2613528072834015
] |
[
0.976814329624176,
0.02176135778427124,
0.0011699870228767395,
0.00025426968932151794
] |
en
| 0.999996 |
The COVID-19 pandemic has further complicated this dynamic, with studies indicating varying impacts on women's PWB based on their occupational status . The lockdown periods particularly affected women's mental health, regardless of their working status, highlighting the need for continued research in this area . Additionally, recent studies have emphasized the strong connection between PWB and physical health outcomes, suggesting that understanding these differences has important implications for overall health .
|
PMC11698264_p3
|
PMC11698264
|
Introduction
| 2.470041 |
biomedical
|
Study
|
[
0.9635981917381287,
0.0010939775966107845,
0.035307738929986954
] |
[
0.4036945104598999,
0.300845205783844,
0.29407572746276855,
0.0013845516368746758
] |
en
| 0.999997 |
The measurement of PWB has evolved significantly since Ryff's pioneering work, which established the theoretical framework for understanding its multiple dimensions . This multidimensional approach has become increasingly important in assessing the complex nature of well-being among different populations . Recent studies in various parts of India have employed this framework to examine PWB among women, considering factors such as marital status, employment, and socioeconomic conditions .
|
PMC11698264_p4
|
PMC11698264
|
Introduction
| 1.913383 |
other
|
Study
|
[
0.1943473517894745,
0.000802554830443114,
0.8048501014709473
] |
[
0.5461276173591614,
0.3990141451358795,
0.05364559218287468,
0.0012125693028792739
] |
en
| 0.999999 |
Recent studies have also highlighted the importance of socioeconomic factors and regional variations in determining PWB among women . The economic development and social changes in Tamil Nadu have created unique opportunities and challenges for both working women and housewives . While some research has explored general well-being patterns, there is limited understanding of how specific domains of PWB might differ between these groups in semi-urban areas like Perambalur district . Furthermore, the use of standardized assessment tools, particularly the Ryff's PWB Scale (PWBS), has proven valuable in capturing these nuanced differences across various populations . Understanding these domain-specific variations could be crucial for developing targeted interventions and support systems for women in different occupational roles .
|
PMC11698264_p5
|
PMC11698264
|
Introduction
| 2.085442 |
other
|
Study
|
[
0.19670094549655914,
0.0006354743964038789,
0.8026636242866516
] |
[
0.7991925477981567,
0.19310776889324188,
0.007103719748556614,
0.0005960142007097602
] |
en
| 0.999995 |
Despite the growing body of research in this area, there remains a need for region-specific studies that can account for local cultural and social contexts . Tamil Nadu in India, with its unique social and cultural landscape, presents an important setting for examining these differences. Furthermore, while several studies have explored overall PWB, fewer have focused on domain-specific differences between homemakers and employed women . Hence, this study aimed to compare the PWB among homemakers and employed women in Perambalur district, Tamil Nadu, India, by assessing both overall and domain-specific scores using the 18-item Ryff's PWBS (PWBS-18).
|
PMC11698264_p6
|
PMC11698264
|
Introduction
| 2.355763 |
other
|
Study
|
[
0.2958906292915344,
0.001030663843266666,
0.7030786871910095
] |
[
0.9974755644798279,
0.0021913843229413033,
0.00023820428759790957,
0.00009491119999438524
] |
en
| 0.999997 |
This was a community-based cross-sectional study, conducted in the field practice area of Dhanalakshmi Srinivasan Medical College and Hospital (DSMCH), a tertiary care teaching hospital in the Perambalur district of Tamil Nadu, India, between March 2023 and February 2024. The study protocol was approved by the Institutional Ethics Committee of DSMCH (approval number: IECHS/IRCHS/No. 392). Written informed consent was obtained from all participants after explaining the study objectives and procedures in their preferred language. Participant confidentiality was maintained throughout the study, and data were stored securely following standard research protocols.
|
PMC11698264_p7
|
PMC11698264
|
Materials and methods
| 3.00228 |
biomedical
|
Study
|
[
0.9969512224197388,
0.0011585218599066138,
0.001890235231257975
] |
[
0.9992375373840332,
0.0005107327597215772,
0.00014760365593247116,
0.00010412187839392573
] |
en
| 0.999997 |
Inclusion and exclusion criteria
|
PMC11698264_p8
|
PMC11698264
|
Materials and methods
| 2.023653 |
biomedical
|
Study
|
[
0.9684056639671326,
0.023301025852560997,
0.0082933334633708
] |
[
0.6835336685180664,
0.3025168478488922,
0.008536490611732006,
0.005413031205534935
] |
en
| 0.999997 |
The study included married women aged 18 years and above who were either homemakers or employed in any occupation. Employed women were defined as individuals who engaged in paid employment outside the home, whether in the formal or informal sectors, and met the specific criteria: (a) They must receive regular monetary compensation; (b) They must work at least 20 hours per week in economic activities outside the home. Homemakers included women who failed to meet these criteria. Women with diagnosed severe psychiatric illnesses, those unable to comprehend the questionnaire, and those unwilling to participate were excluded from the study.
|
PMC11698264_p9
|
PMC11698264
|
Materials and methods
| 2.277236 |
biomedical
|
Study
|
[
0.9257970452308655,
0.0017247877549380064,
0.07247824966907501
] |
[
0.9972122311592102,
0.002567478222772479,
0.0001306572085013613,
0.00008970189810497686
] |
en
| 0.999997 |
Sample size estimation and sampling method
|
PMC11698264_p10
|
PMC11698264
|
Materials and methods
| 1.76141 |
biomedical
|
Study
|
[
0.9840116500854492,
0.002917503472417593,
0.013070917688310146
] |
[
0.7370138764381409,
0.2591747045516968,
0.0019509729463607073,
0.001860424759797752
] |
en
| 0.999994 |
The sample size was calculated based on a previous study by Choudhary and Ahmed in India, which reported that 55% of working women demonstrated high levels of perceived PWB . The minimum required samples in the study were calculated using Cochran's formula for calculating sample size: n = Z 2 × p × (1-p) / d 2 , where p = 0.55 (prevalence) and d = 0.05 (5% absolute precision) at 95% confidence level (Z 2 =3.84). The calculation yielded a minimum required sample size of 380 participants. The study population comprised adult women aged 18 years and above from both urban and rural areas within the field practice area of DSMCH. Participants were recruited using convenience sampling methodology.
|
PMC11698264_p11
|
PMC11698264
|
Materials and methods
| 3.839193 |
biomedical
|
Study
|
[
0.9978249073028564,
0.00047963560791686177,
0.001695545157417655
] |
[
0.9994798302650452,
0.00040269087185151875,
0.00008165220788214356,
0.00003594165536924265
] |
en
| 0.999996 |
Data collection procedure
|
PMC11698264_p12
|
PMC11698264
|
Materials and methods
| 1.54202 |
biomedical
|
Other
|
[
0.9449723362922668,
0.006108541041612625,
0.048919107764959335
] |
[
0.49163809418678284,
0.49999701976776123,
0.005547149106860161,
0.0028177574276924133
] |
en
| 0.714283 |
An interviewer conducted face-to-face interviews with participants using a semi-structured questionnaire. The questionnaire included two main sections. The first section included sociodemographic information such as age, marital status, education, family type, residence, socioeconomic status, family size, spouse addiction status, occurrence of vital events, presence of chronic illness, and debt status. Socioeconomic status was classified based on the Modified BG Prasad classification . The second section includes the Ryff's PWBS-18. The interviews were conducted in the local language, with appropriate translation and back-translation procedures followed to ensure content validity. Each interview lasted approximately 30-45 minutes, and privacy was maintained throughout the data collection process.
|
PMC11698264_p13
|
PMC11698264
|
Materials and methods
| 2.645398 |
biomedical
|
Study
|
[
0.9763841032981873,
0.0014462576946243644,
0.022169537842273712
] |
[
0.9966006875038147,
0.0031416933052241802,
0.000154431167175062,
0.00010320714500267059
] |
en
| 0.999998 |
Study tool
|
PMC11698264_p14
|
PMC11698264
|
Materials and methods
| 1.419721 |
biomedical
|
Study
|
[
0.9622673392295837,
0.007864335551857948,
0.02986832894384861
] |
[
0.7573413848876953,
0.22158434987068176,
0.014965515583753586,
0.006108742672950029
] |
en
| 0.857142 |
PWB was assessed using Ryff's PWBS-18 . This scale measures six dimensions of PWB: autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, and self-acceptance . Each subscale item is evaluated on a six-point Likert scale, with 1 representing "strongly disagree" and 6 indicating "strongly agree." The components for the Autonomy subscale are Q15, Q17, and Q18. The items on the Environmental Mastery subscale are Q4, Q8, and Q9. The items for the Personal Growth subscale are Q11, Q12, and Q14. The items for the Positive Relations with Others subscale are Q6, Q13, and Q16. The items for the Purpose in Life subscale are Q3, Q7, and Q10. The items for the Self-Acceptance subscale are Q1, Q2, and Q5.
|
PMC11698264_p15
|
PMC11698264
|
Materials and methods
| 2.790293 |
biomedical
|
Study
|
[
0.5805113911628723,
0.001289455103687942,
0.4181991517543793
] |
[
0.901889443397522,
0.09645652770996094,
0.0013505028327926993,
0.00030359477386809886
] |
en
| 0.999998 |
Questions 1, 2, 3, 8, 9, 11, 12, 13, 17, and 18 need reverse scoring. Reverse-scored items are phrased contrary to the intended measurement of the scale. To compute subscale scores for each participant, aggregate the respondents' responses to the items of each subscale. Elevated scores indicate enhanced PWB across each category.
|
PMC11698264_p16
|
PMC11698264
|
Materials and methods
| 1.974355 |
biomedical
|
Other
|
[
0.7264454364776611,
0.001931147649884224,
0.2716234028339386
] |
[
0.3864559829235077,
0.6115066409111023,
0.0014238395961001515,
0.0006135876174084842
] |
en
| 0.999997 |
Table 1 shows the internal consistency reliability of the PWBS-18 was assessed using Cronbach's alpha (α = 0.682, standardized α = 0.709). While this reliability coefficient is slightly below the conventional threshold of 0.70, it is considered acceptable for a short-form version of an established scale . Therefore, all items were retained for analysis to maintain the scale's validated structure and enable comparability with existing literature.
|
PMC11698264_p17
|
PMC11698264
|
Materials and methods
| 3.516299 |
biomedical
|
Study
|
[
0.9422860145568848,
0.0005696607986465096,
0.0571444109082222
] |
[
0.9983764886856079,
0.0013576822821050882,
0.00022566232655663043,
0.000040176077163778245
] |
en
| 0.999999 |
Data analysis
|
PMC11698264_p18
|
PMC11698264
|
Materials and methods
| 0.918446 |
biomedical
|
Other
|
[
0.7303429245948792,
0.009679527953267097,
0.2599775493144989
] |
[
0.05484337732195854,
0.934048056602478,
0.007585848215967417,
0.0035227094776928425
] |
tl
| 0.999995 |
The data analysis was performed using IBM SPSS Statistics for Windows, Version 26.0 . The descriptive statistics for the sociodemographic variables were presented as frequency and percentage for categorical variables, and for continuous variables it was presented as mean and SD. The overall PWB scores and individual domain scores were presented as mean, standard deviation, median, and IQR for both homemakers and employed women. To compare the overall PWB scores between homemakers and employed women, an independent t-test was conducted. To compare the domain-specific PWB scores between homemakers and employed women, the Mann-Whitney U Test was conducted. For comparing the six domains of PWB between the groups, a one-way multivariate analysis of variance (MANOVA) was performed. The assumptions of MANOVA, including multivariate normality and homogeneity of variance-covariance matrices, were checked using Box's M test. MANOVA showed significant differences (p < 0.05), and subsequent post-hoc analyses using Bonferroni correction were conducted to identify which specific domains differ between the groups.
|
PMC11698264_p19
|
PMC11698264
|
Materials and methods
| 3.98221 |
biomedical
|
Study
|
[
0.9967304468154907,
0.00036895120865665376,
0.0029006008990108967
] |
[
0.9995903372764587,
0.00023897853679955006,
0.00013944137026555836,
0.00003122390990029089
] |
en
| 0.999996 |
Among the total participants (n= 308), 172 (55.8%) were homemakers, while 136 (44.2%) were employed women. Figure 1 illustrates the distribution of participants by their occupational status, showing a relatively balanced representation of both groups in the study population, with a slightly higher proportion of homemakers compared to employed women.
|
PMC11698264_p20
|
PMC11698264
|
Results
| 1.784382 |
biomedical
|
Study
|
[
0.8032634854316711,
0.0021197914611548185,
0.19461682438850403
] |
[
0.9864843487739563,
0.013029200956225395,
0.00028541492065414786,
0.0002010237076319754
] |
en
| 0.999997 |
Table 2 shows the sociodemographic characteristics of study participants based on occupational status. The mean age of homemakers (44.7 ± 14.06 years) was higher than employed women (37.12 ± 9.96 years). Among married women, 143 (56.7%) were homemakers and 109 (43.3%) were employed women. For education status, most illiterate women were homemakers (n=21, 77.8%), while the majority of graduates were employed (n=84, 75%). In terms of family type, joint families had more homemakers (n=88, 71.5%), while nuclear families showed a more balanced distribution with 84 (46.2%) homemakers and 98 (53.8%) employed women. Rural residence was more common among homemakers (n=146, 72.3%), while urban residence was predominant among employed women (n=80, 75.5%). Regarding socioeconomic status, higher classes (Class 1 and 2) had more employed women (n=31, 73.8% and n=49, 69%, respectively), while lower classes (Class 4 and 5) had more homemakers (n=75, 78.1% and n=25, 92.6%, respectively). Family size distribution showed that medium-sized families (four to six members) were most common in both groups, with 104 (52.5%) homemakers and 94 (47.5%) employed women. Additional characteristics showed an equal distribution of spouse addiction between groups (n=36, 50% each). Vital events occurred in 21 (43.8%) homemakers and 27 (56.3%) employed women. Chronic illness was present in 33 (54.1%) homemakers and 28 (45.9%) employed women, while debts were reported by 82 (59%) homemakers and 57 (41%) employed women.
|
PMC11698264_p21
|
PMC11698264
|
Results
| 3.679048 |
biomedical
|
Study
|
[
0.9446897506713867,
0.0013341999147087336,
0.05397612974047661
] |
[
0.9994025230407715,
0.00041478354251012206,
0.00013666543236467987,
0.000045993841922609136
] |
en
| 0.999998 |
Table 3 shows the comparison of overall and domain-specific PWB scores between homemakers and employed women. The overall PWB score was slightly higher among homemakers (mean: 69.35 ± 6.595) compared to employed women (mean: 68.21 ± 6.046). Among the six domains, autonomy scores were higher in employed women (mean: 13.44 ± 2.121; median: 14) compared to homemakers (mean: 12.88 ± 2.716; median: 14). Environmental mastery showed higher scores in homemakers (mean: 9.53 ± 3.879; median: 9) than employed women (mean: 8.88 ± 3.703; median: 9). Personal growth and positive relations with other domains showed slightly higher scores among homemakers (mean: 11.05 ± 2.423 and 12.2 ± 3.051, respectively) compared to employed women (mean: 10.65 ± 3.042 and 11.99 ± 3.607, respectively). Purpose in life was similar between groups, with employed women showing a marginally higher score (mean: 13.4 ± 2.715; median: 14) compared to homemakers (mean: 13.24 ± 2.207; median: 13.5). Self-acceptance was higher among homemakers (mean: 10.44 ± 1.488; median: 10) compared to employed women (mean: 9.86 ± 1.005; median: 10).
|
PMC11698264_p22
|
PMC11698264
|
Results
| 3.50524 |
biomedical
|
Study
|
[
0.5396735668182373,
0.0012530953390523791,
0.45907336473464966
] |
[
0.9983993172645569,
0.0012230508727952838,
0.0003198962949682027,
0.000057762743381317705
] |
en
| 0.999997 |
Table 4 shows the comparison of overall PWB scores between homemakers and employed women. Homemakers demonstrated a marginally higher mean score (69.35 ± 6.60) compared to employed women (68.21 ± 6.05). Levene's test confirmed the homogeneity of variances (F = 1.410, p = 0.236). The independent t-test revealed no statistically significant difference in overall PWB scores between the groups (p = 0.121).
|
PMC11698264_p23
|
PMC11698264
|
Results
| 3.149662 |
biomedical
|
Study
|
[
0.6754691004753113,
0.0010898482287302613,
0.3234409987926483
] |
[
0.9988227486610413,
0.0009682177333161235,
0.0001629352627787739,
0.0000461443851236254
] |
en
| 0.999997 |
Figure 2 shows the comparison of the distribution of PWB domain scores between homemakers (n = 172, shown in blue) and employed women (n = 136, shown in red) using a Mann-Whitney U test. The autonomy domain demonstrated similar median scores between groups, though homemakers showed greater score variability. Environmental mastery exhibited a broader distribution pattern across both groups, indicating diverse levels of mastery over their environments. Personal growth scores revealed comparable patterns between the groups, with subtle differences in their distributions. The positive relations with other domains showed similar central tendencies between homemakers and employed women, suggesting comparable social relationship qualities. Purpose in life scores display relatively symmetrical distributions for both groups, indicating similar levels of life purpose perception. Finally, the self-acceptance domain showed more concentrated scores among employed women compared to a wider distribution among homemakers, suggesting more varied levels of self-acceptance in the latter group. The mean rank values are displayed for each group across all domains, providing quantitative measures of central tendency.
|
PMC11698264_p24
|
PMC11698264
|
Results
| 3.713183 |
biomedical
|
Study
|
[
0.5461581945419312,
0.0010248598409816623,
0.4528169333934784
] |
[
0.9982428550720215,
0.0012358257081359625,
0.0004628215974662453,
0.000058570498367771506
] |
en
| 0.999997 |
Subsets and Splits
SQL Console for rntc/test-pp-aa
The query retrieves a sample of documents that are clinical cases with an educational score above 3, providing limited analytical value.
Clinical Cases Sample
Returns a sample of 100 clinical case documents, providing a basic overview of the document type's content.