A comparison of multiple imputation methods for missing data in longitudinal studies

The distinct longitudinal impact of pain catastrophizing on pain interference among youth living with sickle cell disease and chronic pain

Youth living with chronic sickle cell disease (SCD) pain are at risk for psychosocial distress and high levels of pain catastrophizing that contribute to functional impairment. This study aimed to identify the unique long-term impact of pain catastrophizing on pain impairment among youth with SCD. Youth with chronic SCD pain (N = 63, 10-18 years old, 58.3% female, 95.1% Black or African American) were recruited within comprehensive SCD clinics and completed a battery of measures at baseline and 4-months follow-up. A linear hierarchical regression examined baseline demographic and clinical characteristics (child SCD genotype, age, and average pain intensity), psychosocial functioning (anxiety, depression), and pain catastrophizing as predictors of pain interference at 4-months follow-up. Pain catastrophizing was the only unique predictor of pain interference at 4-months follow-up. Among youth with chronic SCD pain, pain catastrophizing warrants greater consideration as an important predictor that influences pain management and overall functioning.

Sexual and physical revictimization in U.S. military veterans

The present study examined revictimization, defined as sexual or physical assault in adulthood that followed a history of childhood maltreatment. We aimed to identify factors associated with revictimization over time in a group of U.S. military veterans deployed following the September 11, 2001, terrorist attacks (9/11). As revictimization is associated with multiple negative mental health outcomes in the literature, identifying risk and protective factors can aid in the prevention of revictimization and associated poor health outcomes among veterans. In this sample, the proportion of adult revictimization was 2.7% for men, 95% CI [2.0, 3.6] and 22.9% for women, 95% CI [20.5, 25.8]. Using multilevel logistic models, we found that women, β = 2.2, p < .001; Navy veterans, β = 1.5, p < .001; and participants who reported posttraumatic stress symptoms, β = 0.2, p = .028, were at significantly higher risk of revictimization across time compared to nonrevictimized counterparts. Social support while in the military was protective, β = -0.1, p < .001, against revictimization. In addition, childhood abuse experiences combined with characteristics such as female gender were related to an increased risk of revictimization during and following military service. The findings highlight opportunities for intervention and areas of strength within this population; social connection garnered during military service may serve as a protective factor against revictimization. Future research is needed to examine the role of social support in possibly lowering veterans' risk of revictimization over time, particularly for post-9/11 veterans struggling with transitioning from military to civilian life.

Multiple imputation approaches for handling incomplete three-level data with time-varying cluster-memberships

Three-level data arising from repeated measures on individuals clustered within higher-level units are common in medical research. A complexity arises when individuals change clusters over time, resulting in a cross-classified data structure. Missing values in these studies are commonly handled via multiple imputation (MI). If the three-level, cross-classified structure is modeled in the analysis, it also needs to be accommodated in the imputation model to ensure valid results. While incomplete three-level data can be handled using various approaches within MI, the performance of these in the cross-classified data setting remains unclear. We conducted simulations under a range of scenarios to compare these approaches in the context of an acute-effects cross-classified random effects substantive model, which models the time-varying cluster membership via simple additive random effects. The simulation study was based on a case study in a longitudinal cohort of students clustered within schools. We evaluated methods that ignore the time-varying cluster memberships by taking the first or most common cluster for each individual; pragmatic extensions of single- and two-level MI approaches within the joint modeling (JM) and the fully conditional specification (FCS) frameworks, using dummy indicators (DI) and/or imputing repeated measures in wide format to account for the cross-classified structure; and a three-level FCS MI approach developed specifically for cross-classified data. Results indicated that the FCS implementations performed well in terms of bias and precision while JM approaches performed poorly. Under both frameworks approaches using the DI extension should be used with caution in the presence of sparse data.

Effect of acupuncture on long-term outcomes in patients with post-stroke dysphagia

BACKGROUND

Acupuncture has been used to treat patients with post-stroke neurological dysfunction.

OBJECTIVE

The purpose of our observational study was to observe the long-term efficacy of acupuncture and investigate whether the acupuncture treatment could short the recovery time of patients with post-stroke dysphagia.

METHODS

Medical records were reviewed to select patients who met the inclusion criteria for post-stroke dysphagia. Exposure factor was defined as received acupuncture during inpatient. Clinical data were obtained at the 6-month follow-up. The primary outcome was the time to improve the score of Food Intake Level Scale (FILS, 0-10) by 3 grades. Cox regression models were used to assess the relationship between acupuncture and recovery of dysphagia.

RESULTS

In acupuncture group, the median time to achieve clinical improvement of dysphagia was 97 days (95% CI, 93-124) compared with 119 days (95% CI, 108-145) in control group, with a statistically significant difference between the two groups (HR = 1.48; 95% CI 1.14-1.92; P = 0.003). At 6 months, 78 patients (60.5%) in acupuncture group reached excellent function and 61 patients (47.3%) in control group (RR = 1.28; 95% CI, 1.02-1.62; P = 0.045). 106 patients (82.2%) in acupuncture group achieved favorable function and 91 patients (70.5%) in control group (RR = 1.17; 95% CI, 1.02-1.35; P = 0.039). The outcome of adjusted multivariable Cox regression models showed that there was a difference in the recovery time of dysphagia between groups, HR = 1.79, 95% CI 1.34-2.39. The rates of adverse events were similar in both groups.

CONCLUSIONS

Acupuncture can promote the recovery of post-stroke dysphagia, and has a better long-term efficacy. Besides, it can reduce the degree of disability and improve the quality of life.

Bias from questionnaire invitation and response in COVID-19 research: an example using ALSPAC

Background: Longitudinal studies are crucial for identifying potential risk factors for infection with, and consequences of, COVID-19, but relationships can be biased if they are associated with invitation and response to data collection. We describe factors relating to questionnaire invitation and response in COVID-19 questionnaire data collection in a multigenerational birth cohort (the Avon Longitudinal Study of Parents and Children, ALSPAC). Methods: We analysed online questionnaires completed between the beginning of the pandemic and easing of the first UK lockdown by participants with valid email addresses who had not actively disengaged from the study. We assessed associations of pre-pandemic sociodemographic, behavioural, anthropometric and health-related factors with: i) being sent a questionnaire; ii) returning a questionnaire; and iii) item response (for specific questions). Analyses were conducted in three cohorts: the index children born in the early 1990s (now young adults; 41 variables assessed), their mothers (35 variables) and the mothers' partners (27 variables). Results: Of 14,849 young adults, 41% were sent a questionnaire, of whom 57% returned one. Item response was >95%. In this cohort, 78% of factors were associated with being sent a questionnaire, 56% with returning one, and, as an example of item response, 20% with keyworker status response. For instance, children from mothers educated to degree-level had greater odds of being sent a questionnaire (OR=5.59; 95% CI=4.87-6.41), returning one (OR=1.60; 95% CI=1.31-1.95), and responding to items (e.g., keyworker status OR=1.65; 95% CI=0.88-3.04), relative to children from mothers with fewer qualifications. Invitation and response rates and associations were similar in all cohorts. Conclusions: These results highlight the importance of considering potential biases due to non-response when using longitudinal studies in COVID-19 research and interpreting results. We recommend researchers report response rates and factors associated with invitation and response in all COVID-19 observational research studies, which can inform sensitivity analyses.

Two Years of COVID-19 in Austria-Exploratory Longitudinal Study of Mental Health Outcomes and Coping Behaviors in the General Population

Since the long-term mental health impact of COVID-19 is not yet fully understood, the present study explored changes in mental health outcomes and pandemic-related coping behaviors across four pandemic stages. The main objective was to gain insights into the dynamics of mental health and coping, considering different pandemic features at different assessment waves. The final sample consisted of N = 243 adults from the Austrian general population. Data were collected at four timepoints (between June 2020 and December 2021) via LimeSurvey, an open-source online survey tool. Symptoms of posttraumatic stress disorder (PTSD), adjustment disorder (AD), anxiety, and depression were assessed using validated instruments: Primary Care PTSD Screen for DSM-5 (PC-PTSD-5), AD-New Module 8 (ADNM-8), and Patient Health Questionnaire (PHQ4). We also administered the Pandemic Coping Scale (PCS) to address pandemic-related coping behaviors. Cochran’s Q test and repeated measures ANOVAs were applied to assess changes over time. The results indicated that prevalence rates of AD (χ2(2) = 16.88, p = 0.001), depression (χ2(3) = 18.69, p < 0.001), and anxiety (χ2(3) = 19.10, p < 0.001) significantly changed across four assessment waves. Changes in mean scores of the assessed mental health outcomes were also observed. For pandemic-related coping, we found differences in the subscales: healthy lifestyle: F(3, 651) = 5.11, prevention adherence: F(2.73, 592.35) = 21.88, and joyful activities: F(3, 651) = 5.03. Taken together, our study showed a higher mental health burden in wintertime than in summertime, indicating an increased need for psychosocial support in times of stricter measures, higher incidences, and higher death rates. Furthermore, the observed decrease in adaptive coping behaviors suggests that easy-to-implement coping strategies should be actively promoted in order to maintain mental health during and in the aftermath of pandemics.

Duration of Care and Operative Time Are the Primary Drivers of Total Charges After Ambulatory Hip Arthroscopy: A Machine Learning Analysis

PURPOSE

To develop a machine learning algorithm to predict total charges after ambulatory hip arthroscopy and create a risk-adjusted payment model based on patient comorbidities.

METHODS

A retrospective review of the New York State Ambulatory Surgery and Services database was performed to identify patients who underwent elective hip arthroscopy between 2015 and 2016. Features included in initial models consisted of patient characteristics, medical comorbidities, and procedure-specific variables. Models were generated to predict total charges using 5 algorithms. Model performance was assessed by the root-mean-square error, root-mean-square logarithmic error, and coefficient of determination. Global variable importance and partial dependence curves were constructed to show the impact of each input feature on total charges. For performance benchmarking, the best candidate model was compared with a multivariate linear regression using the same input features.

RESULTS

A total of 5,121 patients were included. The median cost after hip arthroscopy was $19,720 (interquartile range, $12,399-$26,439). The gradient-boosted ensemble model showed the best performance (root-mean-square error, $3,800 [95% confidence interval, $3,700-$3,900]; logarithmic root-mean-square error, 0.249 [95% confidence interval, 0.24-0.26]; R² = 0.73). Major cost drivers included total hours in facility less than 12 or more than 15, longer procedure time, performance of a labral repair, age younger than 30 years, Elixhauser Comorbidity Index (ECI) of 1 or greater, African American race, residence in extreme urban and rural areas, and higher household and neighborhood income.

CONCLUSIONS

The gradient-boosted ensemble model effectively predicted total charges after hip arthroscopy. Few modifiable variables were identified other than anesthesia type; nonmodifiable drivers of total charges included duration of care less than 12 hours or more than 15 hours, operating room time more than 100 minutes, age younger than 30 years, performance of a labral repair, and ECI greater than 0. Stratification of patients based on the ECI highlighted the increased financial risk borne by physicians via flat reimbursement schedules given variable degrees of comorbidities.

LEVEL OF EVIDENCE

Level III, retrospective cohort study.

Trajectories of alcohol consumption in relation to all-cause mortality in patients with cardiovascular disease: a 35-year prospective cohort study

BACKGROUND AND AIMS

Research into alcohol consumption and cardiovascular disease (CVD) patients' prognosis has largely ignored the longitudinal dynamics in drinking behaviour. This study measured the association between alcohol consumption trajectories and mortality risk in CVD patients.

DESIGN

Prospective cohort study.

SETTING

UK-based Whitehall II Study.

PARTICIPANTS

A total of 1306 participants with incident non-fatal CVD (coronary heart disease/stroke) events.

MEASUREMENTS

Up to eight repeated measures of alcohol intake were available for each patient from the most recent assessment phase pre-incident CVD and all subsequent phases post-incident CVD, spanning up to three decades. Six trajectory groups of alcohol consumption were identified using group-based trajectory modelling and related to the risk of all-cause mortality, adjusting for demographics and changes in life-style and health status.

FINDINGS

Three hundred and eighty deaths were recorded during a median follow-up of 5 years after patients' last alcohol assessment. Compared with patients who consistently drank moderately (≤ 14 units/week), former drinkers had a greater risk of mortality (hazard ratio = 1.74, 95% confidence interval = 1.19-2.54) after adjustment for covariates. There was no significantly increased risk of mortality in long-term abstainers, reduced moderate drinkers, stable or unstable heavy drinkers. Cross-sectional analyses based only on drinking information at patients' last assessment found no significant differences in mortality risk for abstainers, former or heavy drinkers versus moderate drinkers.

CONCLUSIONS

Cardiovascular disease patients who consistently drink ≤ 14 units/week appear to have a similar risk of mortality to those who are long-term abstainers, which does not support a protective effect of moderate drinking on total mortality. Cardiovascular disease patients who stop drinking appear to have increased mortality risk compared with continuous moderate drinkers, but this may be linked to poor self-rated health before cardiovascular disease onset.

Clarifying the causes of consistent and inconsistent findings in genetics

As research in genetics has advanced, some findings have been unexpected or shown to be inconsistent between studies or datasets. The reasons these inconsistencies arise are complex. Results from genetic studies can be affected by various factors including statistical power, linkage disequilibrium, quality control, confounding and selection bias, as well as real differences from interactions and effect modifiers, which may be informative about the mechanisms of traits and disease. Statistical artefacts can manifest as differences between results but they can also conceal underlying differences, which implies that their critical examination is important for understanding the underpinnings of traits. In this review, we examine these factors and outline how they can be identified and conceptualised with structural causal models. We explain the consequences they have on genetic estimates, such as genetic associations, polygenic scores, family‐ and genome‐wide heritability, and describe methods to address them to aid in the estimation of true effects of genetic variation. Clarifying these factors can help researchers anticipate when results are likely to diverge and aid researchers' understanding of causal relationships between genes and complex traits.

Heterogeneity in Blood Biomarker Trajectories After Mild TBI Revealed by Unsupervised Learning

Concussions, also known as mild traumatic brain injury (mTBI), are a growing health challenge. Approximately four million concussions are diagnosed annually in the United States. Concussion is a heterogeneous disorder in causation, symptoms, and outcome making precision medicine approaches to this disorder important. Persistent disabling symptoms sometimes delay recovery in a difficult to predict subset of mTBI patients. Despite abundant data, clinicians need better tools to assess and predict recovery. Data-driven decision support holds promise for accurate clinical prediction tools for mTBI due to its ability to identify hidden correlations in complex datasets. We apply a Locality-Sensitive Hashing model enhanced by varied statistical methods to cluster blood biomarker level trajectories acquired over multiple time points. Additional features derived from demographics, injury context, neurocognitive assessment, and postural stability assessment are extracted using an autoencoder to augment the model. The data, obtained from FITBIR, consisted of 301 concussed subjects (athletes and cadets). Clustering identified 11 different biomarker trajectories. Two of the trajectories (rising GFAP and rising NF-L) were associated with a greater risk of loss of consciousness or post-traumatic amnesia at onset. The ability to cluster blood biomarker trajectories enhances the possibilities for precision medicine approaches to mTBI.

A Randomized controlled trial of the Effect of intraVenous iron on Anaemia in Malawian Pregnant women (REVAMP): Statistical analysis plan

Background: Anaemia affects more than half of Africa’s pregnancies. Standard care, with oral iron tablets, often fails to achieve results, with compliance and gastrointestinal side-effects being a significant issue. In recent years, intravenous iron formulations have become safe, effective, and quick to administer, allowing the complete iron requirements of pregnancy to be provided in one 15-minute infusion. The Randomized controlled trial of the Effect of intraVenous iron on Anaemia in Malawian Pregnant women (REVAMP) will evaluate whether a modern intravenous iron formulation, ferric carboxymaltose (FCM), given once during the second trimester is effective and safe in improving maternal and neonatal outcomes for treatment of moderate to severe anaemia in sub-Saharan Africa.  

The objective was to publish the detailed statistical analysis plan for the REVAMP trial prior to unblinding the allocated treatments and performing the analysis.  

Methods: REVAMP is a multicentre, two-arm, open-label, parallel-group randomized control trial (RCT) in 862 pregnant women in their second trimester. The trial statistician developed the statistical analysis plan in consultation with the trial management team based on the protocol, data collection forms, and study outcomes available in the blinded study database.  

Results: The detailed statistical analysis plan will support the statistical analyses and reporting of the REVAMP trial after unblinding the treatment allocations.  

Conclusions: A statistical analysis plan allows for transparency as well as reproducibility of reporting and statistical analyses.

Reciprocal Relationships Between Problematic Social Media Use, Problematic Gaming, and Psychological Distress Among University Students: A 9-Month Longitudinal Study

Background

The causal relationships between two specific types of problematic use in internet-related activities [i.e., problematic social media use (PSMU) and problematic gaming (PG)] and psychological distress remain controversial. The present study investigated the temporal relationships between PSMU, PG, and psychological distress (i.e., anxiety, depression) in university students.

Methods

Hong Kong and Taiwan university students [N = 645; n_male = 266; mean = 20.95 years (SD = 5.63)] were recruited for a survey study, with follow-ups at 3, 6, and 9 months after baseline assessment. The Bergen Social Media Addiction Scale, Internet Gaming Disorder Scale-Short Form, and the Hospital Anxiety and Depression Scale were used to assess studied variables. Demographics including age, physical characteristics (i.e., height, weight, and body mass index), and cigarette use were compared between participants who completed all the follow-ups and those who dropped out. Random intercept cross-lagged models were constructed to understand the reciprocal relationships between PSMU, PG, and psychological distress.

Results

No significant differences were found in age, physical characteristics, and cigarette use between participants who completed all the follow-ups and those who dropped out. Findings indicated that a high level of PSMU significantly increased the level of anxiety and a high level of anxiety significantly increased the level of PSMU. A high level of PSMU significantly increased the level of depression but the level of depression did not significantly affect the level of PSMU. A high level of PG significantly increased the level of anxiety, but the level of anxiety did not significantly affect the level of PG. A high level of depression significantly increased the level of PG, but the level of depression did not significantly affect the level of PG.

Conclusion

The patterns of the causal relationship between PIU and psychological distress variables differ. A reciprocal relationship was only found between the level of PSMU and the level of anxiety. Moreover, the longitudinal design found no differences in the waves in terms of gaming by the participants.

Using linear and natural cubic splines, SITAR, and latent trajectory models to characterise nonlinear longitudinal growth trajectories in cohort studies

BACKGROUND

Longitudinal data analysis can improve our understanding of the influences on health trajectories across the life-course. There are a variety of statistical models which can be used, and their fitting and interpretation can be complex, particularly where there is a nonlinear trajectory. Our aim was to provide an accessible guide along with applied examples to using four sophisticated modelling procedures for describing nonlinear growth trajectories.

METHODS

This expository paper provides an illustrative guide to summarising nonlinear growth trajectories for repeatedly measured continuous outcomes using (i) linear spline and (ii) natural cubic spline linear mixed-effects (LME) models, (iii) Super Imposition by Translation and Rotation (SITAR) nonlinear mixed effects models, and (iv) latent trajectory models. The underlying model for each approach, their similarities and differences, and their advantages and disadvantages are described. Their application and correct interpretation of their results is illustrated by analysing repeated bone mass measures to characterise bone growth patterns and their sex differences in three cohort studies from the UK, USA, and Canada comprising 8500 individuals and 37,000 measurements from ages 5-40 years. Recommendations for choosing a modelling approach are provided along with a discussion and signposting on further modelling extensions for analysing trajectory exposures and outcomes, and multiple cohorts.

RESULTS

Linear and natural cubic spline LME models and SITAR provided similar summary of the mean bone growth trajectory and growth velocity, and the sex differences in growth patterns. Growth velocity (in grams/year) peaked during adolescence, and peaked earlier in females than males e.g., mean age at peak bone mineral content accrual from multicohort SITAR models was 12.2 years in females and 13.9 years in males. Latent trajectory models (with trajectory shapes estimated using a natural cubic spline) identified up to four subgroups of individuals with distinct trajectories throughout adolescence.

CONCLUSIONS

LME models with linear and natural cubic splines, SITAR, and latent trajectory models are useful for describing nonlinear growth trajectories, and these methods can be adapted for other complex traits. Choice of method depends on the research aims, complexity of the trajectory, and available data. Scripts and synthetic datasets are provided for readers to replicate trajectory modelling and visualisation using the R statistical computing software.

Fair comparisons: Life course selection bias and the effect of father absence on US children

Studies have shown that father absence in opposite-gender couples has detrimental effects on children's wellbeing, net of selection bias. However, life course informed research suggests that the problem of selection bias may be more complex than currently thought. This paper shows the importance of nonparametrically adjusting for the trajectory of confounder covariates for the estimation of these effects. This paper uses data from the Fragile Families and Child Wellbeing Study to estimate the total effect of the departure of the biological father on children's wellbeing, as well as delayed or fade-out effects of this transition. The paper employs Bayesian additive regression trees, a machine learning and causal inference method suited for statistical models involving high-dimensional data sets. After adjusting for multiple time-invariant and -varying confounder covariates, as well as their history, estimates of father absence's effect on children's wellbeing are reduced substantially, a finding which may be referred to as life course selection bias. Results suggest early and middle childhood are not negatively affected by the departure of the biological father. Life course selection bias mostly affects estimates of this effect on adolescence, which is explained by children directly experiencing changes inparent's socioeconomic trajectories that lead to divorce or separation. This would not be the case when father absence is experienced in early childhood. Results suggest father absence is mostly a marker of life course cumulative socioeconomic disadvantage, not a cause of negative effects.

Virologic outcomes among adults with HIV using integrase inhibitor-based antiretroviral therapy

BACKGROUND

Integrase strand transfer inhibitor (InSTI)-based regimens have been recommended as first-line antiretroviral therapy (ART) for adults with HIV. But data on long-term effects of InSTI-based regimens on virologic outcomes remain limited. Here we examined whether InSTI improved long-term virologic outcomes compared with efavirenz (EFV).

METHODS

We included adults from the North American AIDS Cohort Collaboration on Research and Design who initiated their first ART regimen containing either InSTI or EFV between 2009 and 2016. We estimated differences in the proportion virologically suppressed up to 7 years of follow-up in observational intention-to-treat and per-protocol analyses.

RESULTS

Of 15 318 participants, 5519 (36%) initiated an InSTI-based regimen and 9799 (64%) initiated the EFV-based regimen. In observational intention-to-treat analysis, 81.3% of patients in the InSTI group and 67.3% in the EFV group experienced virologic suppression at 3 months after ART initiation, corresponding to a difference of 14.0% (95% CI 12.4-15.6). At 1 year after ART initiation, the proportion virologically suppressed was 89.5% in the InSTI group and 90.2% in the EFV group, corresponding to a difference of -0.7% (95% CI -2.1 to 0.8). At 7 years, the proportion virologically suppressed was 94.5% in the InSTI group and 92.5% in the EFV group, corresponding to a difference of 2.0% (95% CI -7.3 to 11.3). The observational per-protocol results were similar to intention-to-treat analyses.

CONCLUSIONS

Although InSTI-based initial ART regimens had more rapid virologic response than EFV-based regimens, the long-term virologic effect was similar. Our findings may inform guidelines regarding preferred initial regimens for HIV treatment.

The Role of Comorbid Symptoms in Perceived Stress and Sleep Problems in Adolescent ADHD

We examined perceived stress and sleep problems in adolescent ADHD and whether this varies as a function of ADHD presentation and sex. Further, we mapped structural associations between ADHD symptoms, comorbid symptoms, perceived stress, and sleep problems. Participants were 306 adolescents aged 13-19 years (66.8% females, 193 had an ADHD diagnosis, 113 were controls). Parents rated ADHD symptoms, all other constructs were self-rated. Adolescents with ADHD had elevated levels of perceived stress and sleep problems. Girls with ADHD reported the highest levels of perceived stress. Emotional symptoms mediated the effect of inattention whereas conduct problems mediated the effect of hyperactivity/impulsivity on stress and sleep. Perceived stress and sleep problems should be considered when mapping ADHD-related problems. Comorbid symptoms are potential intervention targets that may increase treatment response.

Disrupted Attention to Other's Eyes is Linked to Symptoms of ADHD in Childhood

Attention-deficit/hyperactivity disorder (ADHD) is associated with impaired social interaction. Other's eyes are important for understanding the social world. Here, we examined concurrent and longitudinal links between attention to other's eyes and symptoms of ADHD and comorbid externalizing and internalizing symptoms. Eighty-two 8 to 13-year-old children (40% with ADHD) participated. The latency to a first gaze shift to and away from the eye region of human faces, when primed to look at either the eyes or the mouth, was recorded with eye tracking. Parents rated ADHD, externalizing and internalizing symptoms at the time of testing and at 2-year follow-up. The results show that longer looking at the eyes before reorienting was specifically associated with concurrent and future symptoms of inattention, even when accounting for comorbid symptoms. We conclude that the temporal microstructure of attention to other's eyes is altered in children with symptoms of ADHD, which may contribute to social impairments.

Health-related quality of life, motivational regulation and Basic Psychological Need Satisfaction in Education Outside the Classroom: an explorative longitudinal pilot study

BACKGROUND

Given a suboptimal state of mental health among children, an urgent need exists to seek approaches related to health promotion in this population's settings, such as in schools. Increased health-related quality of life (HRQoL) and improved school motivation could be crucial for children's mental health. Based on self-determination theory, paths can be identified that could lead to such improvements by strengthening the basic psychological needs (BPN). This study aimed to examine the impact on and the relationships among HRQoL, school motivation and BPN within the promising concept of education outside the classroom (EOtC).

METHODS

In this exploratory study, we employed a between-subjects cohort study design with no blinding or randomisation. We surveyed fifth graders (mean = 10.1 years) attending EOtC (experimental group [EG], n = 25) and normal indoor lessons (control group, [CG], n = 41) at the beginning (T1) and end (T2) of a semester. We used the translations of validated questionnaires and established linear mixed-effects models to evaluate whether the students in EOtC show higher scores of HRQoL and school motivation and, whether the satisfaction of BPN of autonomy (PAut), competence (PCom), social relatedness with classmates (PSRC) and teachers (PSRT) show associations with these outcomes.

RESULTS

Regarding intrinsic and identified motivational regulation, results showed significant increases over time in the overall sample and significant higher scores in the EG than in the CG. For HRQoL, no group differences were found, but a significant decrease over time in the EG. Regarding possible associations between the outcomes and BPN, such could only be found between HRQoL and PSRC, but not for the other BPN and not for motivational regulation and BPN.

CONCLUSIONS

Without having been able to explain this on the basis of increased BPN values, our results show that EOtC can support improvements in specific regulation types of school motivation. This could contribute to an improvement in the mental health situation in children, as school represents a major stressor for them. Future steps in terms of researching HRQoL in this setting are discussed, as this pilot study does preliminary work for necessary examinations, e.g. in structural equation approaches.

Machine Learning Model Identifies Increased Operative Time and Greater BMI as Predictors for Overnight Admission After Outpatient Hip Arthroscopy

Purpose

The purposes of this study were to identify patient characteristics and risk factors for overnight admission following outpatient hip arthroscopy and to develop a machine learning algorithm that can effectively identify patients requiring admission following elective hip arthroscopy.

Methods

A retrospective review of a prospectively collected national surgical outcomes database was performed to identify patients who underwent elective outpatient hip arthroscopy from 2006 to 2018. Patients admitted overnight postoperatively were identified as those with length of stay of 1 or more days. Models were generated using random forest (RF), extreme gradient boosting (XGBoost), adaptive boosting (AdaBoost), elastic net penalized logistic regression, and an additional model was produced as a weighted ensemble of the four final algorithms.

Results

Overall, 1,276 patients were included. The median age was 43 years, and 64.2% (819) were female. Of the included patients, 109 (8.5%) required an overnight stay following elective outpatient hip arthroscopy. The most important factors for inpatient admission were increasing operative time, general anesthesia, age extremes, male gender, greater body mass index (BMI), American Society of Anesthesiologists classification >1, and the following preoperative lab values outside of normal ranges: sodium, platelet count, hematocrit, and leukocyte count. The ensemble model achieved the best performance based on discrimination assessed via internal validation (area under the curve = .71), calibration, and decision curve analysis. The model was integrated into a Web-based open-access application able to provide both personalized predictions and explanations.

Conclusion

A machine learning algorithm developed based on preoperative features identified increasing operative time, age extremes, greater BMI, sodium, hematocrit, platelets, and leukocyte count as the most important variables associated with inpatient admission with fair validity.

Trajectories of alcohol consumption up to 30 years before and after the diagnosis of cardiovascular diseases: a longitudinal case-control study of 12 502 participants

BACKGROUND

To examine the longitudinal trajectories of alcohol consumption prior to and following the diagnosis of cardiovascular diseases (CVD).

METHODS

We conducted a case-control study of 2501 incident cases of angina, myocardial infarction or stroke and 10 001 matched controls without the condition. Repeated measures of alcohol were centred on the date of diagnosis, spanning up to 30 years before and after CVD onset. Mean trajectories of weekly consumption were estimated using growth curve models.

RESULTS

For trajectories prior to diagnosis, mean volume of alcohol consumed among male cases increased over time, peaking at around 8 years before diagnosis at 95 (95% CI 60 to 130) g/week and declining afterwards. Trajectories following diagnosis showed mean consumption in male cases dropped from 87 (95% CI 54 to 120) g/week to 74 (95% CI 45 to 102) g/week after the date of diagnosis and then slightly rose to 78 (95% CI 40 to 116) g/week at the subsequent 3.5 years, before gradually declining to 31 (95% CI 2 to 61) g/week at 30 years after diagnosis. Mean consumption among female cases remained stable prior to diagnosis (at about 30 g/week), fell marginally to 25 (95% CI 20 to 30) g/week after the date of diagnosis and kept decreasing afterwards. Similar trajectories were obtained in cases and controls.

CONCLUSIONS

This is the first attempt to show how patients with CVD change their drinking volume over such a wide time span. Future research needs to establish insight into drinking behaviour in other ways (such as frequency and context) and address the impact of changes in drinking on patients with CVD.

Understanding Anterior Shoulder Instability Through Machine Learning: New Models That Predict Recurrence, Progression to Surgery, and Development of Arthritis

Background

Management of anterior shoulder instability (ASI) aims to reduce risk of future recurrence and prevent complications via nonoperative and surgical management. Machine learning may be able to reliably provide predictions to improve decision making for this condition.

Purpose

To develop and internally validate a machine-learning model to predict the following outcomes after ASI: (1) recurrent instability, (2) progression to surgery, and (3) the development of symptomatic osteoarthritis (OA) over long-term follow-up.

Study Design

Cohort study (prognosis); Level of evidence, 2.

Methods

An established geographic database of >500,000 patients was used to identify 654 patients aged <40 years with an initial diagnosis of ASI between 1994 and 2016; the mean follow-up was 11.1 years. Medical records were reviewed to obtain patient information, and models were generated to predict the outcomes of interest. Five candidate algorithms were trained in the development of each of the models, as well as an additional ensemble of the algorithms. Performance of the algorithms was assessed using discrimination, calibration, and decision curve analysis.

Results

Of the 654 included patients, 443 (67.7%) experienced multiple instability events, 228 (34.9%) underwent surgery, and 39 (5.9%) developed symptomatic OA. The ensemble gradient-boosted machines achieved the best performances based on discrimination (via area under the receiver operating characteristic curve [AUC]: AUC_recurrence = 0.86), AUC_surgery = 0.76, AUC_OA = 0.78), calibration, decision curve analysis, and Brier score (Brier_recurrence = 0.138, Brier_surgery = 0.185, Brier_OA = 0.05). For demonstration purposes, models were integrated into a single web-based open-access application able to provide predictions and explanations for practitioners and researchers.

Conclusion

After identification of key features, including time from initial instability, age at initial instability, sports involvement, and radiographic findings, machine-learning models were developed that effectively and reliably predicted recurrent instability, progression to surgery, and the development of OA in patients with ASI. After careful external validation, these models can be incorporated into open-access digital applications to inform patients, clinicians, and researchers regarding quantifiable risks of relevant outcomes in the clinic.

A Randomized controlled trial of the Effect of intraVenous iron on Anaemia in Malawian Pregnant women (REVAMP): Statistical analysis plan

Background: Anaemia affects more than half of Africa’s pregnancies. Standard care, with oral iron tablets, often fails to achieve results, with compliance and gastrointestinal side-effects being a significant issue. In recent years, intravenous iron formulations have become safe, effective, and quick to administer, allowing the complete iron requirements of pregnancy to be provided in one 15-minute infusion. The Randomized controlled trial of the Effect of intraVenous iron on Anaemia in Malawian Pregnant women (REVAMP) will evaluate whether a modern intravenous iron formulation, ferric carboxymaltose (FCM), given once during the second trimester is effective and safe in improving maternal and neonatal outcomes for treatment of moderate to severe anaemia in sub-Saharan Africa.   The objective was to publish the detailed statistical analysis plan for the REVAMP trial prior to unblinding the allocated treatments and performing the analysis.   Methods: REVAMP is a multicentre, two-arm, open-label, parallel-group randomized control trial (RCT) in 862 pregnant women in their second trimester. The trial statistician developed the statistical analysis plan in consultation with the trial management team based on the protocol, data collection forms, and study outcomes available in the blinded study database.   Results: The detailed statistical analysis plan will support the statistical analyses and reporting of the REVAMP trial after unblinding the treatment allocations.   Conclusions: A statistical analysis plan allows for transparency as well as reproducibility of reporting and statistical analyses.

Artificial Intelligence Predicts Cost After Ambulatory Anterior Cruciate Ligament Reconstruction

Purpose

To develop and internally validate a machine-learning algorithm to reliably predict cost after anterior cruciate ligament reconstruction (ACLR).

Methods

A retrospective review of the New York State Ambulatory Surgery and Services database was performed to identify patients who underwent elective ACLR from 2015 to 2016. Features included in initial models consisted of patient characteristics (age, sex, insurance status, income, medical comorbidities as classified by the Clinical Classifications Software diagnosis code) as well as intraoperative variables (type of anesthesia and procedure-specific factors). Models were generated to predict total charges using 4 algorithms: random forest, extreme gradient boost, elastic net penalized regression, and support vector machines with radial kernels. Training was performed with 10-fold cross-validation followed by internal validation via 0.632 bootstrapping. Model discriminative performance was assessed by area under the receiver operating characteristic curve, calibration, and the Brier score. Decision curve analysis was performed to demonstrate the net benefit of using the final model in practice.

Results

In total, 7,311 patients undergoing ambulatory ACLR were included. The random forest model demonstrated the best performance assessed via internal validation (area under the curve = 0.85), calibration, and the Brier score (0.208). Cost incurred was influenced by anesthesia type, operating room time, and number of chronic comorbidities. Decision curve analysis revealed a net benefit for use of the random forest model and the model was integrated into a web-based open-access application.

Conclusions

The random forest model predicted cost after ambulatory ACLR using a large, statewide database with good performance. The top variables found to predict increased charges were general anesthesia, operating room time, meniscal repair, self-pay insurance, patient neighborhood characteristics, and number of chronic conditions.

Level of Evidence

III, retrospective cohort study.

Handling Missing Responses in Psychometrics: Methods and Software

The presence of missing responses in assessment settings is inevitable and may yield biased parameter estimates in psychometric modeling if ignored or handled improperly. Many methods have been proposed to handle missing responses in assessment data that are often dichotomous or polytomous. Their applications remain nominal, however, partly due to that (1) there is no sufficient support in the literature for an optimal method; (2) many practitioners and researchers are not familiar with these methods; and (3) these methods are usually not employed by psychometric software and missing responses need to be handled separately. This article introduces and reviews the commonly used missing response handling methods in psychometrics, along with the literature that examines and compares the performance of these methods. Further, the use of the TestDataImputation package in R is introduced and illustrated with an example data set and a simulation study. Corresponding R codes are provided.

Addressing missing values in routine health information system data: an evaluation of imputation methods using data from the Democratic Republic of the Congo during the COVID-19 pandemic

BACKGROUND

Poor data quality is limiting the use of data sourced from routine health information systems (RHIS), especially in low- and middle-income countries. An important component of this data quality issue comes from missing values, where health facilities, for a variety of reasons, fail to report to the central system.

METHODS

Using data from the health management information system in the Democratic Republic of the Congo and the advent of COVID-19 pandemic as an illustrative case study, we implemented seven commonly used imputation methods and evaluated their performance in terms of minimizing bias in imputed values and parameter estimates generated through subsequent analytical techniques, namely segmented regression, which is widely used in interrupted time series studies, and pre-post-comparisons through paired Wilcoxon rank-sum tests. We also examined the performance of these imputation methods under different missing mechanisms and tested their stability to changes in the data.

RESULTS

For regression analyses, there were no substantial differences found in the coefficient estimates generated from all methods except mean imputation and exclusion and interpolation when the data contained less than 20% missing values. However, as the missing proportion grew, k-NN started to produce biased estimates. Machine learning algorithms, i.e. missForest and k-NN, were also found to lack robustness to small changes in the data or consecutive missingness. On the other hand, multiple imputation methods generated the overall most unbiased estimates and were the most robust to all changes in data. They also produced smaller standard errors than single imputations. For pre-post-comparisons, all methods produced p values less than 0.01, regardless of the amount of missingness introduced, suggesting low sensitivity of Wilcoxon rank-sum tests to the imputation method used.

CONCLUSIONS

We recommend the use of multiple imputation in addressing missing values in RHIS datasets and appropriate handling of data structure to minimize imputation standard errors. In cases where necessary computing resources are unavailable for multiple imputation, one may consider seasonal decomposition as the next best method. Mean imputation and exclusion and interpolation, however, always produced biased and misleading results in the subsequent analyses, and thus, their use in the handling of missing values should be discouraged.

Analytical challenges in estimating the effect of exposures that are bounded by follow-up time: experiences from the Blood Stream Infection-Focus on Outcomes study

Objective

To illustrate the challenges of estimating the effect of an exposure that is bounded by duration of follow-up on all-cause 28-day mortality, whilst simultaneously addressing missing data and time-varying covariates.

Study design and methods

BSI-FOO is a multicentre cohort study with the primary aim of quantifying the effect of modifiable risk factors, including time to initiation of therapy, on all-cause 28-day mortality in patients with bloodstream infection. The primary analysis involved two Cox proportional hazard models, first one for non-modifiable risk factors and second one for modifiable risk factors, with a risk score calculated from the first model included as a covariate in the second model. Modifiable risk factors considered in this study were recorded daily for a maximum of 28 days after infection. Follow-up was split at daily intervals from day 0 to 28 with values of daily collected data updated at each interval (i.e., one row per patient per day).

Analytical challenges

Estimating the effect of time to initiation of treatment on survival is analytically challenging since only those who survive to time t can wait until time t to start treatment, introducing immortal time bias. Time-varying covariates representing cumulative counts were used for variables bounded by survival time e.g. the cumulative count of days before first receipt of treatment. Multiple imputation using chained equations was used to impute missing data, using conditional imputation to avoid imputing non-applicable data e.g. ward data after discharge.

Conclusion

Using time-varying covariates represented by cumulative counts within a one row per day per patient framework can reduce the risk of bias in effect estimates. The approach followed uses established methodology and is easily implemented in standard statistical packages.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-021-01393-9.

The association of social frailty with intrinsic capacity in community-dwelling older adults: a prospective cohort study

Background

Social frailty is associated with poor health outcomes; however, its effects on healthy aging indicators have not been adequately investigated. This study assessed the longitudinal association between social frailty and the intrinsic capacity of community-dwelling older adults.

Methods

A total of 663 participants (56.7% women) aged ≥60 years from in Nagoya, Japan, were included in the study. The first measurement occurred in 2014, and annual follow-ups occurred until 2017. Social frailty was determined based on four items: financial difficulty, household status, social activity, and regular contact with others. A deficit score of 0 represented social robustness, 1 represented social prefrailty, and ≥ 2 represented social frailty. Intrinsic capacity was evaluated by the locomotion, cognition, psychological function, vitality, and sensory function domains. The longitudinal association was analyzed using generalized estimating equations.

Results

The prevalence of social prefrailty and social frailty at baseline was 31.2 and 6.3%, respectively. The social prefrailty group (β = − 0.132, P < 0.001) and social frailty group (β = − 0.258, P < 0.001) were associated with a greater reduction in the composite intrinsic capacity scores than the social robustness group, especially in the cognition, psychological function, and vitality domains. Men with social prefrailty/social frailty demonstrated a greater decrease in the psychological function domain score (− 0.512 vs. − 0.278) than women. Additionally, the cognition domain score only decreased in men in the social prefrailty/social frailty group (β = − 0.122, P = 0.016).

Conclusions

Social frailty was associated with intrinsic capacity and its subdomains longitudinally. Men with social frailty were more vulnerable than women to a decline in their psychological function and cognition domains. Therefore, the advanced management of social frailty is necessary to facilitate healthy aging.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-021-02466-6.

Association Between Anxiety and Cognitive Decline Over 12 Years in a Population-Based Cohort

BACKGROUND

Findings on the associations between anxiety and cognitive decline are mixed and often confounded.

OBJECTIVE

We studied whether anxiety symptoms were associated with the risk of cognitive decline after adequate adjustment of confounding factors.

METHODS

Our study consists of 2,551 community-dwelling older adults recruited between the ages of 60-64 years and followed up for 12 years in the PATH Through Life cohort study. Anxiety symptoms were measured using the Goldberg Anxiety Scale (GAS; range 0-9). General cognitive function, episodic memory, working memory, verbal intelligence, processing speed, and psychomotor speed were measured. Multilevel analyses were carried out to investigate the association between anxiety symptoms and cognitive decline over 12 years, taking into account confounding variables.

RESULTS

We did not find a significant association between baseline anxiety symptoms and cognitive decline over 12 years. Although some associations between anxiety symptoms with psychomotor speed (β= -0.04, 99% CI: -0.08, 0.00) and processing speed (β= -0.27, 99% CI: -0.48, -0.07) were found, these were attenuated after adjusting for depression. We also did not find an association between cumulative anxiety and decline in cognitive performance.

CONCLUSION

In this sample of cognitively healthy men and women aged 60 years and above, anxiety symptoms were not associated with the risk of cognitive decline. Long follow-up study time, appropriate selection of confounding factors, and estimating the effect of cumulative anxiety are important to establish the association between anxiety and cognitive symptoms.

Expectation-induced enhancement of pain, itch and quality of life in psoriasis patients: study protocol of a randomised controlled trial

INTRODUCTION

Experimental and clinical data demonstrate that skin diseases like psoriasis are affected by psychological factors and can be modulated by interventions other than conventional drug therapy. The expectation of patients towards the benefit of a forthcoming treatment as well as treatment pre-experiences have been demonstrated as crucial factors mediating placebo responses in inflammatory skin diseases. However, it is unknown whether and to what extent treatment outcomes of psoriasis patients under therapy with monoclonal antibodies like secukinumab can be experimentally modulated at subjective and physiological levels by modifying the expectation of patients via verbal instruction or prior experience.

METHODS AND ANALYSIS

Treatment expectations will be modulated in patients with moderate-to-severe psoriasis undergoing treatment with the anti-interleukin-17A monoclonal antibody secukinumab. Patients with a Psoriasis Area and Severity Index (PASI) >12 will be randomly allocated to one of three groups (N=40 each). As a standard schedule, patients in the pharmacological control group (group 1) will be treated weekly with 300 mg secukinumab, while patients in groups 2 and 3 will receive only 75 mg secukinumab (75% dose reduction) during all treatment weeks. In addition to the injections, patients in group 3 will ingest a novel tasting drink, with a cover story explaining that previous studies showed additional beneficial effects of this combination (drug and drink). Patients will be assessed and treated at nine visits over a 16-week period, during which the severity of pain and itch symptoms, skin lesions and quality of life will be analysed with standardised questionnaires and the PASI.

ETHICS AND DISSEMINATION

This study was approved by the Ethics committee of the Medical Faculty of the University Duisburg-Essen. Study outcomes will be published in peer-reviewed scientific journals.

Machine learning can reliably identify patients at risk of overnight hospital admission following anterior cruciate ligament reconstruction

PURPOSE

Overnight admission following anterior cruciate ligament reconstruction has implications on clinical outcomes as well as cost benefit, yet there are few validated risk calculators for reliable identification of appropriate candidates. The purpose of this study is to develop and validate a machine learning algorithm that can effectively identify patients requiring admission following elective anterior cruciate ligament (ACL) reconstruction.

METHODS

A retrospective review of a national surgical outcomes database was performed to identify patients who underwent elective ACL reconstruction from 2006 to 2018. Patients admitted overnight postoperatively were identified as those with length of stay of 1 or more days. Models were generated using random forest (RF), extreme gradient boosting (XGBoost), linear discriminant classifier (LDA), and adaptive boosting algorithms (AdaBoost), and an additional model was produced as a weighted ensemble of the four final algorithms.

RESULTS

Overall, of the 4,709 patients included, 531 patients (11.3%) required at least one overnight stay following ACL reconstruction. The factors determined most important for identification of candidates for inpatient admission were operative time, anesthesia type, age, gender, and BMI. Smoking history, history of COPD, and history of coagulopathy were identified as less important variables. The following factors supported overnight admission: operative time > 200 min, age < 35.8 or > 53.5 years, male gender, BMI < 25 or > 31.2 kg/m², positive smoking history, history of COPD and the presence of preoperative coagulopathy. The ensemble model achieved the best performance based on discrimination assessed via internal validation (AUC = 0.76), calibration, and decision curve analysis. The model was integrated into a web-based open-access application able to provide both predictions and explanations.

CONCLUSION

Modifiable risk factors identified by the model such as increased BMI, operative time, anesthesia type, and comorbidities can help clinicians optimize preoperative status to prevent costs associated with unnecessary admissions. If externally validated in independent populations, this algorithm could use these inputs to guide preoperative screening and risk stratification to identify patients requiring overnight admission for observation following ACL reconstruction.

LEVEL OF EVIDENCE

IV.

Community supported agriculture plus nutrition education improves skills, self-efficacy, and eating behaviors among low-income caregivers but not their children: a randomized controlled trial

Background

Adults and children in the U.S. consume inadequate quantities of fruit and vegetables (FV), in part, due to poor access among households with lower socioeconomic status. One approach to improving access to FV is community supported agriculture (CSA) in which households purchase a ‘share’ of local farm produce throughout the growing season. This study examined the effects of cost-offset (half-price) CSA plus tailored nutrition education for low-income households with children.

Methods

The Farm Fresh Foods for Healthy Kids (F3HK) randomized controlled trial in New York, North Carolina, Vermont, and Washington (2016–2018) assigned caregiver-child dyads (n = 305) into cost-offset CSA plus education intervention or control (delayed intervention) groups. Following one growing season of CSA participation, changes in children’s diet quality, body mass index (BMI), and physical activity; caregivers’ nutrition knowledge, attitudes, behaviors, and diet quality; and household food access and security were examined using multiple linear or logistic regression, with adjustment for baseline value within an intent-to-treat (ITT) framework in which missing data were multiply imputed.

Results

No significant net effects on children’s dietary intake, BMI, or physical activity were observed. Statistically significant net improvements were observed after one growing season for caregivers’ cooking attitudes, skills, and self-efficacy; FV intake and skin carotenoid levels; and household food security. Changes in attitudes and self-efficacy remained one-year after baseline, but improvements in caregiver diet and household food security did not. The number of weeks that participants picked up a CSA share (but not number of education sessions attended) was associated with improvements in caregiver FV intake and household food security.

Conclusions

Cost-offset CSA plus tailored nutrition education for low-income households improved important caregiver and household outcomes within just one season of participation; most notably, both self-reported and objectively measured caregiver FV intake and household food security improved. Households that picked up more shares also reported larger improvements. However, these changes were not maintained after the CSA season ended. These results suggest that cost-offset CSA is a viable approach to improving adult, but not child, FV intake and household food security for low-income families, but the seasonality of most CSAs may limit their potential to improve year-round dietary behavior and food security.

Trial registration

ClinicalTrials.gov. NCT02770196. Registered 5 April 2016. Retrospectively registered.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12966-021-01168-x.

Cannabis and tobacco use prior to pregnancy and subsequent offspring birth outcomes: a 20-year intergenerational prospective cohort study

There is increasing evidence that the life-course origins of health and development begin before conception. We examined associations between timing and frequency of preconception cannabis and tobacco use and next generation preterm birth (PTB), low birth weight (LBW) and small for gestational age. 665 participants in a general population cohort were repeatedly assessed on tobacco and cannabis use between ages 14-29 years, before pregnancy. Associations were estimated using logistic regression. Preconception parent (either maternal or paternal) daily cannabis use age 15-17 was associated with sixfold increases in the odds of offspring PTB (aOR 6.65, 95% CI 1.92, 23.09), and offspring LBW (aOR 5.84, 95% CI 1.70-20.08), after adjusting for baseline sociodemographic factors, parent sex, offspring sex, family socioeconomic status, parent mental health at baseline, and concurrent tobacco use. There was little evidence of associations with preconception parental cannabis use at other ages or preconception parental tobacco use. Findings support the hypothesis that the early life origins of growth begin before conception and provide a compelling rationale for prevention of frequent use during adolescence. This is pertinent given liberalisation of cannabis policy.

Coastal exposure and residents' mental health in the affected areas by the 2011 Great East Japan Earthquake and Tsunami

No previous study has ever explored the association between coastal exposure and the mental health of residents in a post-disaster context. Therefore, we aimed to confirm whether there was an association between sea visibility and coastal proximity and the mental health of coastal residents a devastating tsunami. We targeted 15 coastal municipalities located in the Miyagi Prefecture, and obtained data from a community-based cohort study. The baseline survey was initiated 2 years after the 2011 Great East Japan Earthquake and Tsunami and the secondary survey was initiated 6 years after the disaster. We applied multilevel mixed-effects models to the longitudinal data. Our outcome measure was the Kessler Psychological Distress Scale (K6) score. We assessed the data collected from 2,327 respondents on both surveys as of April 2018 for this ongoing cohort study. We found that neither sea visibility nor coastal proximity was significantly associated with the recovery of mental health after the disaster. However, we found a distinctive trend of mental health recovery in people who lived alone with a sea view, indicating that visibility of the sea had a negative effect on their mental health immediately after the GEJET, but that the negative effect was subsequently eliminated.

Multiple imputation for handling missing outcome data in randomized trials involving a mixture of independent and paired data

Randomized trials involving independent and paired observations occur in many areas of health research, for example in paediatrics, where studies can include infants from both single and twin births. Multiple imputation (MI) is often used to address missing outcome data in randomized trials, yet its performance in trials with independent and paired observations, where design effects can be less than or greater than one, remains to be explored. Using simulated data and through application to a trial dataset, we investigated the performance of different methods of MI for a continuous or binary outcome when followed by analysis using generalized estimating equations to account for clustering due to the pairs. We found that imputing data separately for independent and paired data, with paired data imputed in wide format, was the best performing MI method, producing unbiased point and standard error estimates for the treatment effect throughout. Ignoring clustering in the imputation model performed well in settings where the design effect due to the inclusion of paired data was close to one, but otherwise led to moderately biased variance estimates. Including a random cluster effect in the imputation model led to slightly biased point estimates for binary outcome data and variance estimates that were too small in some settings. Based on these results, we recommend researchers impute independent and paired data separately where feasible to do so. The exception is if the design effect due to the inclusion of paired data is close to one, where ignoring clustering may be appropriate.

Are Changes in Beliefs About Rumination and in Emotion Regulation Skills Mediators of the Effects of Internet-Delivered Cognitive-Behavioral Therapy for Depression and Anxiety? Results from a Randomized Controlled Trial

Background

Internet-delivered cognitive behavior therapy (iCBT) is effective in treating anxiety and depression. Research on how these interventions operate is scarce. This study explored whether emotion regulation skills and positive beliefs about rumination were affected by iCBT and if these constructs mediated changes in depression and anxiety.

Methods

This is a secondary analysis of a pragmatic randomized waitlist-controlled trial testing the effectiveness of supported iCBT. Adults with at least mild symptoms of depression or anxiety were included. Depression (PHQ-9), anxiety (GAD-7), positive beliefs about rumination (PBRS-A) and two emotion regulation skills: cognitive reappraisal (ERQ-A) and expressive suppression (ERQ-S), were measured at baseline and 8-weeks post-treatment.

Results

The analyses included 358 participants, 71% were female. Median age was 29. Linear mixed models showed statistically significant differences along ERQ-A in favor of the iCBT group (b = 1.83, SE = 0.82, p = .026). Mediation analyses showed reductions in depression (b = 0.31, SE = 0.15, p = 0.043) and anxiety symptoms (b = 0.27, SE = 0.14, p = 0.057) were partially mediated by gains in ERQ-A. No effects were observed for PBRS-A and ERQ-S.

Conclusions

These results align with findings from face-to-face therapy and add to the scarce literature on mediators of effects of iCBT, contributing to the understanding of how these interventions operate. Since mediator and outcome variables were measured at the same time, partial mediation results should be interpreted with caution since the study design did not account for temporality and therefore causality effects cannot be confirmed.

Trial Registration

ISRCTN91967124.

Bias from questionnaire invitation and response in COVID-19 research: an example using ALSPAC

Background: Longitudinal studies are crucial for identifying potential risk factors for infection with, and consequences of, COVID-19, but relationships can be biased if they are associated with invitation and response to data collection. We describe factors relating to questionnaire invitation and response in COVID-19 questionnaire data collection in a multigenerational birth cohort (the Avon Longitudinal Study of Parents and Children, ALSPAC). Methods: We analysed online questionnaires completed between the beginning of the pandemic and easing of the first UK lockdown by participants with valid email addresses who had not actively disengaged from the study. We assessed associations of pre-pandemic sociodemographic, behavioural, anthropometric and health-related factors with: i) being sent a questionnaire; ii) returning a questionnaire; and iii) item response (for specific questions). Analyses were conducted in three cohorts: the index children born in the early 1990s (now young adults; 41 variables assessed), their mothers (35 variables) and the mothers' partners (27 variables). Results: Of 14,849 young adults, 41% were sent a questionnaire, of whom 57% returned one. Item response was >95%. In this cohort, 78% of factors were associated with being sent a questionnaire, 56% with returning one, and, as an example of item response, 20% with keyworker status response. For instance, children from mothers educated to degree-level had greater odds of being sent a questionnaire (OR=5.59; 95% CI=4.87-6.41), returning one (OR=1.60; 95% CI=1.31-1.95), and responding to items (e.g., keyworker status OR=1.65; 95% CI=0.88-3.04), relative to children from mothers with fewer qualifications. Invitation and response rates and associations were similar in all cohorts. Conclusions: These results highlight the importance of considering potential biases due to non-response when using longitudinal studies in COVID-19 research and interpreting results. We recommend researchers report response rates and factors associated with invitation and response in all COVID-19 observational research studies, which can inform sensitivity analyses.

Longitudinal pre-exposure prophylaxis (PrEP) acceptability, initiation and adherence among criminal justice-involved adults in the USA: the Southern PrEP Cohort Study (SPECS) protocol

INTRODUCTION

HIV prevalence among criminal justice (CJ)-involved adults is five times higher than the general population. Following incarceration, CJ-involved individuals experience multilevel barriers to HIV prevention. Pre-exposure prophylaxis (PrEP) is a widely available, daily medication efficacious in preventing HIV. Little is known about PrEP knowledge, acceptability, initiation and sustained use among CJ-involved persons or about how these outcomes vary by multilevel factors. The Southern Pre-Exposure Prophylaxis Study (SPECS) will investigate barriers and facilitators for PrEP initiation and sustained use among CJ-involved adults, building a foundation for PrEP interventions for this underserved population.

METHODS AND ANALYSIS

SPECS uses a mixed-methods sequential design, including a multisite, prospective cohort study in three southern states-North Carolina, Florida and Kentucky-and subsequent qualitative interviews. HIV-negative adults clinically indicated for PrEP with CJ-involvement in the past year (n=660; 220 per site)-will be recruited for four quantitative interviews separated by 6 months, with 18 months of follow-up. Interviews will measure CJ involvement, substance use, sexual behaviours, PrEP acceptability and use, healthcare access and utilisation, support systems and psychological and emotional well-being. We will estimate probabilities of PrEP acceptability and use in a CJ-involved population using descriptive and multivariable analyses. After the follow-up, a subsample that never initiated PrEP, initiated but did not sustain PrEP or sustained PrEP will be asked to participate in a qualitative interview to contextualise their experiences and decisions around PrEP. An inductive approach will guide qualitative analyses.

ETHICS AND DISSEMINATION

PrEP initiation and sustained use rates are unknown among CJ-involved adults. This research will identify individual, social and structural factors that predict PrEP initiation and use. Data generated from the study have the potential to guide research and the development and tailoring of PrEP interventions to CJ-involved populations and provide context to HIV-related outcomes for those with CJ experiences.

Incidence of Thyroid Dysfunction Facing Metabolic Syndrome: A Prospective Comparative Study with 9 Years of Follow-Up

BACKGROUND

Studies assessing thyroid hormones in metabolic syndrome (MetS) patients are contradictory. Also, the effect of MetS on thyroid function over time is not yet evaluated. This study investigated the prevalence and incidence of thyroid dysfunction (TD) as well as time trends of thyroid hormones in subjects with and without MetS, during a 10-year follow-up in Tehranian adult population.

METHODS

This is a prospective cohort study conducted in the framework of Tehran Thyroid Study on 5,786 subjects aged ≥20 years: 4,905 eligible participants entered the study after excluding those with corticosteroid or radioactive iodine use, pregnancy, thyrotropin (TSH) <0.1 and >10 mU/L, and missing data. Physical examinations were performed and serum concentrations of TSH, free thyroxine (FT4), thyroid peroxidase antibody (TPOAb), fasting plasma glucose, insulin, and lipid profile were assessed at baseline and 3-year intervals during the follow-up. MetS was defined according to the Joint Interim Statement Definition.

RESULTS

At baseline, there were no difference in median serum concentrations of FT4 and TSH between MetS and non-MetS group after adjusting for age, sex, BMI, smoking, and TPOAb positivity. Although there was higher risk of overt (42%) and subclinical hypothyroidism (16%) in MetS compared with non-MetS subjects, no significant difference was observed in adjusted ORs for any TD between 2 groups. There were also no significant differences in time trends of TSH, FT4, TPOAb positivity, and incidence rates of TDs between MetS and non-MetS groups during 10 years, after adjustment for age, sex, BMI, smoking status, and TPOAb positivity.

CONCLUSION

MetS is not associated with thyroid hypofunction considering other important confounders such as age, sex, smoking, BMI, and TPOAb positivity. There is also no difference in the trend of thyroid hormones and incidence of TD between MetS and non-MetS subjects during a 10-year follow-up.

Can attachment predict core and comorbid symptoms of attention-deficit/hyperactivity disorder beyond executive functions and emotion regulation?

OBJECTIVE

Previous findings propose an association between attachment and symptoms of attention-deficit/hyperactivity disorder (ADHD). However, the predictive ability of attachment beyond deficits in executive functioning (EF) and emotion regulation is understudied.

METHOD

Using a dimensional perspective, we examined the longitudinal role of attachment on ADHD and comorbid symptoms, accounting for EF and emotion regulation. The sample consisted of 84 children (aged 8-13 years), oversampled for ADHD symptoms (42% had a diagnosis of ADHD). We assessed attachment with the Child Attachment Interview, EF with laboratory tests, and emotion regulation with parental ratings. Parents and teachers rated symptoms at baseline (T1) and at follow-up 2 years later (T2).

RESULTS

Attachment insecurity was positively correlated with ADHD symptoms at T2 but with no unique contribution to symptoms beyond EF and emotion regulation. In contrast, poor EF and emotion regulation contributed to more ADHD and ODD/CD symptoms at T2. Poor emotion regulation contributed to more anxiety at T2.

CONCLUSIONS

The results have important implications for understanding the mechanisms underpinning symptom expression in middle childhood/early adolescence and may guide the search for tailored interventions to reduce and prevent symptoms.

PRACTITIONER POINTS

Executive functions and emotion regulation should be explored as targets for intervention in personalized treatment. The current results do not support attachment as a target for intervention at a group level, although this does not rule out that individuals/families with attachment difficulties may benefit from training programmes promoting secure attachment bonds.

Burden of caregivers who care for oldest-old parents with disability: A cross-sectional study

OBJECTIVE

To describe the characteristics of oldest-old Chinese with disability and their adult-child caregivers, and the extent to which these characteristics were associated with caregiver burden.

METHODS

The study was based on 168 pairs of disabled oldest-old adults and their adult-child caregivers, derived from the Chinese Longitudinal Healthy Longevity Survey. Descriptive analyses of care recipients' and caregivers' characteristics were conducted respectively, in reference to caregiver burden. Statistically significant characteristics identified in these bivariate analyses were then jointly evaluated in multiple linear regression models with caregiver burden as the outcome.

RESULTS

Care recipients positive emotion status [(β = -0.227 (-0.412, -0.042)], multiple chronic disease [(β = 0.513 (0.081, 0.945)], and caregivers spent more caregiving time [(β = 0.225 (0.061, 0.389)] were main factors associated with caregiver burden.

CONCLUSION

Adult-children caregivers perceived heavier burden if care recipients had low positive emotions, had multiple chronic diseases, and caregivers spent more time caregiving.

Framework for the treatment and reporting of missing data in observational studies: The Treatment And Reporting of Missing data in Observational Studies framework

Missing data are ubiquitous in medical research. Although there is increasing guidance on how to handle missing data, practice is changing slowly and misapprehensions abound, particularly in observational research. Importantly, the lack of transparency around methodological decisions is threatening the validity and reproducibility of modern research. We present a practical framework for handling and reporting the analysis of incomplete data in observational studies, which we illustrate using a case study from the Avon Longitudinal Study of Parents and Children. The framework consists of three steps: 1) Develop an analysis plan specifying the analysis model and how missing data are going to be addressed. An important consideration is whether a complete records' analysis is likely to be valid, whether multiple imputation or an alternative approach is likely to offer benefits and whether a sensitivity analysis regarding the missingness mechanism is required; 2) Examine the data, checking the methods outlined in the analysis plan are appropriate, and conduct the preplanned analysis; and 3) Report the results, including a description of the missing data, details on how the missing data were addressed, and the results from all analyses, interpreted in light of the missing data and the clinical relevance. This framework seeks to support researchers in thinking systematically about missing data and transparently reporting the potential effect on the study results, therefore increasing the confidence in and reproducibility of research findings.

Association of sex differences in dementia risk factors with sex differences in memory decline in a population-based cohort spanning 20-76 years

Sex differences in late-life memory decline may be explained by sex differences in dementia risk factors. Episodic memory and dementia risk factors were assessed in young, middle-aged and older adults over 12 years in a population-based sample (N = 7485). For men in midlife and old age, physical, cognitive and social activities were associated with less memory decline, and financial hardship was associated with more. APOE e4 and vascular risk factors were associated with memory decline for women in midlife. Depression, cognitive and physical activity were associated with memory change in older women. Incident midlife hypertension (β = - 0.48, 95% CI - 0.87, - 0.09, p = 0.02) was associated with greater memory decline in women and incident late-life stroke accounted for greater memory decline in men (β = - 0.56, 95% CI - 1.12, - 0.01), p = 0.05). Women have fewer modifiable risk factors than men. Stroke and hypertension explained sex differences in memory decline for men and women respectively.

E-health StandingTall balance exercise for fall prevention in older people: results of a two year randomised controlled trial

OBJECTIVE

To test whether StandingTall, a home based, e-health balance exercise programme delivered through an app, could provide an effective, self-managed fall prevention programme for community dwelling older people.

DESIGN

Assessor blinded, randomised controlled trial.

SETTING

Older people living independently in the community in Sydney, Australia.

PARTICIPANTS

503 people aged 70 years and older who were independent in activities of daily living, without cognitive impairment, progressive neurological disease, or any other unstable or acute medical condition precluding exercise.

INTERVENTIONS

Participants were block randomised to an intervention group (two hours of StandingTall per week and health education; n=254) or a control group (health education; n=249) for two years.

MAIN OUTCOME MEASURES

The primary outcomes were the rate of falls (number of falls per person year) and the proportion of people who had a fall over 12 months. Secondary outcomes were the number of people who had a fall and the number who had an injurious fall (resulting in any injury or requiring medical care), adherence, mood, health related quality of life, and activity levels over 24 months; and balance and mobility outcomes over 12 months.

RESULTS

The fall rates were not statistically different in the two groups after the first 12 months (0.60 falls per year (standard deviation 1.05) in the intervention group; 0.76 (1.25) in the control group; incidence rate ratio 0.84, 95% confidence interval 0.62 to 1.13, P=0.071). Additionally, the proportion of people who fell was not statistically different at 12 months (34.6% in intervention group, 40.2% in control group; relative risk 0.90, 95% confidence interval 0.67 to 1.20, P=0.461). However, the intervention group had a 16% lower rate of falls over 24 months compared with the control group (incidence rate ratio 0.84, 95% confidence interval 0.72 to 0.98, P=0.027). Both groups had a similar proportion of people who fell over 24 months (relative risk 0.87, 95% confidence interval 0.68 to 1.10, P=0.239), but the proportion of people who had an injurious fall over 24 months was 20% lower in the intervention group compared with the control group (relative risk 0.80, 95% confidence interval 0.66 to 0.98, P=0.031). In the intervention group, 68.1% and 52.0% of participants exercised for a median of 114.0 min/week (interquartile range 53.5) after 12 months and 120.4 min/week (38.6) after 24 months, respectively. Groups remained similar in mood and activity levels. The intervention group had a 0.03 (95% confidence interval 0.01 to 0.06) improvement on the EQ-5D-5L (EuroQol five dimension five level) utility score at six months, and an improvement in standing balance of 11 s (95% confidence interval 2 to 19 s) at six months and 10 s (1 to 19 s) at 12 months. No serious training related adverse events occurred.

CONCLUSIONS

The StandingTall balance exercise programme did not significantly affect the primary outcomes of this study. However, the programme significantly reduced the rate of falls and the number of injurious falls over two years, with similar but not statistically significant effects at 12 months. E-health exercise programmes could provide promising scalable fall prevention strategies.

TRIAL REGISTRATION

ACTRN12615000138583.

Practical strategies for handling breakdown of multiple imputation procedures

Multiple imputation is a recommended method for handling incomplete data problems. One of the barriers to its successful use is the breakdown of the multiple imputation procedure, often due to numerical problems with the algorithms used within the imputation process. These problems frequently occur when imputation models contain large numbers of variables, especially with the popular approach of multivariate imputation by chained equations. This paper describes common causes of failure of the imputation procedure including perfect prediction and collinearity, focusing on issues when using Stata software. We outline a number of strategies for addressing these issues, including imputation of composite variables instead of individual components, introducing prior information and changing the form of the imputation model. These strategies are illustrated using a case study based on data from the Longitudinal Study of Australian Children.

Effects of neurological music therapy on behavioural and emotional recovery after traumatic brain injury: A randomized controlled cross-over trial

Traumatic brain injury (TBI) causes deficits in executive function (EF), as well as problems in behavioural and emotional self-regulation. Neurological music therapy may aid these aspects of recovery. We performed a cross-over randomized controlled trial where 40 persons with moderate-severe TBI received a 3-month neurological music therapy intervention (2 times/week, 60 min/session), either during the first (AB, n = 20) or second (BA, n = 20) half of a 6-month follow-up period. The evidence from this RCT previously demonstrated that music therapy enhanced general EF and set shifting. In the current study, outcome was assessed with self-report and caregiver-report questionnaires performed at baseline, 3-month, 6-month, and 18-month stages. The results showed that the self-reported Behavioural Regulation Index of the Behaviour Rating Inventory of Executive Function (BRIEF-A) improved more in the AB than BA group from baseline to 3-month stage and the effect was maintained in the 6-month follow-up. No changes in mood or quality of life questionnaires were observed. However, a qualitative content analysis of the feedback revealed that many participants experienced the intervention as helpful in terms of emotional well-being and activity. Our results suggest that music therapy has a positive effect on everyday behavioural regulation skills after TBI.

Examining longitudinal associations between self-reported depression, anxiety and stress symptoms and hair cortisol among mothers of young children

BACKGROUND

Maternal mental health is critically important given its impacts on both women's and children's outcomes. Hair cortisol concentrations (HCC) may provide insight into physiological processes underpinning mental health. This study investigated associations between mothers' self-reported mental health symptoms and their HCC at 1, 2 and 3 years postpartum.

METHODS

Longitudinal study of Australian mothers recruited for their experience of adversity in pregnancy ('right@home' trial, N=722). Mental health symptoms were self-reported using the Depression, Anxiety and Stress Scales (DASS). Associations between DASS total and subscale scores and HCC were estimated using linear regression and generalized estimating equation (GEE) models, examining associations: at each age; across all ages (multivariate GEE); and with persistence of high symptom severity. Missing data were addressed using multiple imputation.

RESULTS

546/722 (76%) women provided at least one hair sample (71% at 1, 61% at 2, 49% at 3 years). Associations between DASS total or subscale scores and HCC were not evident across time points. Only dichotomized high depression symptom severity was associated with higher HCC in the GEE models (β=0.12, p=0.04). There was no evidence of associations between persistence of high DASS symptom severity and HCC at 3 years.

LIMITATIONS

The DASS measured self-reported symptoms for the preceding week whereas HCC captured average cortisol over three months. Associations amongst mothers experiencing adversity may not represent patterns in the general population.

CONCLUSIONS

Considered in context with existing literature, these findings suggest that HCC provides limited insight into the mental health of mothers experiencing adversity across the early postpartum years.