Recruitment bias resulted in poorer overall health status in a community-based control group

Effects of portable pedal machines at work on lipoprotein subfraction profile in sedentary workers - the REMOVE study

BACKGROUND

Sedentary behaviour at work is a major cause of atherosclerosis, particularly in tertiary workers. However, no studies have ever assessed the effect of active workstation on lipoprotein subfraction profile. This study aimed to evaluate the effect of 12-week portable pedal machines (PPMs) on lipoprotein subfraction profile among healthy sedentary workers.

METHODS

Healthy administrative workers were randomized into an intervention group using PPMs for 12 weeks or a control group using normal-desk. Lipoprotein subfractions were assessed using Lipoprint® electrophoresis. Main outcomes were explored using mixed models with sensitivity analyses (four models).

RESULTS

We included 40 participants (43.7 ± 8.6 years old, 100% women, BMI 23.8 ± 3.4 kg/m2; sedentary time at work 7.7 ± 1.8 h/day). Groups did not differ at baseline in any outcomes. 32 participants finished the trial. Changes in lipoprotein subfractions were especially marked for LDL profile. There was an interaction time x group for all parameters related to LDL and their subfractions: total LDL-cholesterol (p = 0.012), LDL particle size (p = 0.027), large LDL subfractions 1 and 2 (p = 0.001), and small dense LDL subfractions 3 to 7 (p = 0.046), using the crude model. The interaction reflects difference in the direction of changes between groups. The LDL particle size significantly increased in the intervention group (from 271.9 ± 2.5 at t0 to 272.8 ± 1.9 Ångström at t1, p = 0.037) while it did not change in the control group (272.5 ± 1.7 at t0 to 271.8 ± 1.5Å at t1, p = 0.52). All interactions were constantly significant whatever the models. Influencing variables were mainly stress at work that was associated with an increase in total LDL-cholesterol (coefficient 3.15, 95CI 0.20 to 6.11 mg/dl, p = 0.038), and BMI that was associated with Large-LDL, Large-HDL, IDL-C and triglycerides.

CONCLUSIONS

Lipoprotein profile was improved after a 12-week PPMs intervention at work in healthy administrative workers. Changes were mainly showed for LDL and LDL subfractions. Lipoprotein profile was worsened by stress at work, BMI and age.

TRIAL REGISTRATION

NCT04153214.

Participant Bias in Community-Based Physical Activity Research: A Consistent Limitation?

Physical activity is a beneficial, yet complex, health behavior. To ensure more people experience the benefits of physical activity, we develop and test interventions to promote physical activity and its associated benefits. Nevertheless, we continue to see certain groups of people who choose not to, or are unable to, take part in research, resulting in "recruitment bias." In fact, we (and others) are seemingly missing large segments of people and are doing little to promote physical activity research to equity-deserving populations. So, how can we better address recruitment bias in the physical activity research we conduct? Based on our experience, we have identified 5 broad, interrelated, and applicable strategies to enhance recruitment and engagement within physical activity interventions: (1) gain trust, (2) increase community support and participation, (3) consider alternative approaches and designs, (4) rethink recruitment strategies, and (5) incentivize participants. While we recognize there is still a long way to go, and there are broader community and societal issues underlying recruitment to research, we hope this commentary prompts researchers to consider what they can do to try to address the ever-present limitation of "recruitment bias" and support greater participation among equity-deserving groups.

The incidence, baseline predictors, and outcomes of dementia in an incident cohort of Parkinson's disease and controls

BACKGROUND

There are few long-term data on the incidence, baseline predictors, and outcomes of dementia in Parkinson's disease (PD) from prospective community-based incident cohorts.

METHODS

The PINE study prospectively identified all incident PD patients in Aberdeen along with age-sex-matched, community-based controls who consented to standardized annual life-long follow-up. Each year, a clinical expert reviewed the diagnosis of PD and the presence of dementia according to DSM-IV-based criteria. Age-sex stratified incidence rates for dementia in PD and controls were calculated and compared with hazard ratios (HR) adjusted for age, sex, education, and socioeconomic status. Cox proportional-hazard modelling was used to assess baseline predictors for PD dementia and the influence of dementia on survival and institutionalization.

RESULTS

201 patients (mean age 72.6yrs) and 260 controls (mean age 75.4yrs) were followed for median 9.5 years. The incidence of dementia was 7.4 (PD) versus 2.1 (controls) per 100 person-years (adjusted HR 6.0, 95%CI 4.1-8.7), with a sixfold increase from under 60 to over 80 years in PD but no sex difference. Independent baseline predictors of PD dementia were older age at diagnosis, self-reported cognitive symptoms, dream enactment, lower MMSE scores, worse motor UPDRS scores, and the ApoE genotype. PD dementia increased the rates of subsequent death and institutionalization (32.0 and 26.9 per 100 person-years, respectively).

CONCLUSION

The incidence of dementia in PD is high, increases markedly with age, is increased in those with baseline subjective cognitive symptoms as well as other established risk factors, and is associated with high rates of death and institutionalization.

Early weight loss in parkinsonism predicts poor outcomes: Evidence from an incident cohort study

Objective:

To compare weight change over time in patients with Parkinson disease (PD), those with atypical parkinsonism, and matched controls; to identify baseline factors that influence weight loss in parkinsonism; and to examine whether it predicts poor outcome.

Methods:

We analyzed data from the Parkinsonism Incidence in North-East Scotland (PINE) study, an incident, population-based prospective cohort of parkinsonian patients and age- and sex-matched controls with annual follow-up. Mixed-model analysis described weight change in patients with PD, those with atypical parkinsonism, and controls. Baseline determinants of sustained clinically significant weight loss (>5% loss from baseline) and associations between early sustained weight loss and death, dementia, and dependency in parkinsonism were studied with Cox regression.

Results:

A total of 515 participants (240 controls, 187 with PD, 88 with atypical parkinsonism) were followed up for a median of 5 years. At diagnosis, atypical parkinsonian patients had lower body weights than patients with PD, who were lighter than controls. Patients with PD lost weight more rapidly than controls, and weight loss was most rapid in atypical parkinsonism. After multivariable adjustment for potential confounders, only age was independently associated with sustained clinically significant weight loss (hazard ratio [HR] for 10-year age increase 1.83, 95% confidence interval [CI] 1.44–2.32). Weight loss occurring within 1 year of diagnosis was independently associated with increased risk of dependency (HR 2.11, 95% CI 1.00–4.42), dementia (HR 3.23, 95% CI 1.40–7.44), and death (HR 2.23, 95% CI 1.46–3.41).

Conclusion:

Weight loss occurs in early parkinsonism and is greater in atypical parkinsonism than in PD. Early weight loss in parkinsonism has prognostic significance, and targeted dietary interventions to prevent it may improve long-term outcomes.

Medium-term prognosis of an incident cohort of parkinsonian patients compared to controls

Background

The best data on prognosis comes from population-based incident cohorts but few such cohorts exist for Parkinson's disease and atypical parkinsonism.

Methods

The PINE study is a prospective follow-up study of an incident cohort of people with degenerative or vascular parkinsonism and age-sex matched controls. Participants have annual follow-up from diagnosis until death with review of primary/secondary care records and linkage to the UK death register. Data are collected on survival, disability (dependency on others for activities of daily living) and institutionalization. Research criteria are used to guide the clinical diagnosis, which is updated annually. We compared all-cause mortality, disability and institutionalization in patients (subdivided by diagnosis) and controls, adjusted for important confounders.

Results

323 incident parkinsonian patients (199 Parkinson's disease, 124 atypical parkinsonism, mean age at diagnosis 75yrs) and 262 controls (mean age 75yrs) had 1349 and 1334 person-years follow-up respectively (maximum follow-up 10 years). All outcomes were worse in parkinsonian patients than controls, especially in atypical parkinsonism (adjusted mortality hazards ratios Parkinson's disease 2.49, 95%CI 1.72–3.58, atypical parkinsonism, 6.85, 95%CI 4.78–9.81). Median survival times for Parkinson's disease and atypical parkinsonism were 7.8 and 2.7 years respectively but were very age-dependent. At three years the rates of death or dependency were controls 21%, Parkinson's disease 46%, atypical parkinsonism 96% whilst overall institutionalization rates were 5%, 15% and 55% respectively.

Conclusion

The prognosis of Parkinson's disease and atypical parkinsonism in this unselected incident cohort was significantly worse than previously reported. This has important implications for patient management.

•

323 incident parkinsonian patients were followed for up to 10 years.

•

Mortality was 2.5–6.8 times higher in patients compared to controls.

•

After 3 years 46% (PD) and 96% (atypical parkinsonism) were dead or dependent.

•

14% (PD) and 55% (atypical parkinsonism) were institutionalized during follow-up.

•

Prognosis was significantly worse than in previous non-incident younger cohorts.

Analysis of recruitment in a pragmatic observational study on C-reactive protein point-of-care testing in primary care

INTRODUCTION

Failure to recruit all eligible study patients can lead to biased results. Little is known on selective patient recruitment in studies on implementation of diagnostic devices.

OBJECTIVES

The aim of this observational study was to measure recruitment of patients in an implementation study in primary care on use of point-of-care (POC) C-reactive protein (CRP) and to evaluate recruitment bias and its impact on the study endpoint.

METHODS

In a cross-sectional observational study on POC CRP implementation and related antibiotics prescribing, we compared included patients with all eligible patients to assess the representativeness of the included subjects. Eligible patients were adults presenting with acute cough in primary care between March and September 2012. The frequency of POC CRP testing and the proportion of prescribed antibiotics were compared between recruited and non-recruited patients. As measure of bias, odds ratios (ORs) with accompanying 95% confidence intervals (CIs) for the association between CRP level (<20 mg/l or not) and antibiotic prescribing were computed.

RESULTS

Of all 1473 eligible patients 348 (24%) were recruited. In recruited patients, POC CRP tests were conducted and antibiotics prescribed more frequently as compared to non-recruited patients (81% versus 6% and 44% versus 29%, respectively). The ORs were 18.2 (95%CI: 9.6-34.3), 30.5 (95%CI: 13.2-70.3) and 3.8 (95%CI: 0.9-14.8) respectively in all eligible patients, the recruited and the non-recruited patients.

CONCLUSION

Selective recruitment resulted in an overestimation of POC CRP test use and antibiotic prescribing.

Co-morbidity burden in Parkinson's disease: Comparison with controls and its influence on prognosis

Background

Many aspects of co-morbidity burden in Parkinson’s disease (PD) are unclear, but it may be an important predictor of prognosis or confounder of associations in epidemiological studies.

Objectives

To determine how best to assess co-morbidity burden in PD, to compare with matched controls, and investigate its association with prognostic outcomes.

Methods

Data from an incident, community-based cohort with prospective follow-up (the PINE study) were used (198 patients with PD and 151 controls). The reliability of three co-morbidity scales (the Charlson co-morbidity index (CCI), the Cumulative Illness Rating scale and a simple disease count) were evaluated. The association with mortality and development of dependency was assessed with Cox regression. The co-morbidity burden in PD and controls was compared at baseline and over 5 years of follow-up using linear mixed modelling.

Results

The CCI was more reliable and was an independent predictor of mortality with a time-dependent effect (hazard ratio = 1.27 [1.08–1.49] in first four years of follow-up; no significant association after four years). Associations between the other scales and mortality and between each scale and development of dependency were non-significant once adjusted for confounders. Co-morbidity burden was similar between cases and controls at baseline and there was no evidence of differential accrual of co-morbidity between patients and controls (p = 0.94).

Conclusions

The CCI is probably the better scale for measuring co-morbidity burden in PD. There were no differences between PD and controls. Co-morbidity burden at diagnosis was associated with mortality in the early part of the disease course, but not later.

•

The Charlson Index was a reliable and valid way to measure co-morbidity burden in PD.

•

There was no difference in co-morbidity burden between PD and matched controls.

•

Higher co-morbidity at diagnosis was associated with higher early mortality.

Appealing to altruism is not enough: motivators for participating in health services research

This pilot study sought to identify motivators and barriers to participating in a longitudinal survey; we interviewed patients and practitioners at a multidisciplinary primary care clinic where the proposed project would be based. While altruism motivates participation in medical research, we found that for many potential participants, the opportunity to benefit directly was the primary, and sometimes the only motive to participate or encourage participation in the research project. Patients often wanted direct feedback from their individual results, and they expected to provide consent before the results were forwarded to other parties such as their practitioners. Similarly, some practitioners were more likely to support the project if participation benefited patients directly. Other factors were also identified that influenced the acceptability and perceived risks and benefits of participating. More work is needed to understand these motivators and how patients might benefit directly from participating in health services research, especially when direct medical benefit is not possible.

Barriers to recruitment of professionals into a general practice childhood obesity program

Recruiting general practitioners to take part in research in primary care is important and challenging. This paper describes the process, barriers and achievements experienced by a research team whilst recruiting Divisions and general practitioners (GPs)/practices into a project related to management of obesity in children, which used Breakthrough Series methodology in the intervention arm. The research team chose to recruit GPs through Divisions of General Practice, because of the existing positive relationship between the research team and the Divisions, and the Divisions’ positive relationships with the GPs/practices in their area. The project aimed to recruit four Divisions of General Practice, who in turn would recruit a total of 24 GPs/practices. Four Divisions (two rural, two metropolitan) were recruited initially, but the two rural Divisions withdrew before the project commenced and were replaced by two metropolitan Divisions. As Divisions were unable to recruit sufficient GPs/practices, two additional strategies were used, but despite all efforts, only 17 practices were recruited. Reflection on recruitment failure focussed on reasons from the perspective of Divisions, GPs, other projects using the Breakthrough Series methodology with a chronic disease focus and other projects conducted in Australia where the focus was childhood obesity. The results showed that even Divisions who were enthusiastic about joining a project may be unable to recruit sufficient GPs and practices, for reasons including staff changeover and GP reluctance. Caution is suggested when setting recruitment targets where the condition is sensitive, is not the presenting problem, is not routinely tackled in general practice, involves minors, where treatment is perceived to be of doubtful effectiveness, or where major government policies may need to be considered.