Accounting for bias due to selective attrition: the example of smoking and cognitive decline

Potential Role of Community-Based Healthcare System Data in Research on Survivors of Adolescent and Young Adult Cancer

PURPOSE

We sought to examine issues of generalizability in research on adolescent and young adult (AYA) cancer survivorship that relies on using community-based healthcare delivery system data.

METHODS

Individuals aged 15 to 39 diagnosed with cancer between 1992 and 2006 were identified using data from community-based healthcare systems in California and Seattle. Loss to follow-up was defined as the first disenrollment (the end) of membership in the healthcare systems after cancer. Censoring occurred at death or study end (2009). We used Kaplan-Meier analysis to quantify follow-up, and multiple Cox regression to examine the association of follow-up loss with demographic and cancer characteristics.

RESULTS

Of 6828 eligible AYAs, most (93%) were aged between 20 and 39 years at diagnosis; 62% were female and 39% were non-White. Solid tumors accounted for 81% of diagnoses. The majority (89%) of patients continued to be members of the healthcare systems and available for follow-up 1 year after diagnosis. Approximately 60% remained enrolled 5 years after diagnosis. Loss to follow-up was associated with younger age at diagnosis, male gender, and African American or Hispanic race/ethnicity.

CONCLUSION

Data from community-based healthcare delivery systems offer an efficient way to identify large and diverse samples of AYA-onset cancer survivors. Differential loss to follow-up can threaten the generalizability of results from these studies and should be assessed quantitatively. Healthcare system data offer an alternative to studies requiring direct contact with participants.

Dietary patterns and relationship to obesity-related health outcomes and mortality in adults 75 years of age or greater

BACKGROUND

The prevalence of obesity-related adverse health outcomes is increasing among older adults. Because it is thought that nutrition plays an important role in successful aging, there has been considerable interest in the association between dietary patterns of older adults and obesity-related health outcomes.

OBJECTIVE

This study examined the association between dietary patterns and mortality and prevalence of obesity-related health outcomes, namely cardiovascular disease (CVD), type 2 diabetes mellitus, hypertension, and metabolic syndrome (MetSyn), over a 5-year follow-up period in adults aged 75 years or greater.

DESIGN

A longitudinal observational study with cross-sectional dietary assessment.

SETTING

Rural Central Pennsylvania.

PARTICIPANTS

Community-dwelling older adults (N = 449; 76.5 years old; 57% female).

MEASUREMENTS

Multiple, unannounced, 24-hour dietary recalls were used to collect dietary intake. Cluster analysis was used to derive dietary patterns. Prevalence of CVD, diabetes mellitus, hypertension, and MetSyn was extracted from outpatient electronic medical records. Logistic regression was used to examine the associations between dietary patterns and health outcomes and mortality.

RESULTS

'Sweets and Dairy', 'Health-Conscious' and 'Western' dietary patterns were identified. Compared to the 'Health-Conscious' pattern, those in the 'Sweets and Dairy' pattern had increased odds of hypertension over the follow-up period; adjusted odds ratio (95% CI) was 2.18 (1.11-4.30). No significant associations were found for CVD, diabetes mellitus, MetSyn or mortality with dietary patterns.

CONCLUSIONS

These findings support the potential value of healthy dietary patterns in the management of hypertension in older adults. We did not observe any other strong associations between dietary patterns and health outcomes or mortality in persons ≥ 75 years of age; thus failing to support the use of overly restrictive diet prescriptions for older persons, especially where food intake may be inadequate.

Brain MRI markers and dropout in a longitudinal study of cognitive aging: the Three-City Dijon Study

OBJECTIVE

Longitudinal studies of dementia rely on the assumption that individuals who drop out are comparable to those who remain in the study, adjusting for measured covariates. Existing methods to handle dropouts account for differences based on past health and cognitive measures. We assess whether such adjustments fully account for differences in future dementia risk.

METHODS

Among Three-City Study participants in Dijon, France, with 1 (n = 1,633) or 2 (n = 1,168) brain MRI scans, we tested whether white matter lesion volume (WMLV), hippocampal volume, or brain CSF volume predicted dropout ("unable to contact" or "refused interview") in repeated-measures logistic regression with up to 4 follow-ups (average 3.5 waves). Using linear regression, we also estimated differences in MRI volumes and MRI changes by subsequent dropout status and estimated plausible ranges for selective attrition bias based on these associations. Models were adjusted for demographic, health, and cognitive score covariates.

RESULTS

Baseline greater WMLV predicted increased odds of dropping out (adjusted odds ratio = 1.71; 95% confidence interval [CI] 1.20-2.43). Among participants with 2 MRI scans, individuals who subsequently dropped out had significantly worse declines in hippocampal volume (-0.30 SD difference; 95% CI -0.43 to -0.17) between the first and second MRI scans.

CONCLUSIONS

Higher future dementia risk, indicated by worse past brain MRI findings, predicted future study dropout. Adjustment for selective attrition, based on MRI markers when available, may help reduce bias in estimates of dementia incidence and improve research on dementia risk factors. MRI findings may also help prospectively identify cohort members at elevated risk of attrition.