Application of multiple validated algorithms for identifying incident breast cancer among individuals with atopic dermatitis.

PURPOSE: Validated algorithms (VAs) in insurance claims databases are often used to estimate the prevalence and incidence of comorbidities and evaluate safety signals. However, although they are then used in different data sources or subpopulations from those in which they were developed the replicability of these VAs are rarely tested, making their application and performance in these settings potentially unknown. This paper describes testing multiple VAs used to identify incident breast cancer cases in a general population and in an indication-specific population, patients with atopic dermatitis (AD). METHODS: Two algorithms were tested in multiple insurance claims databases and four cohorts were created. Modifications were made to account for the US insurance setting. The resulting incidence rates (IRs) were then compared across algorithms and against surveillance, epidemiology, and end results (SEER) estimates to assess reliability. RESULTS: Algorithm 1 produced low IRs compared to Algorithm 2. Algorithm 2 provided similar estimates to those of SEER. Individuals in the AD cohorts experienced lower incident breast cancer cases than those in the general population cohorts. CONCLUSION: Regardless of an algorithm's reported accuracy, the original study setting and targeted population for the VAs may matter when attempting to replicate the algorithm in an indication-specific subpopulation or varying data sources. Investigators should use caution and conduct sensitivity analyses or use multiple algorithms when attempting to calculate incidence or prevalence estimates using VAs.

Cohabitation as a determinant of adaptive and innate immune cell profiles: Findings from the Health and Retirement Study.

Introduction: Non-genetic factors are important but poorly understood determinants of immune profiles. Age and Cytomegalovirus (CMV) infection remain two well documented non-genetic determinants of the immune profile. Recently, one study identified cohabitation in the same household as an important determinant of immune profiles. Methods: We used immunophenotyping data from the Health and Retirement Study (HRS) to evaluate the association between cohabitation and the adaptive (subsets of T-cells, B-cells) and innate immune profiles (subsets of monocytes, natural killer cells and neutrophils). We compared adaptive and innate immune cell profiles using immunophenotyping data from 1184 same-household pairs (cohabitating partners) to 1184 non-household pairs to evaluate the association between cohabitation and adaptive immune cell profiles. We used data from 1737 same-household pairs and 1737 non-household pairs to evaluate the association between cohabitation and innate cell profiles. Household and non-household pairs were matched on age (±2years), educational background and race/ethnicity to minimize confounding due to these factors. The adaptive immune cells and innate immune cell profiles were compressed to two coordinates using multidimensional scaling (MDS). The Euclidean distances between same-household pairs were compared to the distances between non-household pairs for the adaptive and innate cell profiles separately using two sample independent t-tests. We also performed additional adjustment for age and BMI differences, CMV serostatus and smoking concordance/discordance status among household members. Results: For adaptive immune cell profiles, the mean Euclidean distance between same-household pairs was 4% lower than the non-household pairs (p = 0.03). When stratified by concordance for CMV serostatus among household pairs, the Euclidean distance was significantly lower by 8% in the same-household pairs as compared to non-household pairs among those who were discordant for CMV serostatus (p = 0.01) and among same-household pairs who were CMV seronegative (p = 0.02) after covariate adjustment. The mean Euclidian distance between same-household pairs was also 8% lower than non-household pairs for the innate immune cell profiles (p-value <0.0001) and this difference remained consistent across all strata of CMV infection. Discussion: This study confirms that cohabitation is associated with similarity in immune cell profiles. The differential effects of cohabitation on the adaptive and innate immune profiles suggest that further studies into the common environmental factors that influence individual immune cell subsets need to be evaluated in greater detail.

Age-related Differences in T-cell Subsets and Markers of Subclinical Inflammation in Aging Are Independently Associated With Type 2 Diabetes in the Health and Retirement Study.

OBJECTIVES: Age-related changes in adaptive immunity and subclinical inflammation are both important risk factors for diabetes in older adults. We evaluated the independent association between T-cell subsets, subclinical inflammation, and diabetes risk in the Health and Retirement Study (HRS). METHODS: We measured 11 T-cell subsets, 5 pro-inflammatory markers, and 2 anti-inflammatory markers from the 2016 wave of the HRS (baseline). Diabetes/prediabetes status was estimated at the 2016, 2018, and 2020 waves of HRS, based on levels of blood glucose/glycated hemoglobin in plasma or self-reported status. We used survey generalized logit models to evaluate the cross-sectional associations and Cox proportional hazard models to evaluate longitudinal associations. RESULTS: Among 8,540 participants (56 to 107 years of age), 27.6% had prevalent type 2 diabetes and 31.1% had prediabetes in the 2016 survey. After adjusting for age, sex, race/ethnicity, education, obesity, smoking, comorbidity index, and cytomegalovirus seropositivity, individuals with type 2 diabetes had lower naive T cells and higher memory and terminal effector T cells as compared with normoglycemic individuals. Among 3,230 normoglycemic participants in the 2016 survey, the incidence of diabetes was 1.8% over 4 years of follow-up. The baseline percentage of CD4+ effector memory T cells was associated with a lower risk of incident diabetes (hazard ratio [HR]=0.63, 95% confidence interval [CI] 0.49 to 0.80, p=0.0003) after adjustment for covariates. Baseline level of interleukin-6 (IL-6) was associated with risk of incident diabetes (HR=1.52, 95% CI 1.18 to 1.97, p=0.002). The associations between age-related changes in CD4+ effector memory T cells and risk of incident diabetes remained unchanged after adjustment for subclinical inflammation, although adjusting for CD4+ effector memory T cells nullified the association between IL-6 and incident diabetes. CONCLUSIONS: This study showed that the baseline percentage of CD4+ effector memory T cells was inversely associated with incident diabetes independent of subclinical inflammation, but CD4+ effector memory T-cell subsets affected the relationship between IL-6 and incident diabetes. Further studies are needed to confirm and investigate mechanisms by which T-cell immunity affects diabetes risk.

Socioeconomic status and immune aging in older US adults in the health and retirement study.

Socioeconomic and demographic factors including educational attainment, race and ethnicity, and childhood socioeconomic status (SES) are powerful predictors of inequalities in aging, morbidity, and mortality. Immune aging, including accumulation of late-differentiated, senescent-like lymphocytes and lower levels of naïve lymphocytes, may play a role in the development of the age-related health inequalities. This study used nationally representative data from more than 9,000 US adults from the Health and Retirement Study to investigate associations between educational attainment, race and ethnicity, and childhood SES and lymphocyte percentages. Respondents with lower educational attainment, Hispanic adults, and those who had a parent with less than a high school education had lymphocyte percentages consistent with more immune aging compared to those with greater educational attainment, non-Hispanic White adults, and respondents who had parents with a high school education, respectively. Associations between education, Hispanic ethnicity, and parents' education and late differentiated senescent-like T lymphocytes (TemRA) and B cells were largely driven by cytomegalovirus (CMV), suggesting it is a factor in observed SES inequalities in immunosenescence. Naïve T lymphocytes may be particularly affected by socioeconomic position and may therefore be of particular interest to research interested in inequalities in health and aging.

Evaluation of T-cell aging-related immune phenotypes in the context of biological aging and multimorbidity in the Health and Retirement Study.

BACKGROUND: Cellular changes in adaptive immune system accompany the process of aging and contribute to an aging-related immune phenotype (ARIP) characterized by decrease in naïve T-cells (TN) and increase in memory T-cells (TM). A population-representative marker of ARIP and its associations with biological aging and age-related chronic conditions have not been studied previously. METHODS: We developed two ARIP indicators based on well understood age-related changes in T cell distribution: TN/(TCM (Central Memory) + TEM (Effector Memory) + TEFF (Effector)) (referred as TN/TM) in CD4 + and CD8 + T-cells. We compared them with existing ARIP measures including CD4/CD8 ratio and CD8 + TN cells by evaluating associations with chronological age and the Klemera Doubal measure of biological age (measured in years) using linear regression, multimorbidity using multinomial logistic regression and two-year mortality using logistic regression. RESULTS: CD8 + TN and CD8 + TN/TM had the strongest inverse association with chronological age (beta estimates: -3.41 and -3.61 respectively; p-value < 0.0001) after adjustment for sex, race/ethnicity and CMV status. CD4 + TN/TM and CD4 + TN had the strongest inverse association with biological age (ß = -0.23; p = 0.003 and ß = -0.24; p = 0.004 respectively) after adjustment for age, sex, race/ethnicity and CMV serostatus. CD4/CD8 ratio was not associated with chronological age or biological age. CD4 + TN/TM and CD4 + TN was inversely associated with multimorbidity. For CD4 + TN/TM, people with 2 chronic conditions had an odds ratio of for 0.74 (95%CI: 0.63-0.86 p = 0.0003) compared to those without any chronic conditions while those with 3 chronic conditions had an odds ratio of 0.75 (95% CI: 0.63-0.90; p = 0.003) after adjustment for age, sex, race/ethnicity, CMV serostatus, smoking, and BMI. The results for the CD4 + TN subset were very similar to the associations seen with the CD4 + TN/TM. CD4 + TN/TM and CD4 + TN were both associated with two-year mortality (OR = 0.80 (95% CI: 0.67-0.95; p = 0.01) and 0.81 (0.70-0.94; p = 0.01), respectively). CONCLUSION: CD4 + TN/TM and CD4 + TN had a stronger association with biological age, age-related morbidity and mortality compared to other ARIP measures. Future longitudinal studies are needed to evaluate the utility of the CD4 + subsets in predicting the risk of aging-related outcomes.

Gene expression of oxidative stress markers and lung function: A CARDIA lung study.

BACKGROUND: Circulating markers of oxidative stress have been associated with lower lung function. Our objective was to study the association of gene expression levels of oxidative stress pathway genes (ALOX12, ALOX15, ARG2, GSTT1, LPO, MPO, NDUFB3, PLA2G7, and SOD3) and lung function forced expiratory volume in one second (FEV1 ), forced vital capacity (FVC) in Coronary Artery Risk Development in Young Adults study. METHODS: Lung function was measured using spirometry and the Nanostring platform was used to estimate gene expression levels. Linear regression models were used to study association of lung function measured at year 30, 10-year decline in lung function and gene expression after adjustment for center, smoking, and BMI, measured at year 25. RESULTS: The 10-year decline of FEV1 was faster in highest NDUFB3 quartile compared to the lowest (difference = -2.09%; p = 0.001) after adjustment for multiple comparisons. The 10-year decline in FEV1 and FVC was nominally slower in highest versus lowest quartile of PLA2G7 (difference = 1.14%; p = 0.02, and difference = 1.06%; p = 0.005, respectively). The other genes in the study were not associated with FEV1 or FVC. CONCLUSION: Higher gene expression levels in oxidative stress pathway genes are associated with faster 10-year FEV1 decline.

Lung Function and Gene Expression of Pathogen Recognition Pathway Receptors: the Cardia Lung Study.

Activation of toll-like receptors (TLR1, TLR5, TLR6) and downstream markers (CCR1, MAPK14, ICAM1) leads to increased systemic inflammation. Our objective was to study the association between the gene expression levels of these six genes and lung function (Forced Expiratory Volume in one second (FEV1), Forced Vital Capacity (FVC) and FEV1/FVC). We studied gene expression levels and lung function in the Coronary Artery Risk Development in Young Adults study. Spirometry testing was used to measure lung function and gene expression levels were measured using the Nanostring platform. Multivariate linear regression models were used to study the association between lung function measured at year 30, 10-year decline from year 20 to year 30, and gene expression levels (highest quartile divided into two levels - 75th to 95th and>95th to 100th percentile) adjusting for center, smoking and BMI, measured at year 25. Year 30 FEV1 and FVC were lower in the highest level of TLR5 compared to the lowest quartile with difference of 4.00% (p for trend: 0.04) and 3.90% (p for trend: 0.05), respectively. The 10-year decline of FEV1 was faster in the highest level of CCR1 as compared to the lowest quartile with a difference of 1.69% (p for trend: 0.01). There was no association between gene expression and FEV1/FVC. Higher gene expression levels in TLR5 and CCR1 are associated with lower lung function and faster decline in FEV1 over 10 years, in a threshold manner, providing new insights into the role of inflammation in lung function.

Validation of a hybrid approach to standardize immunophenotyping analysis in large population studies: The Health and Retirement Study.

Traditional manual gating strategies are often time-intensive, place a high burden on the analyzer, and are susceptible to bias between analyzers. Several automated gating methods have shown to exceed performance of manual gating for a limited number of cell subsets. However, many of the automated algorithms still require significant manual interventions or have yet to demonstrate their utility in large datasets. Therefore, we developed an approach that utilizes a previously published automated algorithm (OpenCyto framework) with a manually created hierarchically cell gating template implemented, along with a custom developed visualization software (FlowAnnotator) to rapidly and efficiently analyze immunophenotyping data in large population studies. This approach allows pre-defining populations that can be analyzed solely by automated analysis and incorporating manual refinement for smaller downstream populations. We validated this method with traditional manual gating strategies for 24 subsets of T cells, B cells, NK cells, monocytes and dendritic cells in 931 participants from the Health and Retirement Study (HRS). Our results show a high degree of correlation (r ≥ 0.80) for 18 (78%) of the 24 cell subsets. For the remaining subsets, the correlation was low (<0.80) primarily because of the low numbers of events recorded in these subsets. The mean difference in the absolute counts between the hybrid method and manual gating strategy of these cell subsets showed results that were very similar to the traditional manual gating method. We describe a practical method for standardization of immunophenotyping methods in large scale population studies that provides a rapid, accurate and reproducible alternative to labor intensive manual gating strategies.