Sex differences and predictors of anti-osteoporosis medication use in the 12 months after hip fracture surgery in adults 65 or older.

PURPOSE: This study evaluates sex differences and predictors of anti-osteoporosis medication (AOM) use following a hip fracture, with a focus on older men who exhibit higher mortality rates post-fracture compared to women over the age of 65. METHODS: Participants included 151 men and 161 women aged 65 and older with hip fractures. The outcome, AOM use, was assessed at baseline (≤ 22 days of hospitalization) and at 2, 6, and 12 months post-hip fracture. Generalized estimating equations (GEE) modeled sex differences and predictors of AOM use during the year post-fracture in 255 participants with complete baseline data and ≥ 1 follow-up observation. RESULTS: Of the 312 participants, only 53 used AOM at baseline, and 35 initiated use during follow-up. In the unadjusted GEE model, AOM use was significantly less likely in men (OR = 0.42; 95% CI, 0.22-0.78) compared to women. For both men and women, baseline use of AOM was a significant predictor (OR = 28.3; 95% CI, 5.4-148.0 vs. 41.6; 95% CI, 14.0-123.0). The other significant predictors by sex were osteoporosis diagnosis (OR = 3.19; 95% CI, 1.16-8.77) and minimal alcohol use (OR = 3.26; 95% CI, 1.34-7.94) for women versus age (OR = 1.09; 95% CI, 1.01-1.18) for men. CONCLUSION: In older adults with hip fractures, AOM use is low over the year post-fracture and men are less likely to report AOM use compared to women which has implications for important sex differences in predictors of use. Further research is needed to address overall disparities and sex differences in AOM use.

Integrating external summary information in the presence of prior probability shift: an application to assessing essential hypertension.

Recent years have witnessed a rise in the popularity of information integration without sharing of raw data. By leveraging and incorporating summary information from external sources, internal studies can achieve enhanced estimation efficiency and prediction accuracy. However, a noteworthy challenge in utilizing summary-level information is accommodating the inherent heterogeneity across diverse data sources. In this study, we delve into the issue of prior probability shift between two cohorts, wherein the difference of two data distributions depends on the outcome. We introduce a novel semi-parametric constrained optimization-based approach to integrate information within this framework, which has not been extensively explored in existing literature. Our proposed method tackles the prior probability shift by introducing the outcome-dependent selection function and effectively addresses the estimation uncertainty associated with summary information from the external source. Our approach facilitates valid inference even in the absence of a known variance-covariance estimate from the external source. Through extensive simulation studies, we observe the superiority of our method over existing ones, showcasing minimal estimation bias and reduced variance for both binary and continuous outcomes. We further demonstrate the utility of our method through its application in investigating risk factors related to essential hypertension, where the reduced estimation variability is observed after integrating summary information from an external data.

Analyzing risk factors for post-acute recovery in older adults with Alzheimer's disease and related dementia: A new semi-parametric model for large-scale medicare claims.

Nearly 300,000 older adults experience a hip fracture every year, the majority of which occur following a fall. Unfortunately, recovery after fall-related trauma such as hip fracture is poor, where older adults diagnosed with Alzheimer's disease and related dementia (ADRD) spend a particularly long time in hospitals or rehabilitation facilities during the post-operative recuperation period. Because older adults value functional recovery and spending time at home versus facilities as key outcomes after hospitalization, identifying factors that influence days spent at home after hospitalization is imperative. While several individual-level factors have been identified, the characteristics of the treating hospital have recently been identified as contributors. However, few methodological rigorous approaches are available to help overcome potential sources of bias such as hospital-level unmeasured confounders, informative hospital size, and loss to follow-up due to death. This article develops a useful tool equipped with unsupervised learning to simultaneously handle statistical complexities that are often encountered in health services research, especially when using large administrative claims databases. The proposed estimator has a closed form, thus only requiring light computation load in a large-scale study. We further develop its asymptotic properties with stabilized inference assisted by unsupervised clustering. Extensive simulation studies demonstrate superiority of the proposed estimator compared to existing estimators.

The inverse association between cancer history and incident cognitive impairment: Addressing attrition bias.

INTRODUCTION: Cancer is inversely associated with cognitive impairment. Whether this is due to statistical handling of attrition (death and censoring) is unknown. METHODS: We quantified associations between cancer history and incident cognitive impairment among Health, Aging, and Body Composition Study participants without baseline cognitive impairment or stroke (n = 2604) using multiple competing-risks models and their corresponding estimands: cause-specific, subdistribution, and marginal hazards, plus composite-outcome (cognitive impairment or all-cause mortality) hazards. All-cause mortality was also modeled. RESULTS: After covariate adjustment (demographics, apolipoprotein E ε4, lifestyle, health conditions), cause-specific and marginal hazard ratios (HRs) were similar to each other (≈ 0.84; P values < 0.05). The subdistribution HR was 0.764 (95% confidence interval [CI] = 0.645-0.906), and composite-outcome Cox model HR was 1.149 (95% CI = 1.016-1.299). Cancer history was positively associated with all-cause mortality (HR = 1.813; 95% CI = 1.525-2.156). DISCUSSION: Cause-specific, subdistribution, and marginal hazards models produced inverse associations between cancer and cognitive impairment. Competing risk models answer slightly different questions, and estimand choice influenced findings here. HIGHLIGHTS: Cancer history is inversely associated with incident cognitive impairment. Findings were robust to handling of competing risks of death. All models also addressed possible informative censoring bias. Cancer history was associated with 16% lower hazard of cognitive impairment. Cancer history was associated with 81% higher all-cause mortality hazard.

Bacterial Vaginosis and Spontaneous Clearance of Chlamydia trachomatis in the Longitudinal Study of Vaginal Flora.

BACKGROUND: Up to 26% of urogenital Chlamydia trachomatis infections spontaneously resolve between detection and treatment. Mechanisms governing natural resolution are unknown. We examined whether bacterial vaginosis (BV) was associated with greater chlamydia persistence versus spontaneous clearance in a large, longitudinal study. METHODS: Between 1999 and 2003, the Longitudinal Study of Vaginal Flora followed reproductive-age women quarterly for 1 year. Baseline chlamydia screening and treatment were initiated after ligase chain reaction testing became available midstudy, and unscreened endocervical samples were tested after study completion. Chlamydia clearance and persistence were defined between consecutive visits without chlamydia-active antibiotics (n = 320 persistence/n = 310 clearance). Associations between Nugent score (0-3, no BV; 4-10, intermediate/BV), Amsel-BV, and chlamydia persistence versus clearance were modeled with alternating and conditional logistic regression. RESULTS: Of chlamydia cases, 48% spontaneously cleared by the next visit (310/630). Nugent-intermediate/BV was associated with higher odds of chlamydia persistence (adjusted odds ratio [aOR] = 1.89; 95% confidence interval [CI], 1.30-2.74), and the findings were similar for Amsel-BV (aOR 1.39; 95% CI, .99-1.96). The association between Nugent-intermediate/BV and chlamydia persistence was stronger in a within-participant analysis of 67 participants with both clearance/persistence intervals (aOR = 4.77; 95% CI, 1.39-16.35). BV symptoms did not affect any results. CONCLUSIONS: BV is associated with greater chlamydia persistence. Optimizing the vaginal microbiome may promote chlamydia clearance.

Douching cessation and molecular bacterial vaginosis: a reanalysis of archived specimens.

OBJECTIVES: Observational studies demonstrate an association between vaginal douching and bacterial vaginosis (BV) characterised by Gram stain. We sought to describe the effect of a douching cessation intervention on the composition and structure of the vaginal microbiota and molecular-BV, a state defined by low levels of Lactobacillus spp evaluated by molecular tools. METHODS: 33 women self-collected mid-vaginal swabs twice weekly (982 samples) during a douching observation phase (4 weeks) followed by a douching cessation phase (12 weeks) in a 2005 single crossover pilot study conducted in Baltimore, Maryland. Vaginal microbiota were characterised by 16S rRNA gene amplicon sequencing (V3-V4) and clustered into community state types (CSTs). Conditional logistic regression modelling allowed each participant to serve as their own control. Wilcoxon signed-rank tests were used to evaluate changes in microbiota between phases. Broad-range qPCR assays provided estimates of bacterial absolute abundance per swab in a subsample of seven participants before and after douching. A piecewise linear mixed effects model was used to assess rates of change in bacterial absolute abundance before and after douching. RESULTS: There was no statistically significant change in the odds of molecular-BV versus Lactobacillus-dominated CSTs comparing the douching cessation interval to douching observation (adjusted OR 1.77, 95% CI 0.89 to 3.55). Removal of L. iners-dominated CST III from the outcome did not affect the results. There were no significant changes in the relative abundance of four Lactobacillus spp and no meaningful changes in other taxa investigated. There was no significant change in bacterial absolute abundance between a participant's sample collected 3 days prior to and following douching (p=0.46). CONCLUSIONS: In this pilot study, douching cessation was not associated with major changes in vaginal microbiota. Douching cessation alone may not durably shift the vaginal microbiota and additional interventions may be needed to restore optimal vaginal microbiota among those who douche.

Factors Associated With Incidence and Spontaneous Clearance of Molecular-Bacterial Vaginosis: Results From a Longitudinal Frequent-Sampling Observational Study.

BACKGROUND: We sought to assess time-independent and time-varying factors associated with incidence and spontaneous clearance of molecular-bacterial vaginosis (BV; without treatment). METHODS: Midvaginal samples were self-collected daily by 100 participants recruited at the University of Alabama Birmingham for 10 weeks (4778 samples). Vaginal microbiota was characterized by 16S rRNA gene amplicon sequencing and clustered into community state types (CSTs). A low-Lactobacillus CST defined the molecular-BV outcome in this study. Factors associated with molecular-BV incidence and spontaneous clearance were modeled using Andersen-Gill recurrent event Cox models. Community class identified the predominant CST of a participant during follow-up. RESULTS: Menstruations (adjusted odds ratio [aHR], 2.09 [95% confidence interval, 1.51-2.89] in the prior 24 hours) and CST III (Lactobacillus iners dominated) at the previous sample (aHR, 2.25 [1.48-3.40]) were associated with increased molecular-BV incidence. Participants with a majority of L. iners-dominated samples longitudinally (community class LI) displayed less stable patterns of vaginal microbiota. In LI participants, reduced molecular-BV spontaneous clearance was observed in African American participants (aHR, 0.44 [0.26-0.75]) compared with White participants, older participants (age, 40-49 years [aHR, 0.38; 0.23-0.61]; age, 30-39 years [aHR, 0.48; 0.28-0.83]) compared with participants aged 18 to 29 years, and after douching (0.45 [0.28-0.73] within prior 72 hours). CONCLUSIONS: Although it is now well documented that vaginal microbiota are dynamic, there are few available data on factors associated with spontaneous clearance of molecular-BV. Lactobacillus iners-dominated vaginal microbiota are more likely to be dynamic and associated with different risk factors for incidence and clearance of BV. Among L. iners-dominated participants, age, race, and douching were linked to reduced clearance. Most transitions to molecular-BV during menstruations were short-lived.

Biogenic Amines Increase the Odds of Bacterial Vaginosis and Affect the Growth of and Lactic Acid Production by Vaginal Lactobacillus spp.

Bacterial vaginosis (BV) is the most common vaginal disorder of reproductive-aged women, yet its etiology remains enigmatic. One clinical symptom of BV, malodor, is linked to the microbial production of biogenic amines (BA). Using targeted liquid chromatography mass spectrometry, we analyzed 149 longitudinally collected vaginal samples to determine the in vivo concentrations of the most common BAs and then assessed their relationship to BV and effect upon the growth kinetics of axenically cultured vaginal Lactobacillus species. Increases in cadaverine, putrescine, and tyramine were associated with greater odds of women transitioning from L. crispatus-dominated vaginal microbiota to microbiota that have a paucity of Lactobacillus spp. and from Nugent scores of 0 to 3 to Nugent scores of 7 to 10, consistent with BV. Exposure to putrescine lengthened the lag time and/or slowed the growth of all vaginal Lactobacillus spp. except L. jensenii 62G. L. iners AB107's lag time was lengthened by cadaverine but reduced in the presence of spermidine and spermine. The growth rate of L. crispatus VPI 3199 was slowed by cadaverine and tyramine, and strain-specific responses to spermine and spermidine were observed. BAs were associated with reduced production of d- and l-lactic acid by vaginal Lactobacillus spp., and this effect was independent of their effect upon Lactobacillus species growth. The exceptions were higher levels of d- and l-lactic acid by two strains of L. crispatus when grown in the presence of spermine. Results of this study provide evidence of a direct impact of common biogenic amines on vaginal Lactobacillus spp.IMPORTANCELactobacillus spp. are credited with providing the primary defense against gynecological conditions, including BV, most notably through the acidification of the vaginal microenvironment, which results from their production of lactic acid. The microbial production of BAs has been hypothesized to play a mechanistic role in diminishing Lactobacillus species-mediated protection, enabling the colonization and outgrowth of diverse anaerobic bacterial species associated with BV. Here, we demonstrate that in vivo increases in the most commonly observed BAs are associated with a loss of Lactobacillus spp. and the development of BV, measured by Nugent score. Further, we show that BAs formed by amino acid decarboxylase enzymes negatively affect the growth of type strains of the most common vaginal Lactobacillus spp. and separately alter their production of lactic acid. These results suggest that BAs destabilize vaginal Lactobacillus spp. and play an important and direct role in diminishing their protection of the vaginal microenvironment.

Depressive symptom heterogeneity among older adults after hip fracture.

OBJECTIVE: to evaluate patterns of depressive symptoms after hip fracture and examine their impact on functional recovery. METHODS: participants (n = 304) included older adults from the Baltimore Hip Studies 7th cohort who experienced a hip fracture. Depressive symptoms were measured at baseline or 2-, 6- or 12-month post-hip fracture using the 20-item Center for Epidemiologic Studies Depression scale. Gait speed was measured after hip fracture at 2-, 6- or 12-month follow-up. Latent class analysis was used to identify individuals with similar patterns of depressive symptoms after hip fracture. Item response probabilities characterised symptom profiles, and posterior probability estimates were used to assign participants to a baseline depressive symptom subtype. Weighted estimated equations compared post-fracture gait speed between baseline symptomatic and asymptomatic groups. RESULTS: four patterns of depressive symptoms were identified: asymptomatic (50.8%), somatic (28.6%), melancholic (11.4%) and anhedonic (9.2%). The somatic subtype was characterised by difficultly concentrating and reduced energy and movement, whereas anhedonic symptoms were associated with the inability to experience pleasure. Melancholic symptoms corresponded to anhedonia, decreased physical activity and other psychological and somatic complaints. Compared with the asymptomatic group, somatic symptoms were consistently associated with slower gait speed, -0.03 metres per second (m/s) and between-group differences for melancholic symptomology were as large as -0.05 m/s, but the associations were not statistically significant. CONCLUSION: findings demonstrate unique depressive symptom subtypes in older adults after hip fracture and provide confirmatory evidence of unique clinical phenotypes; however, their impact on functional recovery after hip fracture remains unclear.

Association between disease progression and depression onset in persons with radiographic knee osteoarthritis.

OBJECTIVES: Osteoarthritis (OA) disease progression may lead to deteriorating psychosocial function, but it is unclear what aspects of disease severity are related to the onset of depression. This study assessed which components of OA disease progression cumulatively contribute to depression onset in persons with radiographic knee OA. METHODS: Osteoarthritis Initiative participants (n = 1651) with radiographic disease (Kellgren-Lawrence grade ≥2) in one or both knees and below the screening threshold for probable depression [Center for Epidemiological Studies Depression (CES-D) scale <16] at baseline were included. Disease severity was measured from baseline to the third annual follow-up visit using joint space width, 20-meter gait speed, and the Western Ontario and McMaster Universities Osteoarthritis Index pain subscale, each categorized into quintiles. Depression onset (CES-D ≥ 16) was assessed annually at four follow-up visits. Marginal structural models that account for time-dependent confounding and attrition evaluated the association between each time-varying disease severity measure and depression onset. RESULTS: Each disease severity measure exhibited a non-linear relationship concerning the probability of depression onset, with the higher quintiles generally being associated with a larger risk. The highest quintile (relative to the lowest) of joint space width and gait speed were both significantly associated with depression onset. By contrast, none of the higher pain quintiles compared with the lowest were significantly associated with the onset of depression. CONCLUSION: Faster disease progression as measured by either worsening structural severity or decreasing physical performance corresponds to an increased risk of depression among individuals with radiographic knee OA.

Extracellular vesicle biomarkers of Alzheimer's disease associated with sub-clinical cognitive decline in late middle age.

INTRODUCTION: Neuronal extracellular vesicle (nEV) tau and insulin signaling biomarkers may detect preclinical Alzheimer's disease and age-associated cognitive decline. METHODS: This case-control study used repeated serum samples from 73 cognitively declining and 73 stable Wisconsin Registry for Alzheimer's Prevention participants (62.4 ± 6.3 years old). We immunocaptured nEVs; measured tau and insulin signaling biomarkers; and examined biomarker differences by group, their performance in group classification in training and test datasets (97, 49 individuals, respectively), and whether they predict cognitive performance change. RESULTS: Declining compared to stable individuals showed higher baseline total, p231-, and p181-tau with older age and higher annualized change for p-IR and p-IGF-1R. Combining biomarkers classified decliners with 94% area under the curve (AUC), 86.0% sensitivity and 86.7% specificity, in training data, and 75% AUC, 71.4% sensitivity, and 77.3% specificity, in test data. Insulin biomarkers predicted cognitive performance change prospectively. DISCUSSION: Combining nEV biomarkers can identify individuals with age-associated cognitive decline.

Diffusion-weighted MRI with intravoxel incoherent motion modeling for assessment of muscle perfusion in the thigh during post-exercise hyperemia in younger and older adults.

Aging is associated with impaired endothelium-dependent vasodilation that leads to muscle perfusion impairment and contributes to organ dysfunction. Impaired muscle perfusion may result in inadequate delivery of oxygen and nutrients during and after muscle contraction, leading to muscle damage. The ability to study the relationship between perfusion and muscle damage has been limited using traditional muscle perfusion measures, which are invasive and risky. To overcome this limitation, we optimized a diffusion-weighted MRI sequence and validated an intravoxel incoherent motion (IVIM) analysis based on Monte Carlo simulation to study muscle perfusion impairment with aging during post-exercise hyperemia. Simulation results demonstrated that the bias of IVIM-derived perfusion fraction (fp ) and diffusion of water molecules in extra-vascular tissue (D) ranged from -3.3% to 14% and from -16.5% to 0.002%, respectively, in the optimized experimental condition. The dispersion in fp and D ranged from 3.2% to 9.5% and from 0.9% to 1.1%, respectively. The mid-thigh of the left leg of four younger (21-30 year old) and four older (60-90 year old) healthy females was studied using the optimized protocol at baseline and at seven time increments occurring every 3.25 min following in-magnet dynamic knee extension exercise performed using a MR-compatible ergometer with a workload of 0.4 bar for 2.5 min. After exercise, both fp and D significantly increased in the rectus femoris (active muscle during exercise) but not in adductor magnus (inactive muscle), reflecting the fact that the local increase in perfusion with both groups showed a maximum value in the second post-exercise time-point. A significantly greater increase in perfusion from the baseline (p < 0.05) was observed in the younger group (37 ± 12.05%) compared with the older group (17.57 ± 15.92%) at the first post-exercise measurement. This work establishes a reliable non-invasive method that can be used to study the effects of aging on dynamic changes in muscle perfusion as they relate to important measures of physical function.

Trunk Muscle Composition 2 Months After Hip Fracture: Findings From the Baltimore Hip Studies.

OBJECTIVE: To determine if hip fracture patients would have smaller cross-sectional area (CSA) and lower radiological attenuation (suggesting greater fat infiltration) in all trunk muscles as compared to older adults without hip fractures. DESIGN: Cross-sectional analysis of computed tomography (CT) scans. SETTING: Clinical imaging facility. PARTICIPANTS: Forty-one white participants (19 men, 22 women) from the Baltimore Hip Studies seventh cohort at 2 months postfracture were compared to 693 white participants (424 men, 269 women) from the Health, Aging and Body Composition (Health ABC) study at the year 6 visit (N=734). INTERVENTION: Not applicable. MAIN OUTCOME MEASURES: Trunk muscle CSA and attenuation values were obtained from a single 10-mm, axial CT scan completed at the L4-L5 disc space in each participant. RESULTS: The hip fracture cohort had significantly smaller CSA for all trunk muscles (range: 12.1%-38% smaller) compared to the Health ABC cohort (P<.01), with the exception of the rectus abdominus muscle in men (P=.12). But, hip fracture patients, particularly female patients, had higher attenuation levels (lower intramuscular fat) in all trunk muscles (P<.0001). CONCLUSIONS: Findings are consistent with atrophy of the trunk muscles in the hip fracture population without a high level of intramuscular fat. Future work should evaluate the role of trunk muscle composition in the functional recovery of older adults after hip fracture.

Effect of a Multicomponent Home-Based Physical Therapy Intervention on Ambulation After Hip Fracture in Older Adults: The CAP Randomized Clinical Trial.

Importance: Disability persists after hip fracture in older persons. Current rehabilitation may not be sufficient to restore ability to walk in the community. Objective: To compare a multicomponent home-based physical therapy intervention (training) with an active control on ability to walk in the community. Design, Setting, and Participants: Parallel, 2-group randomized clinical trial conducted at 3 US clinical centers (Arcadia University, University of Connecticut Health Center, and University of Maryland, Baltimore). Randomization began on September 16, 2013, and ended on June 20, 2017; follow-up ended on October 17, 2017. Patients aged 60 years and older were enrolled after nonpathologic, minimal trauma hip fracture, if they were living in the community and walking without human assistance before the fracture, were assessed within 26 weeks of hospitalization, and were not able to walk during daily activities at the time of enrollment. A total of 210 participants were randomized and reassessed 16 and 40 weeks later. Interventions: The training intervention (active treatment) (n = 105) included aerobic, strength, balance, and functional training. The active control group (n = 105) received transcutaneous electrical nerve stimulation and active range-of-motion exercises. Both groups received 2 to 3 home visits from a physical therapist weekly for 16 weeks; nutritional counseling; and daily vitamin D (2000 IU), calcium (600 mg), and multivitamins. Main Outcomes and Measures: The primary outcome (community ambulation) was defined as walking 300 m or more in 6 minutes at 16 weeks after randomization. The study was designed to test a 1-sided hypothesis of superiority of training compared with active control. Results: Among 210 randomized participants (mean age, 80.8 years; 161 women [76.7%]), 197 (93.8%) completed the trial (187 [89.0%] by completing the 6-minute walk test at 16 weeks and 10 [4.8%] by adjudication of the primary outcome). Among these, 22 of 96 training participants (22.9%) and 18 of 101 active control participants (17.8%) (difference, 5.1% [1-sided 97.5% CI, -∞ to 16.3%]; 1-sided P = .19) became community ambulators. Seventeen training participants (16.2%) and 15 control participants (14.3%) had 1 or more reportable adverse events during the intervention period. The most common reportable adverse events reported were falls (training: 6 [5.7%], control: 4 [3.8%]), femur/hip fracture (2 in each group), pneumonia (training: 2, control: 0), urinary tract infection (training: 2, control: 0), dehydration (training: 0, control: 2), and dyspnea (training: 0, control: 2). Conclusions and Relevance: Among older adults with a hip fracture, a multicomponent home-based physical therapy intervention compared with an active control that included transcutaneous electrical nerve stimulation and active range-of-motion exercises did not result in a statistically significant improvement in the ability to walk 300 m or more in 6 minutes after 16 weeks. Trial Registration: ClinicalTrials.gov Identifier: NCT01783704.

Regular low-calorie sweetener consumption is associated with increased secretion of glucose-dependent insulinotropic polypeptide.

Low-calorie sweeteners (LCSs) are widely used for weight control despite limited evidence of their effectiveness and studies linking LCS consumption with incident obesity. We tested the hypothesis that regular LCS consumption is associated with higher postprandial glucose-dependent insulinotropic polypeptide (GIP) secretion, which has been linked to obesity. We used data from participants in the Baltimore Longitudinal Study of Aging who had completed a diet diary, had at least one visit during which they underwent an oral glucose tolerance test (OGTT), and had no diabetes. Of 232 participants, 166 contributed 1, 39 contributed 2, and 27 contributed 3 visits, and 96 (41%) reported using LCS. Plasma OGTT samples were analysed for glucose, insulin and GIP. Fasting glucose, insulin and GIP levels were no different between LCS users and non-users. The association of LCS use with 2-hour OGTT responses after adjustment for covariates was non-significant for glucose (P = .98) and insulin (P = .18), but significant for greater increase in GIP in LCS users (P = .037). Regular consumption of LCSs was associated with greater increases in GIP secretion after food intake, which may potentially lead to weight gain through the lipogenic properties of GIP.

Joint mixed-effects models for causal inference with longitudinal data.

Causal inference with observational longitudinal data and time-varying exposures is complicated due to the potential for time-dependent confounding and unmeasured confounding. Most causal inference methods that handle time-dependent confounding rely on either the assumption of no unmeasured confounders or the availability of an unconfounded variable that is associated with the exposure (eg, an instrumental variable). Furthermore, when data are incomplete, validity of many methods often depends on the assumption of missing at random. We propose an approach that combines a parametric joint mixed-effects model for the study outcome and the exposure with g-computation to identify and estimate causal effects in the presence of time-dependent confounding and unmeasured confounding. G-computation can estimate participant-specific or population-average causal effects using parameters of the joint model. The joint model is a type of shared parameter model where the outcome and exposure-selection models share common random effect(s). We also extend the joint model to handle missing data and truncation by death when missingness is possibly not at random. We evaluate the performance of the proposed method using simulation studies and compare the method to both linear mixed- and fixed-effects models combined with g-computation as well as to targeted maximum likelihood estimation. We apply the method to an epidemiologic study of vitamin D and depressive symptoms in older adults and include code using SAS PROC NLMIXED software to enhance the accessibility of the method to applied researchers.

Persistence of depressive symptoms and gait speed recovery in older adults after hip fracture.

OBJECTIVE: Depression after hip fracture in older adults is associated with worse physical performance; however, depressive symptoms are dynamic, fluctuating during the recovery period. The study aim was to determine how the persistence of depressive symptoms over time cumulatively affects the recovery of physical performance. METHODS: Marginal structural models estimated the cumulative effect of persistence of depressive symptoms on gait speed during hip fracture recovery among older adults (n = 284) enrolled in the Baltimore Hip Studies 7th cohort. Depressive symptoms at baseline and at 2-month and 6-month postadmission for hip fracture were evaluated by using the Center for Epidemiological Studies Depression Scale, and persistence of symptoms was assessed as a time-averaged severity lagged to standardized 3 m gait speed at 2, 6, and 12 months. RESULTS: A 1-unit increase in time-averaged Center for Epidemiological Studies Depression score was associated with a mean difference in gait speed of -0.0076 standard deviations (95% confidence interval [CI]: -0.0184, 0.0032; P = .166). The association was largest in magnitude from baseline to 6 months: -0.0144 standard deviations (95% CI: -0.0303, 0.0015; P = 0.076). Associations for the other time intervals were smaller: -0.0028 standard deviations (95% CI: -0.0138, 0.0083; P = .621) at 2 months and -0.0121 standard deviations (95% CI: -0.0324, 0.0082; P = .238) at 12 months. CONCLUSION: Although not statistically significant, the magnitude of the numerical estimates suggests that expressing more depressive symptoms during the first 6 months after hip fracture has a meaningful impact on functional recovery.

A novel, multiplexed targeted mass spectrometry assay for quantification of complement factor H (CFH) variants and CFH-related proteins 1-5 in human plasma.

Age-related macular degeneration (AMD) is a leading cause of visual loss among older adults. Two variants in the complement factor H (CFH) gene, Y402H and I62V, are strongly associated with risk of AMD. CFH is encoded in regulator of complement activation gene cluster in chromosome 1q32, which includes complement factor related (CFHR) proteins, CFHR1 to CFHR5, with high amino acid sequence homology to CFH. Our goal was to build a SRM assay to measure plasma concentrations of CFH variants Y402, H402, I62, and V62, and CFHR1-5. The final assay consisted of 24 peptides and 72 interference-free SRM transition ion pairs. Most peptides showed good linearity over 0.3-200 fmol/µL concentration range. Plasma concentrations of CFH variants and CFHR1-5 were measured using the SRM assay in 344 adults. Plasma CFH concentrations (mean, SE in µg/mL) by inferred genotype were: YY402, II62 (170.1, 31.4), YY402, VV62 (188.8, 38.5), HH402, VV62 (144.0, 37.0), HY402, VV62 (164.2, 42.3), YY402, IV62 (194.8, 36.8), HY402, IV62 (181.3, 44.7). Mean (SE) plasma concentrations of CFHR1-5 were 1.63 (0.04), 3.64 (1.20), 0.020 (0.001), 2.42 (0.18), and 5.49 (1.55) µg/mL, respectively. This SRM assay should facilitate the study of the role of systemic complement and risk of AMD.

Spontaneous Viral Load Decline and Subsequent Clearance of Chronic Hepatitis C Virus in Postpartum Women Correlates With Favorable Interleukin-28B Gene Allele.

BACKGROUND: Postpartum hepatitis C viral (HCV) load decline followed by spontaneous clearance has been previously described. Herein we identify predictors for viral decline in a cohort of HCV-infected postpartum women. METHODS: Pregnant women at Cairo University were screened for anti-HCV antibodies and HCV RNA, and viremic women were tested for quantitative HCV RNA at 3, 6, 9, and 12 months postpartum. Spontaneous clearance was defined as undetectable viremia twice at least 6-months apart. Associations between viral load and demographic, obstetrical, HCV risk factors, and interleukin-28B gene (IL28B) polymorphism (rs12979860) were assessed. RESULTS: Of 2514 women, 97 (3.9%) had anti-HCV antibodies, 54 (2.1%) were viremic and of those, 52 (2.1%) agreed to IL28B testing. From pregnancy until 12 months postpartum, IL28B-CC allele women had a significant viral decline (P = .009). After adjusting, the IL28B-CC allele had a near significant difference compared to the CT allele (odds ratio [OR], 0.75; 95% confidence interval [CI], 0.75,1.00; P = .05), but not the TT allele (OR, 0.91; 95% CI, 0.61,1.38; P = .64). All 14/52 (26.9%) women who subsequently cleared were among the 15 with undetectable viremia at 12 months, making that time point a strong predictor of subsequent clearance (sensitivity = 100%, specificity = 97.4%, positive predictive value = 93.3%, negative predictive value = 100%). CONCLUSIONS: IL28B-CC genotype and 12-month postpartum undetectable viremia were the best predictors for viral decline and subsequent clearance. These 2 predictors should influence clinical decision making.

Doubly robust estimation and causal inference in longitudinal studies with dropout and truncation by death.

Motivated by aging research, we propose an estimator of the effect of a time-varying exposure on an outcome in longitudinal studies with dropout and truncation by death. We use an inverse-probability weighted (IPW) estimator to derive a doubly robust augmented inverse-probability weighted (AIPW) estimator. IPW estimation involves weights for the exposure mechanism, dropout, and mortality; AIPW estimation additionally involves estimating data-generating models via regression. We demonstrate that the estimators identify a causal contrast that is a function of principal strata effects under a set of assumptions. Simulations show that AIPW estimation is unbiased when weights or outcome regressions are correct, and that AIPW estimation is more efficient than IPW estimation when all models are correct. We apply the method to a study of vitamin D and gait speed among older adults.