Multivariable Ordinal Logistic Regression Models Research Articles

Two-week continuous glucose monitoring-derived metrics and degree of hepatic steatosis: a cross-sectional study among Chinese middle-aged and elderly participants

BackgroundContinuous glucose monitoring (CGM) devices provide detailed information on daily glucose control and glycemic variability. Yet limited population-based studies have explored the association between CGM metrics and fatty liver. We aimed to investigate the associations of CGM metrics with the degree of hepatic steatosis.MethodsThis cross-sectional study included 1180 participants from the Guangzhou Nutrition and Health Study. CGM metrics, covering mean glucose level, glycemic variability, and in-range measures, were separately processed for all-day, nighttime, and daytime periods. Hepatic steatosis degree (healthy: n = 698; mild steatosis: n = 242; moderate/severe steatosis: n = 240) was determined by magnetic resonance imaging proton density fat fraction. Multivariate ordinal logistic regression models were conducted to estimate the associations between CGM metrics and steatosis degree. Machine learning models were employed to evaluate the predictive performance of CGM metrics for steatosis degree.ResultsMean blood glucose, coefficient of variation (CV) of glucose, mean amplitude of glucose excursions (MAGE), and mean of daily differences (MODD) were positively associated with steatosis degree, with corresponding odds ratios (ORs) and 95% confidence intervals (CIs) of 1.35 (1.17, 1.56), 1.21 (1.06, 1.39), 1.37 (1.19, 1.57), and 1.35 (1.17, 1.56) during all-day period. Notably, lower daytime time in range (TIR) and higher nighttime TIR were associated with higher steatosis degree, with ORs (95% CIs) of 0.83 (0.73, 0.95) and 1.16 (1.00, 1.33), respectively. For moderate/severe steatosis (vs. healthy) prediction, the average area under the receiver operating characteristic curves were higher for the nighttime (0.69) and daytime (0.66) metrics than that of all-day metrics (0.63, P < 0.001 for all comparisons). The model combining both nighttime and daytime metrics achieved the highest predictive capacity (0.73), with nighttime MODD emerging as the most important predictor.ConclusionsHigher CGM-derived mean glucose and glycemic variability were linked with higher steatosis degree. CGM-derived metrics during nighttime and daytime provided distinct and complementary insights into hepatic steatosis.

More Movement, Less Bodily Pain? Findings From a Large, Representative Multi-District Aging Study in Ghana

ObjectiveThe association of physical activity (PA) with pain has been characterized. Although the literature largely comes from high-income countries, the mechanistic pathways underlying this association remain unknown, particularly in low- and middle-income countries (LMICs). We aim to examine the association between meeting the World Health Organization (WHO) PA guidelines and pain among aging adults and identify the factors that may mediate this association. DesignWe used a quantitative cross-sectional study design. Setting and ParticipantsRepresentative data from adults aged ≥50 years who participated in the Aging, Health, Psychological Well-being, and Health-seeking Behavior Study were used. MethodsPA was defined using the International Physical Activity Questionnaire (IPAQ). Self-reported pain experience using a cross-culturally validated item over the past month assessed pain severity. Adjusted multivariable ordinal logistic regression and mediation models quantified the hypothesized associations. ResultsThe analysis included 1201 adults (mean 66.1 ± 11.9 years; 63.3% female). After adjusting for confounders, adhering to the WHO-recommended PA guidelines was associated with 42% lower odds for severe/extreme pain [odds ratio (OR) 0.58, 95% CI 0.44-0.77]. The association was much stronger among men (OR 0.52, 95% CI 0.31-0.85) than women (OR 0.60, 95% CI 0.42-0.87). The PA-pain association was explained by functional limitations (84.7%), self-rated health (76.6%), sleep problems (20.4%), and injury (6.6%). Conclusions and ImplicationsMoving more was associated with less pain in older adults from LMICs. Interventions for pain management in old age may focus on enhancing compliance with PA doses, particularly in LMICs. However, longitudinal data will need to confirm these findings.

Managing Children and Adolescents with Type 1 Diabetes and Coexisting Celiac Disease: Real-World Data from a Global Survey.

Celiac disease (CD) is among the diseases most commonly associated with type 1 diabetes (T1D). This study aimed to evaluate the worldwide practices and attitudes of physicians involved in pediatric diabetes care regarding diagnosing and managing CD in children with T1D. The 30-item survey was conducted between July and December 2023 aimed at targeting pediatricians with special interest in T1D and CD. It was shared by the JENIOUS - young investigators group of the International Society of Pediatric and Adolescent Diabetes (ISPAD) - and the YES - early career group of the European Society for Pediatric Endocrinology (ESPE). Overall, 180 physicians (67.8% female) from 25 countries responded. Among respondents, 62.2% expected sustaining optimal glycemic control in children with T1D and CD (T1D + CD) to be more difficult than in children with T1D alone. Majority (81.1%) agreed that more specific guidelines are needed. The follow-up routine for patients with T1D + CD differed, and one-quarter of physicians scheduled more frequent follow-up checkups for these patients. Seventy percent agreed multidisciplinary outpatient clinics for their follow-up is needed. In the multivariate ordinal logistic regression model, a statistically significant predictor of a higher degree of practice according to ISPAD 2022 guidelines was a higher level of country income (OR = 3.34; p &lt; 0.001). These results showed variations in physicians' practices regarding managing CD in children with T1D, emphasizing the need for more specific guidelines and intensive education of physicians in managing this population, especially in lower-income countries. Our data also suggest the implementation of multidisciplinary outpatient clinics for their follow-up.

Severity of Trachomatous Scarring among Adults in Trachoma-Endemic Amhara Region of Ethiopia.

Trachomatous scarring has been shown to progress regardless of active ocular Chlamydia trachomatis infection, indicating that scarring drivers may be unrelated to ongoing transmission. Although scarring prevalence is commonly associated with older age and female sex, less is known about other potential contributors to its development. This study identified and assessed risk factors associated with scarring magnitude in a trachoma-endemic setting, utilizing a five-point photographic scale (S0-S4). During 2017 trachoma surveys of Amhara, Ethiopia, photographers captured left and right conjunctival images of adults (ages 15 years and older) from 10 districts. Subsequently, two graders independently assessed photographs for scarring, with discrepancies adjudicated by an expert grader. Scarring scores for 729 individuals were aggregated from the eye level to the participant level, excluding 17 participants because of poor photograph quality. Among those with scarring, most cases (20.4%) were severe (S4, comprising more than 90% of the tarsal conjunctiva) compared with the prevalence of moderate S3-A/B (11.2%), S2 (8.3%), and mild S1 (19.2%). The youngest group (ages 15-19 years) exhibited all scarring stages. Older participants (60 years and older) experienced a greater burden of severe scarring (S4 prevalence: 32.6%) than their younger (15-19 years) counterparts (6.2%). Multivariate ordinal logistic regression models indicated female sex, increasing age, and district-level trachomatous follicular-inflammation prevalence were significant predictors of scarring severity. Trachomatous scarring and its progression to trichiasis, may prove a barrier to meeting WHO timelines for trachoma elimination and will necessitate ongoing surveillance and interventions after elimination thresholds have been met.

The Association of Delayed Milk Ejection with Milking Performance in Holstein Cows in a Large Dairy Herd with Suboptimal Premilking Teat Stimulation.

The primary objective was to investigate the association between delayed milk ejection (DME) and the average milk flow rate, milking unit-on time, and duration in a low milk flow rate in Holstein dairy cows in a large dairy herd with suboptimal premilking teat stimulation. Our second objective was to study the association between peak lactation milk yield and the occurrence of DME. This longitudinal field study was conducted at a 4300-cow dairy farm with a thrice-daily milking schedule over a 1-week period. We analyzed data from 61,677 cow milking observations from 2937 cows. Delayed milk ejection was defined as present if the 30-60 s milk flow rate was ≤3.1 kg/min. The mean average milk flow rate (MAMF, kg/min), mean milking unit-on time (MMUT, s), and mean duration of a low milk flow rate (MLMF, s) were calculated as the mean values from the 21 milking observations. General linear multivariable models revealed associations of DME with MAMF, MMUT, and MLMF. A multivariable ordinal logistic regression model revealed an association between peak lactation milk yield and DME. Cows with lower peak lactation milk yield had greater odds of exhibiting a higher frequency level of DME. The observed associations between DME and milking performance indices suggest that DME can negatively affect milking and parlor efficiency. Peak lactation milk yield may serve as a proxy to estimate cows' risk of recurrent DME. Future research is warranted to test if alleviating DME through, for example, a modified milking routine influences the milking performance indices described herein.

Early predictive factors in routine clinical practice for rituximab therapy response in patients with rheumatoid arthritis.

Identifying early predictive factors of how rheumatoid arthritis (RA) patients respond to rituximab (RTX) treatment is crucial for both individual treatment outcome and the improvement of clinical practice overall. This study aimed to identify early predictive factors available in standard clinical practice for predicting RTX treatment outcomes in RA patients. Data on seventy patients diagnosed with RA treated with RTX (two 1,000 mg doses 2 weeks apart or two 500 mg doses 2 weeks apart) were retrospectively collected. Baseline information collected at the initiation of RTX treatment included patient characteristics such as age, sex, disease duration, disease activity, Health Assessment Questionnaire score, erythrocyte sedimentation rate, C-reactive protein, and serological status regarding rheumatoid factor (RF) and anti-cyclic citrullinated protein antibodies (ACPA). Clinical responses were analyzed 6 months after RTX initiation using the European Alliance of Associations for Rheumatology criteria. Potential predictors associated with positive RTX response at 6 months were identified using a multivariate ordinal logistic regression model. The analysis showed that persistently active RA disease, Disease Activity Score with 28-joint count (DAS28) values at the treatment onset and after 3 months, along with erythrocyte sedimentation rate at treatment initiation, were negatively correlated with the response to RTX therapy (p < 0.05). All these correlations were statistically significant at the 99% confidence interval. The correlation and logistic regression analyses indicate that there are no significant association between RF and ACPA concerning therapy response, despite a higher number of RTX responders in the seropositive groups. Additionally, the study emphasizes the prognostic significance of the DAS28 value at treatment initiation in predicting therapy response at 6 months. The optimal model for predicting RTX response at 6 months involves the interaction of all clinical factors examined in this study, as revealed by the analysis of multiple variables.

Census tract–level social determinants of health and variability in stage at diagnosis of bladder, breast, and non-small cell lung cancers.

11170 Background: Social determinants of health (SDOH) can identify underserved populations, inform the patient journey, and identify opportunities to improve outcomes. In cancer, SDOH may influence stage at diagnosis through exposure to risk factors for aggressive cancer and decreased access to care. However, privacy restrictions can limit research with SDOH in real world datasets, which often only know a patient’s broad location. Methods: Geocoding was undertaken on addresses of patients diagnosed between 1/1/17 and 9/30/22 with bladder (n=6078), breast (n=28,958), and non-small cell lung cancer (NSCLC, n=14,957) in the Syapse Learning Health Network of U.S. community health systems. Location was linked to five indicators of SDOH at the census tract level: the Social Vulnerability Index (SVI); percent (%) of tract spending &gt;30% of income on housing; % of tract with broadband internet; county designation as a primary care shortage area; and rural urban commuting area (RUCA). Nested multivariable ordinal logistic regression models estimated independent associations between demographic, clinical, and SDOH factors with stage at initial cancer diagnosis. The statistical significance of including SDOH variables was assessed using a likelihood ratio test (LRT) comparing the models before and after including SDOH. Results: After successful geocoding and linkage to census tracts, there were marginal differences in the distribution of the five SDOH across stages at diagnosis within the three tumor types. In multivariable models adjusted for year of diagnosis, age, sex, race, ethnicity, smoking, and primary payor, among the tested SDOH measures, county designated as primary care shortage (proportional odds ratios (OR): 1.31, 95% confidence interval (95% CI): 1.05-1.64), and non-metropolitan area (0.69, 0.50-0.94) were statistically significantly associated with stage at bladder cancer diagnosis. SVI (1.52, 1.31-1.76) was also statistically significantly associated and directly proportional with stage at breast cancer diagnosis. Among patients with NSCLC, broadband internet access (1.01, 1.00-1.01), primary care shortage (1.15, 1.03-1.27) and non-metropolitan area (0.85, 0.74-0.99) were statistically associated with stage at diagnosis. In all three tumors, LRTs found that models with SDOH were statistically significantly better at predicting stage at diagnosis than those without (p &lt;0.001). Conclusions: Using granular, census-tract geospatial resolution for geocoding of SDOH, we find that patients who live in greater vulnerability are diagnosed at later stages than those who live in areas of less vulnerability, suggesting that consideration of SDOH variables is important in assessing stage at diagnosis and identify patients with an unmet need for outreach and screening.

Association of symptom severity, anxiety, and health literacy with portal usage among patients with cancer.

1567 Background: Patient portals within electronic health records (EHR) have become increasingly essential in promoting patient-centered oncology care. They offer opportunities for completing symptom surveys that can aid in symptom management, facilitate patient-clinician communication, and enable health monitoring. As part of the Northwestern University Improving the Management of Symptoms During and Following Cancer Treatment (NU IMPACT) study, we examined the relationship between symptom severity, anxiety, health literacy, and shared decision-making with usage of three common features in the patient portal (portal). Methods: This study included adults diagnosed with cancer in the NU IMPACT study. Participants completed baseline surveys on socio-demographics, health literacy (Single Item Literacy Screener), anxiety (Patient-Reported Outcomes Measurement Information System; PROMIS), with high anxiety scored as ≥60 vs &lt;60, shared decision-making (CollaboRATE), and symptom severity (Patient-Reported Outcome version of the Common Terminology Criteria for Adverse Events; PRO-CTCAE). Portal activities one year following enrollment were extracted from EHRs. Portal usage patterns of patient-initiated messages, portal lab and clinical note views were counted as number of days the activity was used over the year and categorized as tertiles. Multivariable ordinal logistic regression models were used to examine associations between survey measures and ordered tertiles of patient portal activities. Results: Participants (N=3,464) were mostly female (66%), White (85%), with a mean (SD) age of 61(13) years old. The frequency ranges of patient-initiated messages over a year were 0-163 days, labs viewed were 0-350 days, and clinical notes viewed were 0-103 days. Participants with at least one severe or very severe symptom on the PRO-CTCAE were more likely to be in higher tertiles of patient-initiated portal messages (OR: 1.37, 95% CI: 1.17-1.61), lab views (OR: 1.48, 95% CI: 1.26-1.74), and clinical note views (OR: 1.19, 95% CI: 1.01-1.40). Participants with higher levels of anxiety were more likely to be in higher tertiles of patient-initiated portal messages (OR: 1.54, 95% CI: 1.28-1.85), lab views (OR: 1.22, 95% CI: 1.01-1.46), and clinical note views (1.42, 95% CI: 1.18-1.71). Participants with higher health literacy were more likely to be in higher tertiles of patient-initiated portal messages (OR: 1.38, 95% CI: 1.09-1.74) and lab views (OR: 1.34, 95% CI: 1.06-1.69), but not note views (OR: 1.03, 95% CI: 0.82-1.30). Shared decision making was not associated with portal usage. Conclusions: Patients with greater symptom severity and anxiety may turn to the portal for symptom management. Certain features within the portal may demand a higher level of health literacy for engagement. Understanding the factors that promote and impede portal usage can help optimize cancer care delivery.

Abstract PO5-11-07: Short- and long-term treatment-related effects on physical health in breast cancer survivors

Abstract Background. Breast cancer (BC) survivors experience greater physical health declines compared to cancer-free women of the same age, which is thought to be caused by their cancer and its treatment. However, the short- and long-term effects of different types of treatment on physical health are not well understood, especially for hormone therapy. Moreover, whether specific aspects of physical health, such as activities of daily living (ADL), are affected by treatment is unknown. Methods. This analysis was conducted in the Cancer Prevention Study 3 (CPS-3), a prospective cohort study with approximately 300,000 participants enrolled between 2006-2013. Participants returned a baseline survey and triennial follow-up surveys beginning in 2015. Incident, non-metastatic, female BC patients were included in this analysis if they returned a survey at least 90 days after their diagnosis through June 30, 2020, the end of follow-up for this analysis. All 2,551 BC patients included in this study had pre- and post-diagnostic surveys and underwent breast surgery if they were diagnosed with invasive disease. They were matched to up to five cancer-free women on age and year of survey returns (N=12,704). Self-reported physical health and ADL were assessed using the Patient-Reported Outcomes Measurement Information System (PROMIS) Global Health Scale instrument. Treatment information was obtained from cancer registries and self-report and included surgery, chemotherapy, radiation and hormonal therapy. Odds Ratios (OR) and 95% confidence intervals (CI) between treatment received and self-reported physical function were estimated using multivariable ordinal logistic regression models at three time points relative to the case’s diagnosis date: before diagnosis, within 90 days to 5 years after diagnosis, and more than 5 years after diagnosis. Results. Median age at diagnosis was 56 years (interquartile range (IQR): 50-62). Approximately 82% of the tumors were hormone-receptor positive. The median times from the three physical health measurements to diagnoses were 1.7 years before diagnosis (IQR: 0.9 – 2.7), 1.7 years (IQR: 0.9 – 2.5) from 90 days to 5 years post-diagnosis and 6.6 years (IQR: 5.5-7.0) for more than 5 years post-diagnosis. There was a significant risk of physical decline within 5 years of diagnosis for women who received chemotherapy or hormonal therapy compared to cancer-free women (Table 1). However, more than 5 years after diagnosis, the risk of physical decline was only observed among those who received chemotherapy. Treatment combinations that included chemotherapy were also associated with a decline in the ability to carry out ADL during the first 5 years. Patients who received hormonal therapy without chemotherapy were comparable to cancer-free women in their ability to carry out ADL during the post-diagnosis period. BC patients who received surgery only or surgery and radiation reported physical health and ADLs similar to cancer-free women within 5 and &amp;gt;5 years after diagnosis. Conclusions. Post-diagnosis physical health decline is time and treatment dependent. Patients receiving chemotherapy have the greatest overall physical health decline and it is long-lasting. Beyond five years from diagnosis, physical health of breast cancer patients who did not receive chemotherapy were similar to cancer-free women. Strategies are needed to help BC patients receiving chemotherapy overcome the long-term physical function limitations of their cancer treatment. Table. Associations between treatment received and physical health and ability to carry out activities of daily living. OR: Odds Ratio; CI: confidence interval; chemo: chemotherapy; radio: radiation therapy; horm: hormonal therapy. *Associations were adjusted for matching factors, which included age and time from diagnosis to measurement of physical health. Citation Format: Clara Bodelon, Matthew Masters, Lauren McCullough, Alpa Patel, Lauren Teras. Short- and long-term treatment-related effects on physical health in breast cancer survivors [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-11-07.

Left ventricular systolic dysfunction predicts clinical prognosis in patients with acute ischemic stroke after intravenous thrombolysis.

Although intravenous recombinant tissue plasminogen activator (rt-PA) thrombolysis is the most effective early treatment for acute ischemic stroke (AIS), outcomes vary greatly among patients. Left ventricular systolic dysfunction (LVSD) is prone to distant organ ischemia and may be a predictor for poor prognosis in AIS patients undergoing intravenous thrombolysis (IVT). Our aim was to investigate the predictivity of LVSD diagnosis (as measured by left ventricular ejection fraction (LVEF)) on 90-day clinical outcomes in AIS patients undergoing thrombolysis. The current prospective cohort study continuously enrolled 273 AIS patients from the National Stroke Prevention and Treatment Engineering Management Special Database who underwent IVT and completed echocardiography within 24 h of admission between 2021 and 2023. LVSD was examined by evaluation of the echocardiographic LVEF values using Simpson's biplane method of discs in line with international guidelines, and defined as a LVEF value < 50%. Multivariable ordinal logistic regression model was performed to analyze the association between LVEF and functional outcome at 3 months. Restricted cubic spline (RCS) was used to examine the shape of the dose-response association between reduced LVEF and poor functional outcomes. Subgroup analysis was also employed to further verify the reliability and practicability of the results. Baseline data analysis showed LVSD patients had more comorbidities including on multivariate analyses, LVSD (OR 2.78, 95% CI 1.23 to 6.24, P=0.014), pre-existing diabetes mellitus (OR 2.08, 95% CI 1.11 to 3.90, P=0.023) and NIHSS on arrival (OR 1.31, 95% CI 1.21 to 1.49, P<0.001) were independent predictors of poor functional outcomes (mRS ≥ 3) at 3 months. Multivariable-adjusted spline regression indicated a linear dose-response association between LVEF after IVT and poor functional outcomes (p for linearity < 0.001), with the optimal cutoff values of LVEF being 0.48. Our finding indicated that AIS patients with LVSD after IVT had poorer outcomes, suggesting the need to monitor and optimize LVEF in stroke management.

Assessing health-related quality of life of Chinese population using CQ-11D

PurposeThis study aimed to assess the health-related quality of life (HRQoL) of the Chinese population by using the Chinese medicine quality of life-11 dimensions (CQ-11D) questionnaire and to identify factors associated with HRQoL.MethodsThe data was derived from a survey conducted by the Institute of Pharmacoeconomics Evaluation at Beijing University of Chinese Medicine on the quality of life of the Chinese population. The sex and age of respondents were considered through quota sampling. Demographic, socioeconomic, and health indicators were collected using the structured questionnaire. We performed bivariate analyses first to examine the associations between the above factors and the HRQoL of respondents measured by the CQ-11D. Multivariate linear regression and ordinal logistic regression models were established to analyze the factors (demographic, socioeconomic, and health indicators) differences in HRQoL, as well as the risk of each group reporting problems across the 11 dimensions of CQ-11D.ResultsFrom February 2021 to November 2022, a total of 7,604 respondents were involved and 7,498 respondents were included. The sample approximated the general adult Chinese population in terms of age, sex, and district of residence, and each geographic distribution ranged from 9.71 to 25.54%. Of the respondents, 45.84% were male, and 89.82% were Han ethnicity. The mean utility score ranged from 0.796 to 0.921 as age increased. According to the respondents, most health problems were identified in the PL (fatigue) (70.16%) and SM (quality of sleep) (63.63%) dimensions. The CQ-11D index scores varied with the demographic and socioeconomic characteristics of respondents, except for ethnicity (p > 0.05) and income (p > 0.05). The multivariate analysis revealed significant negative associations between health utility scores and various factors. These factors include sex (female), age over 65, belonging to ethnic minorities, rural household registration, being widowed or divorced, having a primary school education or below, being a student or unemployed, having a low income of 0–1,300, engaging in smoking or drinking, limited participation in physical activities, experiencing changes in self-perceived health status compared to the previous year, and having chronic diseases. The odds of respondents reporting problems in 11 dimensions varied among different socio-demographic groups.ConclusionsThis study reports the first Chinese population norms for the CQ-11D derived using a representative sample of the Chinese general population. Self-reported health status measured by the CQ-11D varies among different socio-economic groups. In addition to participation a physical activity and the presence of chronic disease, smoking and drinking also significantly influence HRQoL.

Multivariate mixed-effects ordinal logistic regression models with difference-in-differences estimator of the impact of WORTH Yetu on household hunger and socioeconomic status among OVC caregivers in Tanzania.

Although most of the livelihood programmes target women, those that involve women and men have been evaluated as though men and women were a single homogenous population, with a mere inclusion of gender as an explanatory variable. This study evaluated the impact of WORTH Yetu (an economic empowerment intervention to improve livelihood outcomes) on household hunger, and household socioeconomic status (SES) among caregivers (both women and men) of orphaned and vulnerable children (OVC) in Tanzania. The study hypothesized that women and men respond to livelihood interventions differently, hence a need for gender-disaggregated impact evaluation of such interventions. This is a secondary analysis of longitudinal data, involving caregivers' baseline (2016-2019) and follow-up (2019-2020) data from the USAID Kizazi Kipya project in 25 regions of Tanzania. Two dependent variables (ie, outcomes) were assessed; household hunger which was measured using the Household Hunger Scale (HHS), and Socioeconomic Status (SES) using the Principal Component Analysis (PCA). WORTH Yetu, a livelihood intervention implemented by the USAID Kizazi Kipya project was the main independent variable whose impact on the two outcomes was evaluated using multivariate analysis with a multilevel mixed-effects, ordinal logistic regression model with difference-in-differences (DiD) estimator for impact estimation. The analysis was based on 497,293 observations from 249,655 caregivers of OVC at baseline, and 247,638 of them at the follow-up survey. In both surveys, 70% were women and 30% were men. Their mean age was 49.3 (±14.5) years at baseline and 52.7 (±14.8) years at the follow-up survey. Caregivers' membership in WORTH Yetu was 10.1% at the follow-up. After adjusting for important confounders there was a significant decline in the severity of household hunger by 46.4% among WORTH Yetu members at the follow-up compared to the situation at the baseline (adjusted Odds Ratio (aOR) = 0.536, 95% Confidence Interval (CI) [0.521, 0.553]). The decline was 45.7% among women (aOR = 0.543 [0.524, 0.563]) and 47.5% among men (aOR = 0.525 [0.497, 0.556]). Regarding SES, WORTH Yetu members were 15.9% more likely to be in higher wealth quintiles at the follow-up compared to the situation at the baseline (aOR = 1.159 [1.128, 1.190]). This impact was 20.8% among women (aOR = 1.208 [1.170, 1.247]) and 4.6% among men (aOR = 1.046 [0.995, 1.101]). WORTH Yetu was associated with a significant reduction in household hunger, and a significant increase in household SES among OVC caregivers in Tanzania within an average follow-up period of 1.6 years. The estimated impacts differed significantly by gender, suggesting that women and men responded to the WORTH Yetu intervention differently. This implied that the design, delivery, and evaluation of such programmes should happen in a gender responsive manner, recognising that women and men are not the same with respect to the programmes.

Intelligence, academic achievement, and pediatric injury among a large sample of kindergartners

Objectives: Previous work reports mixed results on the roles of intelligence and academic achievement on child injury risk. Some early research offered evidence of links between lower levels of intelligence and higher rates of pediatric injury, but other work has suggested there is no relation between the variables. This study investigated the matter further, in the context of a large sample and with strong statistical power. Methods: Data from a national sample of 8002 kindergartners in the United States, all from low-income families, were analyzed. Intelligence was estimated using the Peabody Picture Vocabulary Test-Revised and academic achievement with the Reading and Mathematics scales of the Woodcock Johnson Psycho- Educational Battery-Revised. Mothers reported whether children had experienced an injury in the previous year that required professional medical attention. Results: Results of both bivariate and multivariate ordinal logistic regression models suggest higher levels of reading achievement and of mathematics achievement were associated with slightly reduced risk of pediatric injury. Intelligence was not related to injury history. Conclusions: Academic achievement, but not intelligence, appears to relate in a small but statistically significant inverse manner with risk for pediatric injury among kindergartners from a low-income background. Possible mediators include children’s ability to learn and remember safety-related rules or the fact that parents who encourage learning in the household might also safeguard children more effectively.

Influence of Infarct Morphology and Patterns on Cognitive Outcomes After Endovascular Thrombectomy.

To assess the association of qualitative and quantitative infarct characteristics and 3 cognitive outcome tests, namely the Montreal Cognitive Assessment (MOCA) for mild cognitive impairment, the Boston Naming Test for visual confrontation naming, and the Sunnybrook Neglect Assessment Procedure for neglect, in large vessel occlusion stroke. Secondary observational cohort study using data from the randomized-controlled ESCAPE-NA1 trial (Safety and Efficacy of Nerinetide in Subjects Undergoing Endovascular Thrombectomy for Stroke), in which patients with large vessel occlusion undergoing endovascular treatment were randomized to receive either intravenous Nerinetide or placebo. MOCA, Sunnybrook Neglect Assessment Procedure, and 15-item Boston Naming Test were obtained at 90 days. Total infarct volume, gray matter, and white matter infarct volumes were manually measured on 24-hour follow-up imaging. Infarcts were also visually classified as either involving the gray matter only or both the gray and white matter and scattered versus territorial. Associations of infarct variables and cognitive outcomes were analyzed using multivariable ordinal or binary logistic regression models. Of 1105 patients enrolled in ESCAPE-NA1, 1026 patients with visible infarcts on 24-hour follow-up imaging were included. MOCA and Sunnybrook Neglect Assessment Procedure were available for 706 (68.8%) patients and the 15-item Boston Naming Test was available for 682 (66.5%) patients. Total infarct volume was associated with worse MOCA scores (adjusted common odds ratio per 10 mL increase, 1.05 [95% CI, 1.04-1.06]). After adjusting for baseline variables and total infarct volume, mixed gray and white matter involvement (versus gray matter-only adjusted common odds ratio, 1.92 [95% CI, 1.37-2.69]), white matter infarct volume (adjusted common odds ratio per 10 mL increase 1.36 [95% CI, 1.18-1.58]) and territorial (versus scattered) infarct pattern (adjusted common odds ratio, 1.65 [95% CI, 1.15-2.38]) were associated with worse MOCA scores. Results for Sunnybrook Neglect Assessment Procedure and 15-item Boston Naming Test were similar, except for the territorial infarct pattern, which did not reach statistical significance in multivariable analysis. Besides total infarct volume, infarcts that involve the white matter and that show a territorial distribution were associated with worse cognitive outcomes, even after adjusting for total infarct volume.

Prevalence and Risk Factors for Renal Insufficiency among Adults Living with HIV in Tanzania: Results from a Cross-Sectional Study in 2020-2021.

With improved survival, adults living with HIV (ALHIV) are increasingly likely to experience age-related and HIV-related comorbidities, including renal insufficiency. Other risk factors for renal insufficiency (high blood pressure (BP), obesity, diabetes, and dyslipidemia) are also growing more common among ALHIV. To determine the prevalence of renal insufficiency (defined as an eGFR < 60 mL/min/1.73 m2) and factors associated with reduced eGFR, we conducted a cross-sectional study at six HIV clinics in Dar-es-Salaam, Tanzania. We applied multivariable (MV) ordinal logistic regression models to identify factors associated with reduced eGFR and examined the interaction of age with BP levels. Among the 450 ALHIV on ART analyzed [26% males; median age 43 (IQR: 18-72) years; 89% on tenofovir-containing ART; 88% HIV viral load ≤50 copies/mL], 34 (7.5%) had renal insufficiency. Prevalence was higher among males (12%) vs. females (6%), p = 0.03; ALHIV ≥50 (21%) vs. <50 years (2.5%), p < 0.001; those with high [≥130/80 mmHg (15%)] vs. normal [<120/80 mmHg (4%)] BP, p < 0.01 and those with dyslipidemia (10%) vs. those without (4.5%), p < 0.03. After adjusting for covariates, age (in years) was the only covariate with a statistically significant association with reduced eGFR (OR = 1.09 (1.07-1.12), p < 0.001). No significant interaction between age and BP was found. Interventions to increase routine screening for renal insufficiency, especially among older ALHIV, and improve BP control are critical to reducing kidney disease-related morbidity and mortality.

Factors affecting different COVID-19 outcomes in patients with systemic lupus erythematosus during the second pandemic wave of COVID-19 in China.

To investigate characteristics associated with different COVID-19 outcomes of people with systemic lupus erythematosus (SLE) and COVID-19 during the second pandemic wave of COVID-19 in China. In this retrospective study, people with SLE and COVID-19 who visited the First Affiliated Hospital of Nanchang University from December 2022 and February 2023 were subjected to this study. The three possible outcomes were listed in order of ordinal severity: (1) not hospitalized, (2) hospitalized but not receiving oxygenation, and (3) hospitalized with any ventilation or oxygenation. A multivariable ordinal logistic regression model was built to examine the association between COVID-19 severity and demographic traits, medications, comorbidities, and disease activity. Furthermore, among the 301 SLE patients included in our study, only two patients experienced mortality. In order to maintain statistical rigor, we have included these two deceased patients in the outcome measure of hospitalized with any ventilation or oxygenation. A total of 301 patients with SLE were enrolled in this study. The multivariate ordinal logistic regression analyses indicated that high SLE disease activity (vs remission; OR 39.04, 95% CI 3.08 to 494.44, p = .005) was associated with more severe outcomes. Three doses of COVID-19 vaccination (OR 0.19, 95% CI 0.07 to 0.51, p = .001), glucocorticoids dose (1-5mg/day 0.14, 0.03 to 0.73, p = .020, and 6-9mg/day 0.12, 0.02 to 0.61, p = .010), and more intensive immunosuppression drugs (0.34, 0.12 to 0.97, p = .044) were associated with better outcomes. In age-adjusted and sex-adjusted models, telitacicept (6.66, 1.35 to 32.86, p = .020) and rituximab (7.81, 1.87 to 32.66, p = .005) were associated with more severe outcomes. Hydroxychloroquine (0.47, 0.25 to 0.88, p = .018) was associated with favorable outcomes. Different COVID-19 outcomes in people with SLE are mostly driven by COVID-19 vaccination, medications, and activity SLE. More importantly, three doses of COVID-19 vaccination may be associated with better outcomes.

Insulinemic potential of lifestyle is associated with depression and anxiety in adults: A large community-based study

BackgroundWe aimed to investigate the association between an empirical lifestyle index for hyperinsulinemia (ELIH), empirical lifestyle index for insulin resistance (ELIR), and depression and anxiety in an adult Iranian population. MethodsIn this cross-sectional study, a total of 6450 participants, aged 35–65 years were recruited as part of the MASHAD cohort study. Dietary intakes were assessed using a validated food frequency questionnaire (FFQ). Depression and anxiety were screened using Beck Depression Inventory (BDI) and Beck Anxiety Inventory (BAI). ELIH and ELIR were calculated using dietary intake, body mass index, and physical activity information. Multivariable ordinal logistic regression models were applied to determine the association between ELIH, ELIR, and depression and anxiety severity. ResultsIn a fully adjusted model, participants with the highest ELIH quartile had a higher odds of more severe depression and anxiety compared to those in the lowest category (OR = 1.44; 95 % CI = 1.22–1.71 and OR = 1.62; 95 % CI = 1.37–1.25, respectively). Participants with the highest ELIR had higher odds of more severe depression and anxiety compared to those in the lowest category (OR = 1.22; 95 % CI = 1.04–1.43 and OR = 1.21; 95 % CI = 1.03–1.42, respectively). LimitationsThe assessment of dietary intake and mental health by questionnaires may increases the rate of misclassification. Due to the study's cross-sectional nature, causal relationships cannot be established. ConclusionThere was a significant positive association between the hyperinsulinemia and insulin resistance potential of lifestyle and severity of depression and anxiety among Iranian adults.

NONCOMBAT AND COMBAT MILITARY SERVICE EXPERIENCES, HEARING DIFFICULTY, AND DIFFICULTY REMEMBERING/CONCENTRATING

Abstract Hearing difficulty is widespread among U.S. veterans. Research documents a strong association between hearing difficulty and a range of cognitive outcomes. Most research that examines cognitive outcomes in older adults does not distinguish veterans from non-veterans or non-combat from combat veterans. We use data on Sample Adults aged 50+ from the 2021 National Health Interview Survey to examine: the associations between military service experiences (non-veteran, non-combat veteran, combat veteran) and difficulty remembering/concentrating (none, some, a lot/cannot do at all) net of sociodemographic controls; the extent to which observed associations are mediated by hearing difficulty (none, some, a lot/cannot do at all); whether military service experiences moderate the association between hearing loss and difficulty remembering and concentrating. This latter hypothesis is motivated by the notion that service-related hearing loss may occur earlier in the life course and therefore be more detrimental to cognitive outcomes. Analyses are weighted; standard errors are adjusted for the complex sampling design. Preliminary results from multivariable ordinal logistic regression models that include sociodemographic controls indicate that: non-combat and combat veterans have significantly more difficulty remembering/concentrating than non-veterans; the associations between non-combat and combat military service experiences and increased difficulty remembering/concentrating are reduced 20% when hearing difficulty is added to the model, but remain statistically significant; hearing difficulty is strongly associated with difficulty remembering/concentrating; and military service experiences moderate the associations between hearing difficulty and difficulty remembering/concentrating. For example, the highest level of hearing difficulty is more strongly associated with difficulty remembering/concentrating among combat veterans than non-combat veterans.

Abstract B054: Impact of health behaviors, health outcomes, and socioeconomic status on sleep duration among gynecologic cancer survivors

Abstract Background: Emerging evidence suggests sleep duration may be associated with health outcomes among cancer survivors. There have been few studies analyzing sleep among gynecologic cancer survivors (GCS). We defined GCS as respondents who reported being told they had a form of gynecologic cancer (cervical, endometrial, and ovarian cancer) in their life. This study aimed to explore the associations between health outcomes, health behaviors, and socioeconomic status (SES) on sleep duration in GCS. Methods: We used the 2016-2020 Behavioral Risk Factor Surveillance System (BRFSS) datasets to evaluate the impacts of health outcomes, health behaviors, and SES on sleep duration among GCS. We categorized sleep duration into Short Sleep Duration (&amp;lt;7 hours) (SSD), Long Sleep Duration (&amp;gt;8 hours) (LSD), and Optimal Sleep Duration (7-8 hours) (OSD). Bivariate analyses including chi-square tests, were used to examine associations between health outcomes, behaviors, SES, and sleep duration. A multivariable ordinal logistic regression model was produced to assess the adjusted odds ratios (aOR) for sleep duration, considering health outcomes, behaviors, and SES, while controlling for confounders. Results: Our sample (N=10222) was primarily composed of cervical (n=8856) followed by endometrial (n=1198) and ovarian (n=168) cancer survivors. Most of the sample was 65+ years old (65.93%), followed by the age categories 40-64 years old (32.17%) and 18-39 years old (1.90%). Regarding race, most of the sample was non-Hispanic White (95.68%), followed by non-Hispanic Black (2.54%) and Hispanic (1.78%). We found significant associations between health outcomes, health behaviors, and SES in relation to sleep duration among GCS. GCS who reported depression had higher odds of both LSD (aOR=1.93, 95% CI: 1.45-2.58) and SSD (aOR=1.57, 95% CI: 1.29-1.92). GCS engaging in exercise were more likely to report LSD (aOR=1.53, 95% CI: 1.15-2.04). Current smokers had an increased likelihood of SSD (aOR=1.47, 95% CI: 1.11-1.95). Unemployment was associated with an increased likelihood of LSD (aOR=2.67, 95% CI: 1.87-3.82). Additionally, GCS with incomes &amp;lt;$25,000 had significantly higher odds of SSD compared to GCS earning &amp;gt;$75,000 (aOR=1.82, 95% CI: 1.38-2.39). For health outcomes, GCS with fair/poor health were significantly more likely to report SSD compared to GCS with excellent health (aOR=1.63, 95% CI: 1.20-2.20). GCS experiencing 14 or more poor health days in a month also had higher odds of SSD (aOR=1.31, 95% CI: 1.11-1.55). Finally, GCS residing in rural areas had an association with SSD (aOR=1.22, 95% CI: 1.03-1.44) in comparison to GCS residing in urban areas. Discussion: Our findings illustrate a complex interplay of health outcomes, health behaviors, and SES contributing to sleep duration among GCS. This underscores the crucial need to address sleep duration to improve health outcomes, reduce disparities, and increase quality of life for all GCS. Opportunities for future research include sleep hygiene interventions that may mitigate these effects and improve sleep among GCS. Citation Format: LaShae D. Rolle, Maurice J. Chery, Alyssa Mercadel, Matthew P. Schlumbrecht, Sophia H.L. George, Tracy E. Crane. Impact of health behaviors, health outcomes, and socioeconomic status on sleep duration among gynecologic cancer survivors [abstract]. In: Proceedings of the 16th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2023 Sep 29-Oct 2;Orlando, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2023;32(12 Suppl):Abstract nr B054.

The circulating sclerostin as a biomarker for brain white matter damage in hemodialysis patients

AbstractBackgroundPatients with kidney disease, especially those receiving hemodialysis, presented a higher risk of cognitive impairment and brain white matter damage than those without kidney disease. Previous research suggests an association between lower bone marrow densities and cognitive impairments. However, the study investigating the relationship between bone turnover biomarkers and brain white matter damage in hemodialysis patients is sparse.MethodsThe enrolled hemodialysis participants were examed on the Fazekas scale of brain magnetic resonance imaging (MRI) white matter lesions and measured 8 bone turnover proteins using the Luminex bead‐based multiplex assay. The Fazekas scale of periventricular lesions was from grade 0 to grade 3. Multivariable ordinal regression and logistic regression models were examined to determine the independent relationship between bone turnover proteins and the Fazekas scale of periventricular lesions.ResultsIn 59 hemodialysis patients, older patients had more periventricular lesions on the Fazekas scale, which was connected with a worse Mini‐Mental State Examination score (MMSE). The serum level of sclerostin, one of eight bone turnover proteins, displayed a dose‐response relationship (P value for trend = 0.004), meaning that a greater sclerostin level was connected with a lower grade on the Fazekas scale. Both the ordinal regression model (odds ratio [OR] 0.886; 95% confidence interval [CI] 0.828‐0.947, P value 0.001) and the logistic regression model (odds ratio [OR] 0.828; 95% 0.712‐0.963, P value = 0.014) determined a negative association between sclerostin and the Fazekas scale of periventricular lesions after complete adjustment for potential confounders.ConclusionsThe brain white matter damage was inversely correlated with the level of circulating sclerostin. Our finding suggested a possible connection between bone turnover proteins and brain damage.