SEARCH · Search PubMed

Results for “instrumental variable”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 91 records · Page 5Linked to original sources

PubMed2002

Addressing self-selection bias in quasi-experimental evaluations of whole-school reform. A comparison of methods.

This article discusses potential sources of self-selection bias in quasi-experimental evaluations of whole-school reform models and considers how individual student-level data might be used to provide valid impact estimates. Although repeated pretreatment and posttreatment measures of student performance can provide unbiased estimates under relatively weak assumptions, such data are difficult to obtain. The article develops an instrumental variable strategy that can be used to improve on common value-added estimators when only posttreatment measures of performance are available. Using data from New York City, the author shows that the instrumental variable strategy can provide estimates of model impacts similar to those provided by a difference-in-differences estimator provided that appropriate instruments are used.

Bias↗

PubMed2026

Modelling time-varying genetic effects on binary disease risk via functional Mendelian randomization.

MOTIVATION: Genome-wide association studies have identified thousands of genetic variants associated with complex traits, establishing Mendelian randomization (MR) as a powerful framework for causal inference using variants as natural experiments. However, existing MR methods treat causal effects as static, relying on cross-sectional exposure measurements and ignoring how genetic predispositions to disease operate dynamically across the life course. Recovering age-specific causal effect functions from longitudinal data requires combining functional data representations of exposure trajectories with instrumental variable estimation strategies suitable for binary disease endpoints, a methodological gap that has remained unaddressed. RESULTS: We develop a functional MR framework for binary outcomes that integrates functional principal component analysis with two-stage residual inclusion (2SRI), ensuring consistent estimation under the nonlinear logistic link function that renders standard instrumental variable estimators inconsistent. Simulations across different causal effect trajectory shapes, varying measurement densities, and varying instrument strengths demonstrate accurate recovery of time-varying genetically predicted effects with minimal bias. Applied to UK Biobank data, the framework identifies an age-specific causal effect of genetically predicted body mass index on type 2 diabetes risk concentrated in early mid-adulthood and progressively attenuating thereafter. Concordance between the proposed 2SRI estimator applied to type 2 diabetes and the established continuous-outcome functional MR estimator applied to the paired glycated haemoglobin marker in the same cohort provides indirect empirical support for the validity of the proposed approach. AVAILABILITY AND IMPLEMENTATION: The method is implemented in the R package mvfmr, with a full tutorial vignette.

Mendelian Randomization Analysis↗

PubMed2026

Novel Insights into Immune Cell Function in Type 2 Diabetes Mediated by Gut Microbiota: A Two-Sample Mendelian Randomization Study.

INTRODUCTION: The role of immune cells in type 2 diabetes mellitus (T2DM) development is well-studied, but their interactions with the gut microbiota and the mediating role in this process remain unclear. METHODS: We analyzed 731 immune cell phenotypes (3,757 Europeans), 473 gut microbiota traits (5,959 Finns), and T2DM data (over 400,000 Finns). Mendelian randomization (MR) was based on three assumptions: the instrumental variable (IV) is associated with exposure, IV is not influenced by confounding, and IV affects the outcome only through exposure. We selected single-nucleotide polymorphisms (SNPs) from genome-wide association studies as instrumental variables (IVs) to infer causal effects in two-sample MR analysis. RESULTS: We identified 36 immune cell phenotypes associated with T2DM, including 29 protective factors and seven risk factors, as well as 10 gut microbiota significantly linked to T2DM, with eight protective factors and two risk factors. MR revealed that five gut microbiota mediated the relationship between immune cells and T2DM. For example, the effects of CD3 on resting Treg (OR: 1.0136), CD3 on CM CD4+ (OR: 1.0180), and CD3 on naive CD4+ cells (OR: 1.0150) in T2DM were found to be partially mediated by the species Bacillus. AYThe corresponding mediation effect proportions were 8.99%, 11.8%, and 11.4%. DISCUSSION: MR analysis identified multiple gut microbiota mediators in the relationship between immune cells and T2DM, addressing previous observational evidence. Limitations included the European ancestry bias, among others. CONCLUSION: This study has highlighted the gut microbiota as a mediator between immune cells and T2DM, offering new insights for its early prevention and intervention.

Diabetes Mellitus, Type 2↗

PubMed2003

Exploring the health-wealth nexus.

The causal links between health and economic resources have long concerned social scientists. We use four waves of data from the Panel Study of Income Dynamics (PSID) to analyze the impact of wealth upon an individual's health status. The difficulty in approaching this task that has bedeviled previous studies is that wealth may be endogenous; a priori, it is just as likely that changes in health affect wealth as vice versa. We argue that inheritance is a suitable instrument for the change in wealth, and implement a straightforward instrumental variables strategy to deal with this problem. Our results suggest that the causal relationship running from wealth to health may not be as strong as first appears. In the data, wealth exerts a positive and statistically significant effect on health status, but it is very small in magnitude. Instrumental variables estimation leaves the point estimate approximately the same, but renders it insignificantly different from zero. And even when the point estimate is increased by twice its standard error (S.E.), the quantitative effect is small. We conclude that the wealth-health connection is not driven by short run changes in wealth.

Causality↗

PubMed2004

The effect of practice budgets on patient waiting times: allowing for selection bias.

Under the UK fundholding scheme, general practices could elect to hold a budget to meet the costs of some types of elective surgery (chargeable admissions) for their patients. It was alleged that patients of fundholding practices had shorter waits for elective surgery than the patients of non-fundholders. Comparison of waiting times between fundholding and non-fundholding practices are potentially confounded by selection bias as fundholding was voluntary. We estimate the effect of a practice's fundholding status on the waiting times of its patients using both cross-sectional methods (OLS, propensity score, instrumental variables, Heckman selection correction and Heckman heterogenous effects estimators) and difference in differences methodologies to correct for selection bias. The estimated effect of fundholding status was to significantly reduce the waiting times for chargeable admissions of the patients of fundholders by 4.1-6.6% (or 4-7 days) with the instrumental variables and Heckman selection correction estimators yielding the highest estimates. We also find that patients of fundholding practices had shorter waits (by 3.7% or 2 days) for non-chargeable elective admissions, suggesting that fundholders were able to obtain shorter waits for all types of elective admissions.

Budgets↗

PubMed2004

Do routine eye exams improve vision?

We use a longitudinal national sample of Medicare claims linked to the National Long-Term Care Survey (NLTCS) to assess the productivity of routine eye examinations. Although such exams are widely recommended by professional organizations for certain populations, there is limited empirical evidence on the productivity of such care. We measure two outcomes, the ability to continue reading, and no onset of blindness or low vision, accounting for potential endogeneity of frequency of eye exams. Using instrumental variables, we find a statistically significant and beneficial effect of routine eye exams for both outcomes. Marginal effects for reading ability are large, but decline in the number of years with eye exams. Effects for blindness/low vision are smaller for the general elderly population, but larger for persons with diabetes. Instrumental variables provide a useful approach for assessing the productivity of particular interventions, particularly in situations in which randomized controlled trials are expensive or perhaps unethical and difficult to conduct over a lengthy time period.

Cost-Benefit Analysis↗

PubMed2004

The 2004 Marshall Urist award: delays until surgery after hip fracture increases mortality.

The objective of this study was to analyze whether a delay in time from admission until surgical treatment increased the mortality rate for patients with a closed hip fracture. We used the day of the week of admission as an instrumental variable to pseudorandomize patients. We analyzed 18,209 Medicare recipients who were 65 years of age or older and had surgical treatment for a closed hip fracture. Patients for whom the delay between admission and surgery was 2 days or more had a 17% higher chance of dying by Day 30. Using instrumental variables analysis, we found a similar 15% increased risk of mortality in patients with delays until surgery of 2 or more days. Based on these results, we found that a delay of 2 or more days significantly increased the mortality rate. This suggests that delay to surgery independently affects mortality, therefore additional study on the effect of smaller delays on outcome is needed.

Aged↗

PubMed2005

Biofeedback treatment of fecal incontinence incorporating a mental variable without instrumentation: a prospective pilot study in Hispanic population.

UNLABELLED: The long-established approaches utilized to treat fecal incontinence always require instrumentation with some type of electronic equipment. This equipment is not always available in every institutions. In addition, no studied protocol principally used as coordination, sensory, or strength training has reached the level of gold standard. The purpose of this study was to describe a simple biofeedback technique that incorporating a mental variable and not requiring electronic equipment with prior adequate training could be used at any medical institution. METHODS: A particular modality of an operant conditioning technique was given once and a home trainer program was established. Forty-eight patients (mean age 37.1 +/- 3.7 years) were recruited. Patients had suffered from total incontinence for a period of 55 +/- 7.5 months, all used two to three pads per day and suffered 2.4 +/- 0.2 episodes of incontinence per day. Patients underwent clinical history recording, laboratory tests, recto-sigmoidoscopy, and double-contrast barium enema. Manometry and rectal sensitivity were performed in 7 and 27 patients, respectively. For physiologic comparisons, 21 healthy volunteers were used. RESULTS: A total of 79.1% of patients became continent in a median period of 3.9 +/- 0.5 months. An average of 3.85 +/- 0.55 sessions was required. Follow-up continued for 3-11 years. Patients with incontinence showed lower basal mean resting pressure, maximum squeeze pressure and rectal sensitivity (p <0. 01) and spontaneous rectoanal inhibitory reflex was absent in 57%. CONCLUSIONS: This biofeedback approach does not employ any type of electronic equipment and can be easily reproduced in any type of medical center. Additionally, this is the first report in which a methodology for biofeedback therapy successfully incorporates a mental variable in addition to sensory and strength training.

Adult↗

PubMed2025

Association between NAFLD and liver cancer: A two-sample Mendelian randomization study.

Observational studies suggest an association between nonalcoholic fatty liver disease (NAFLD) and liver cancer, but its causal nature remains unclear. A 2-sample Mendelian randomization (MR) analysis was performed using NAFLD and liver cancer summary statistics from genome-wide association study databases. Instrumental variables satisfying the 3 core MR assumptions were selected. Causal effects were estimated using inverse-variance weighted, MR-Egger, weighted median, and other methods, followed by sensitivity and power analyses. All 4 MR analyses demonstrated a positive causal association between NAFLD and liver cancer risk [odds ratio > 1, inverse-variance weighted P < .001]. Sensitivity analysis indicated no significant level of multiplicity or heterogeneity in the instrumental variables, and individual single nucleotide polymorphisms had no significant impact on the results. However, statistical power was insufficient. This study provides the first MR evidence demonstrating a genetically predicted causal relationship between NAFLD and liver cancer that is consistent across subtypes. Sensitivity analyses confirmed the absence of horizontal pleiotropy or heterogeneity, strengthening the robustness of the findings. These results offer genetic support for early NAFLD intervention to reduce the risk of liver cancer. However, the limited statistical power highlights the need for larger-scale genome-wide association study to identify more and stronger genetic instruments for a more precise quantification of the causal effect of NAFLD on liver cancer risk.

Humans↗

PubMed2004

Determinants of self-reported mental health using the British household panel survey.

BACKGROUND: The study of self-reported mental health is a fairly recent area for economists, although sociologists, psychologists and public health specialists have been studying it for years. One methodological problem with earlier research is that there are many unobserved characteristics of individuals that may be correlated with self-reported mental health. Neglecting these factors may lead to biased estimates of the effects of variables such as income, education, health, etc. Panel data enables us to control for unobserved individual specific effects, whereas a cross-section study or time series study cannot. AIMS OF THE STUDY: This paper examines the determinants of self-reported mental health in UK using data from the first eight waves of the British Household Panel Survey. In particular, we are interested in assessing the effect of education on self-reported mental health which other studies have ignored. METHODS: The measure of self-reported mental health used in this paper is the General Health Questionnaire (GHQ). To account for the possible correlation between the unobserved individual effects and some explanatory variables, a Hausman Taylor's instrumental variables estimator (HT) is employed. In order to derive this estimator, one has to distinguish between variables that are correlated with the individual specific effects (endogenous) and variables which are uncorrelated with the individual specific effects (exogenous). This HT estimator also allows for estimating the parameters corresponding with time invariant variables such as education and ethnicity. RESULTS: The evidence presented here confirms that mental health scores mentioned on the GHQ are significantly related to job status, age, marital status and self-assessed health status. The results also show no evidence that income impacts on self-reported mental health. Ethnicity is also found to deteriorate self-reported mental health yet the effect is not significant. The results of this paper also show that education had no significant impact on self-reported mental health. IMPLICATIONS FOR MENTAL HEALTH POLICY: Issues related to unemployment and social cohesion may be relevant factors in the prevention of mental illness. Policies aimed at improving these factors have an impact on the mental health status of society. In consideration of the evidence of gender differential in mental health, mental health policies should take into account properly this issue. IMPLICATIONS FOR FURTHER RESEARCH: In order to draw definite conclusions, it is important to formally test the presence of attrition bias as well as expand the sample to include more waves. Still, we are concerned about the issue of weak correlation between the instruments and potential endogenous variables. Additionally, we have to bear in mind that inconsistent estimates may potentially occur if the partition of the variables in subsets of endogenous and exogenous is not correctly specified. These issues need further research. The estimation technique also presented in this paper may be applied to a wide range of health services research.

Family↗

PubMed2004

Estimation of antipsychotic effects on hospitalization risk in a naturalistic study with selection on unobservables.

Estimates of effects of antipsychotic medication on hospitalization risk based on nonexperimental data may be affected by selection bias from either observable or unobservable factors. This study applies a statistical method, using instrumental variables, that controls for both types of possible selection bias. We use data from a large observational study of people under treatment for schizophrenia to estimate models of drug choice and hospitalization, including atypical (versus typical) medication effects on 12-month hospitalization risk. Results for younger patients (<age 45 years) indicate that unobservable factors bias the atypical effect estimate in a positive direction; correcting for this bias yields a significant negative effect on hospitalization risk. With data for older patients, our instrumental variables performed poorly and provided little information about possible selection bias. Obtaining detailed information on treatment history and other determinants of medication choice in future studies is critical for deriving more accurate estimates of medication effects from nonexperimental data.

Adult↗

PubMed2001

The relationship between quality and outcomes in routine depression care.

OBJECTIVE: This longitudinal, nonexperimental study examined whether depression treatment provided in concordance with guidelines developed by the Agency for Healthcare Research and Quality (AHRQ) is associated with improved clinical outcomes. METHODS: The medical, insurance, and pharmacy records of a community-based sample of 435 subjects who screened positive for current major depression were abstracted to ascertain whether depression treatment was received and whether it was provided in accordance with AHRQ guidelines. Regression analyses estimated the impact of guideline-concordant treatment on the change in depression severity and on mental and physical health over a six-month period. An instrumental variables analysis was used to check the sensitivity of the results to selection bias. RESULTS: A total of 106 subjects were treated for depression by 105 different primary care and specialty providers. Sixty percent of the sample had current major depression, and about 40 percent had subthreshold depression. Only 29 percent of the patients received guideline-concordant treatment. For patients with major depression, guideline-concordant care was significantly and substantially associated with improved depression severity but not with improvements in overall mental or physical health. The instrumental variables analysis indicated that the standard regression analysis underestimated the treatment effect by 21 percent. For those with subthreshold depression, guideline-concordant care was not associated with improved outcomes. DISCUSSION AND CONCLUSIONS: This community-based, nonexperimental study found a positive relationship between the quality of care for depression and clinical outcomes for patients with major depression in routine practice settings.

Adult↗

PubMed2001

Alternative solutions to the problem of selection bias in an analysis of federal residential drug treatment programs.

In an evaluation of prison-based residential drug treatment programs, the authors use three different regression-based approaches to estimating treatment effects. Two of the approaches, the instrumental variable and the Heckman approach, attempt to minimize selection bias as an explanation for treatment outcomes. Estimates from these approaches are compared with estimates from a regression in which treatment is represented by a dummy variable. The article discusses the advantage of using more than one method to increase confidence in findings when possible selection bias is a concern. Three-year outcome data for 2,315 federal inmates are used in analyses where the authors separately examine criminal recidivism and relapse to drug use for men and women. Statistical tests lead the authors to conclude that treatment reduces criminal recidivism and relapse to drug use. The treatment effect was largest when the inference was based on the Heckman approach, somewhat smaller when based on the instrumental variable approach, and smallest when based on the traditional dummy variable approach. Treatment effects for females were not statistically significant.

Crime↗

PubMed2024

Assessing the causal link between liver function and acute pancreatitis: A Mendelian randomisation study.

A correlation has been reported to exist between exposure factors (e.g. liver function) and acute pancreatitis. However, the specific causal relationship remains unclear. This study aimed to infer the causal relationship between liver function and acute pancreatitis using the Mendelian randomisation method. We employed summary data from a genome-wide association study involving individuals of European ancestry from the UK Biobank and FinnGen. Single-nucleotide polymorphisms (SCNPs), closely associated with liver function, served as instrumental variables. We used five regression models for causality assessment: MR-Egger regression, the random-effect inverse variance weighting method (IVW), the weighted median method (WME), the weighted model, and the simple model. We assessed the heterogeneity of the SNPs using Cochran's Q test. Multi-effect analysis was performed using the intercept term of the MR-Egger method and leave-one-out detection. Odds ratios (ORs) were used to evaluate the causal relationship between liver function and acute pancreatitis risk. A total of 641 SNPs were incorporated as instrumental variables. The MR-IVW method indicated a causal effect of gamma-glutamyltransferase (GGT) on acute pancreatitis (OR = 1.180, 95%CI [confidence interval]: 1.021-1.365, P = 0.025), suggesting that GGT may influence the incidence of acute pancreatitis. Conversely, the results for alkaline phosphatase (ALP) (OR = 0.997, 95%CI: 0.992-1.002, P = 0.197) and aspartate aminotransferase (AST) (OR = 0.939, 95%CI: 0.794-1.111, P = 0.464) did not show a causal effect on acute pancreatitis. Additionally, neither the intercept term nor the zero difference in the MR-Egger regression attained statistical significance (P = 0.257), and there were no observable gene effects. This study suggests that GGT levels are a potential risk factor for acute pancreatitis and may increase the associated risk. In contrast, ALP and AST levels did not affect the risk of acute pancreatitis.

Humans↗

PubMed2004

Analyzing a randomized trial on breast self-examination with noncompliance and missing outcomes.

Recently, instrumental variables methods have been used to address non-compliance in randomized experiments. Complicating such analyses is often the presence of missing data. The standard model for missing data, missing at random (MAR), has some unattractive features in this context. In this paper we compare MAR-based estimates of the complier average causal effect (CACE) with an estimator based on an alternative, nonignorable model for the missing data process, developed by Frangakis and Rubin (1999, Biometrika, 86, 365-379). We also introduce a new missing data model that, like the Frangakis-Rubin model, is specially suited for models with instrumental variables, but makes different substantive assumptions. We analyze these issues in the context of a randomized trial of breast self-examination (BSE). In the study two methods of teaching BSE, consisting of either mailed information about BSE (the standard treatment) or the attendance of a course involving theoretical and practical sessions (the new treatment), were compared with the aim of assessing whether teaching programs could increase BSE practice and improve examination skills. The study was affected by the two sources of bias mentioned above: only 55% of women assigned to receive the new treatment complied with their assignment and 35% of the women did not respond to the post-test questionnaire. Comparing the causal estimand of the new treatment using the MAR, Frangakis-Rubin, and our new approach, the results suggest that for these data the MAR assumption appears least plausible, and that the new model appears most plausible among the three choices.

Adult↗

PubMed2000

Is more better than less? An analysis of children's mental health services.

OBJECTIVE: To assess the dose-response relationship for outpatient therapy received by children and adolescents-that is, to determine the impact of added outpatient visits on key mental health outcomes (functioning and symptomatology). DATA SOURCES/STUDY SETTING: The results presented involve analyses of data from the Fort Bragg Demonstration and are based on a sample of 301 individuals using outpatient services. STUDY DESIGN: This article provides estimates of the impact of outpatient therapy based on comparisons of individuals receiving differing treatment doses. Those comparisons involve standard multiple regression analyses as well as instrumental variables estimation. The latter provides a means of adjusting comparisons for unobserved or unmeasured differences among individuals receiving differing doses, differences that would otherwise be confounded with the impact of treatment dose. DATA COLLECTION/EXTRACTION METHODS: Using structured diagnostic interviews and behavior checklists completed by the child and his or her caretaker, detailed data on psychopathology, symptomatology, and psychosocial functioning were collected on individuals included in these analyses. Information on the use of mental health services was taken from insurance claims and a management information system. Services data were used to describe the use of outpatient therapy within the year following entry into the study. PRINCIPAL FINDINGS/CONCLUSIONS: Instrumental variables estimation indicates that added outpatient therapy improves functioning among children and adolescents. The effect is statistically significant and of moderate practical magnitude. These results imply that conventional analyses of the dose-response relationship may understate the impact of additional treatment on functioning. This finding is robust to choice of functional form, length of time over which outcomes are measured, and model specification. Dose does not appear to influence symptomatology.

Adolescent↗

PubMed2004

Use of claims data to examine the impact of length of inpatient psychiatric stay on readmission rate.

OBJECTIVE: This study analyzed the impact of length of stay for inpatient treatment of psychiatric disorders on readmission rates. METHODS: Hospitalization data were obtained from the MarketScan data set collected by Medstat. The instrumental variable method, an econometric technique, was used to estimate the impact of length of stay on the rate of readmission for 5,735 persons who had at least one discharge with a primary diagnosis of a psychiatric disorder during 1997 and 1998. RESULTS: Decreasing length of stay below ten days led to an increase in the readmission rate during the 30 days after discharge. Decreasing the length of stay from seven to six days increased the expected readmission rate from.04 to.047 (17.5 percent), whereas decreasing length of stay from four to three days increased the readmission rate from.09 to.136 (51.1 percent). CONCLUSION: Decreasing length of stay for inpatient psychiatric treatment increased the readmission rate. The use of instrumental variables could help better estimate the value of mental health services when using observational data.

Adult↗

PubMed2001

The program effect of batterer programs in three cities.

Recent meta-analyses and experimental designs of batterer program evaluations suggest little or no program effect. This finding may be compromised by a variety of analytical issues. Instrumental variable analysis addresses some of these issues, especially the relationship of program dropout to batterer reassault. This method of analysis was, therefore, used to test for program effect in a multi-site evaluation. The sites were three well-established batterer programs using a gender-based, cognitive-behavioral approach (n = 640). Completing a batterer program reduced the likelihood of reassault by 44% to 64%, depending on the specification used. Completing a 3-month program appeared to be as effective as completing a 5 1/2- or 9-month program. This moderate effect is an accomplishment considering the problems associated with the program participants and the inconsistency in the criminal justice system at the research sites. The findings remain tentative because of weak instrumental variables for reassault, but do confirm the need for more complex analyses of program effect.

Colorado↗