SEARCH · PubMed Health

Results for “instrumental variable assumptions”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

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

At least 19 recordsLinked to original sources

PubMed2025

No More Free Lunch: Challenges to Mendelian Randomization Due to Sample Selection and Complex Methods.

Mendelian randomization (MR) is increasingly used in epidemiological studies to investigate causal relationships. MR depends on 3 fundamental instrumental variable assumptions: relevance, independence, and exclusion restriction. Studies often assume that MR mitigates bias from confounding due to the random allocation of genetic variants at conception. In this perspective, using causal directed acyclic graphs, we discuss several scenarios where biases in MR analyses may arise due to the nature of the data or methods being used. These include (1) collider bias due to the nonrandom selection of participants into study populations used for conducting genome-wide association studies (GWAS), (2) indirect genetic effects arising from population-based GWAS rather than within-family studies, and (3) collider bias due to gene-environment interaction effects on the exposure in nonlinear MR analyses. We provide practical considerations for examining and reducing these biases in MR analyses.

Humans↗

PubMed2006

Evaluating short-term drug effects using a physician-specific prescribing preference as an instrumental variable.

BACKGROUND: Postmarketing observational studies of the safety and effectiveness of prescription medications are critically important but fraught with methodological problems. The data sources available for such research often lack information on indications and other important confounders for the drug exposure under study. Instrumental variable methods have been proposed as a potential approach to control confounding by indication in nonexperimental studies of treatment effects; however, good instruments are hard to find. METHODS: We propose an instrument for use in pharmacoepidemiology that is based on a time-varying estimate of the prescribing physician's preference for one drug relative to a competing therapy. The use of this instrument is illustrated in a study comparing the effect of exposure to COX-2 inhibitors with nonselective, nonsteroidal antiinflammatory medications on gastrointestinal complications. RESULTS: Using conventional multivariable regression adjusting for 17 potential confounders, we found no protective effect due to COX-2 use within 120 days from the initial exposure (risk difference = -0.06 per 100 patients; 95% confidence interval = -0.26 to 0.14). However, the proposed instrumental variable method attributed a protective effect to COX-2 exposure (-1.31 per 100 patients; -2.42 to -0.20) compatible with randomized trial results (-0.65 per 100 patients; -1.08 to -0.22). CONCLUSIONS: The instrumental variable method that we have proposed appears to have substantially reduced the bias due to unobserved confounding. However, more work needs to be done to understand the sensitivity of this approach to possible violations of the instrumental variable assumptions.

Aged↗

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↗

PubMed1994

Measurement error, instrumental variables and corrections for attenuation with applications to meta-analyses.

MacMahon et al. present a meta-analysis of the effect of blood pressure on coronary heart disease, as well as new methods for estimation in measurement error models for the case when a replicate or second measurement is made of the fallible predictor. The correction for attenuation used by these authors is compared to others already existing in the literature, as well as to a new instrumental variable method. The assumptions justifying the various methods are examined and their efficiencies are studied via simulation. Compared to the methods we discuss, the method of MacMahon et al. may have bias in some circumstances because it does not take into account: (i) possible correlations among the predictors within a study; (ii) possible bias in the second measurement; or (iii) possibly differing marginal distributions of the predictors or measurement errors across studies. A unifying asymptotic theory using estimating equations is also presented.

Bias↗

PubMed2006

Instruments for causal inference: an epidemiologist's dream?

The use of instrumental variable (IV) methods is attractive because, even in the presence of unmeasured confounding, such methods may consistently estimate the average causal effect of an exposure on an outcome. However, for this consistent estimation to be achieved, several strong conditions must hold. We review the definition of an instrumental variable, describe the conditions required to obtain consistent estimates of causal effects, and explore their implications in the context of a recent application of the instrumental variables approach. We also present (1) a description of the connection between 4 causal models-counterfactuals, causal directed acyclic graphs, nonparametric structural equation models, and linear structural equation models-that have been used to describe instrumental variables methods; (2) a unified presentation of IV methods for the average causal effect in the study population through structural mean models; and (3) a discussion and new extensions of instrumental variables methods based on assumptions of monotonicity.

Bias↗

PubMed1989

The influence of different methods on basal metabolic rate measurements in human subjects.

While collating basal metabolic rate (BMR) measurements made worldwide it becomes important to know how different instruments compare with each other and whether errors in methodology could account for the differences in BMRs measured. BMRs of 34 healthy individuals were measured by using five different instruments in various combinations. Results show that energy outputs were comparable between the oxylog. Hartmann and Braun Metabolator, ventilated tent and hood, and whole-body indirect calorimeter: small differences, if any, were not statistically significant. However, significant interactions existed between subjects with the ventilated tent and hood and calorimeter when measurements were taken in subjects who were unaccustomed to the apparatus. When converting O2-consumption measurements to energy output, some instruments make assumptions that introduce a variable error in the final result, which may lead to systematic errors during the compilation of large databases of human BMRs.

Adult↗

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↗

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↗

PubMed2006

Instrumental variables: application and limitations.

To correct for confounding, the method of instrumental variables (IV) has been proposed. Its use in medical literature is still rather limited because of unfamiliarity or inapplicability. By introducing the method in a nontechnical way, we show that IV in a linear model is quite easy to understand and easy to apply once an appropriate instrumental variable has been identified. We also point out some limitations of the IV estimator when the instrumental variable is only weakly correlated with the exposure. The IV estimator will be imprecise (large standard error), biased when sample size is small, and biased in large samples when one of the assumptions is only slightly violated. For these reasons, it is advised to use an IV that is strongly correlated with exposure. However, we further show that under the assumptions required for the validity of the method, this correlation between IV and exposure is limited. Its maximum is low when confounding is strong, such as in case of confounding by indication. Finally, we show that in a study in which strong confounding is to be expected and an IV has been used that is moderately or strongly related to exposure, it is likely that the assumptions of IV are violated, resulting in a biased effect estimate. We conclude that instrumental variables can be useful in case of moderate confounding but are less useful when strong confounding exists, because strong instruments cannot be found and assumptions will be easily violated.

Bias↗

PubMed2000

A new bilateral closed chain assessment technique: methods and error analysis.

PURPOSE: The purpose of this paper is to present the Omnikinetic methodology for clinical evaluation of lower extremity function, to characterize its sensitivity to errors, and to present typical data for an assessment protocol. METHODS: A 5-bar, 2-degree of freedom linkage was used to model the geometry of the crank, pedal, and lower extremity. Two-degree force transducers at the pedal were used to calculate center of pressure and force applied at the foot. A Newton-Euler inverse dynamic model was used to calculate net joint torques and powers bilaterally of the ankle, knee, and hip. Ten subjects performed a high velocity evaluation protocol which served as the control. Error sensitivity was determined by adding instrumental error, hip translation, and segmental length errors to the collected data and comparing the outcome to the control. RESULTS: All variables associated with instrumental error had mean errors under 4%. Mean errors associated with violations of the fixed hip assumption were under 15% for all variables. Mean errors associated with anthropometric measures were divided into two types: relative error (overall length unchanged, ratios of segments changed) and absolute (overall length changed, ratios of segment lengths unchanged). Relative anthropometric mean errors were under 5%. Absolute anthropometric mean errors were under 12%. CONCLUSION: The Omnikinetic is a new tool for bilateral lower extremity evaluation that enables the whole lower extremity to be evaluated at the joint level. Instrumental accuracy was excellent. The instrument was most sensitive to violations of the fixed hip position assumption over the last 20 degrees of knee extension.

Adult↗

PubMed2005

Random marginal agreement coefficients: rethinking the adjustment for chance when measuring agreement.

Agreement coefficients quantify how well a set of instruments agree in measuring some response on a population of interest. Many standard agreement coefficients (e.g. kappa for nominal, weighted kappa for ordinal, and the concordance correlation coefficient (CCC) for continuous responses) may indicate increasing agreement as the marginal distributions of the two instruments become more different even as the true cost of disagreement stays the same or increases. This problem has been described for the kappa coefficients; here we describe it for the CCC. We propose a solution for all types of responses in the form of random marginal agreement coefficients (RMACs), which use a different adjustment for chance than the standard agreement coefficients. Standard agreement coefficients model chance agreement using expected agreement between two independent random variables each distributed according to the marginal distribution of one of the instruments. RMACs adjust for chance by modeling two independent readings both from the mixture distribution that averages the two marginal distributions. In other words, both independent readings represent first a random choice of instrument, then a random draw from the marginal distribution of the chosen instrument. The advantage of the resulting RMAC is that differences between the two marginal distributions will not induce greater apparent agreement. As with the standard agreement coefficients, the RMACs do not require any assumptions about the bivariate distribution of the random variables associated with the two instruments. We describe the RMAC for nominal, ordinal and continuous data, and show through the delta method how to approximate the variances of some important special cases.

Data Interpretation, Statistical↗

PubMed2004

Instrumental variables and inverse probability weighting for causal inference from longitudinal observational studies.

Inferring causal effects from longitudinal repeated measures data has high relevance to a number of areas of research, including economics, social sciences and epidemiology. In observational studies in particular, the treatment receipt mechanism is typically not under the control of the investigator; it can depend on various factors, including the outcome of interest. This results in differential selection into treatment levels, and can lead to selection bias when standard routines such as least squares regression are used to estimate causal effects. Interestingly, both the characterization of and methodology for handling selection bias can differ substantially by disciplinary tradition. In social sciences and economics, instrumental variables (IV) is the standard method for estimating linear and nonlinear models in which the error term may be correlated with an observed covariate. When such correlation is not ruled out, the covariate is called endogenous and least squares estimates of the covariate effect are typically biased. The availability of an instrumental variable can be used to reduce or eliminate the bias. In public health and clinical medicine (e.g., epidemiology and biostatistics), selection bias is typically viewed in terms of confounders, and the prevailing methods are geared toward making proper adjustments via explicit use of observed confounders (e.g., stratification, standardization). A class of methods known as inverse probability weighting (IPW) estimators, which relies on modeling selection in terms of confounders, is gaining in popularity for making such adjustments. Our objective is to review and compare IPW and IV for estimating causal treatment effects from longitudinal data, where the treatment may vary with time. We accomplish this by defining the causal estimands in terms of a linear stochastic model of potential outcomes (counterfactuals). Our comparison includes a review of terminology typically used in discussions of causal inference (e.g., confounding, endogeneity); a review of assumptions required to identify causal effects and their implications for estimation and interpretation; description of estimation via inverse weighting and instrumental variables; and a comparative analysis of data from a longitudinal cohort study of HIV-infected women. In our discussion of assumptions and estimation routines, we try to emphasize sufficient conditions needed to implement relatively standard analyses that can essentially be formulated as regression models. In that sense this review is geared toward the quantitative practitioner. The objective of the data analysis is to estimate the causal (therapeutic) effect of receiving combination antiviral therapy on longitudinal CD4 cell counts, where receipt of therapy varies with time and depends on CD4 count and other covariates. Assumptions are reviewed in context, and resulting inferences are compared. The analysis illustrates the importance of considering the existence of unmeasured confounding and of checking for 'weak instruments.' It also suggests that IV methodology may have a role in longitudinal cohort studies where potential instrumental variables are available.

Antiretroviral Therapy, Highly Active↗

PubMed1994

The reliability of manufacturer-determined, instrument-specific international sensitivity index values for calculating the international normalized ratio.

Some thromboplastin manufacturers are currently supplying the instrument-specific international sensitivity index (ISI) values of their reagents, allowing clinical laboratories to calculate instrument-specific international normalized ratio (INR) values on plasma samples from patients receiving coumarin therapy. However, the assumption that systematic interinstrument variability in the INR would be eliminated if manufacturer-determined ISI values were used remains unsubstantiated. This assumption was evaluated by comparing INR values obtained on one instrument that measures a mechanical endpoint (fibrometer) with one that measures a photo-optical endpoint (MLA-700). Three thromboplastin reagents with instrument-specific ISI values supplied by the manufacturer (ISI range, 1.23-2.79) were used. For two of three reagents, the fibrometer INR values were significantly higher than the MLA-700 INR values (P < .01). Analysis of log prothrombin time ratio plots showed that this systematic variability was caused by inaccurate manufacturer ISI values. Of clinical significance is that the inaccurate ISI values produced a high number of discordant INR values between these two instruments (> 47% of plasma samples had one INR value within and one out of the recommended therapeutic range). The implication of these findings for laboratory monitoring of oral anticoagulation is discussed.

Adult↗

PubMed1990

Partitioning macronutrient intake estimates from a food frequency questionnaire.

In the 1984-1985 University of Michigan Food Frequency Study, 228 black and white men and women ages 24-51 years completed 16 days of food records over the course of a year and then completed a quantitative food frequency questionnaire. The relative validity of macronutrient (fat, protein, and carbohydrate) estimates from the questionnaire was evaluated with the records used as a standard. The two methods differed both in estimated group mean intake and in the relative ranking of individuals. Partitioning of between-method disagreement showed that frequency differences were the main source of poor agreement in relative ranking; serving size differences contributed mainly to disagreement in group mean intake; and differences in nutrient composition had only minor effects. After adjusting for total calorie intake, agreement in group mean intake improved. Agreement in relative ranking and classification also improved, but substantial disagreement remained. The strategy of adjusting for total calorie intake depends on the assumption that macronutrients are reported in similar proportions on both instruments. This assumption did not appear to hold true. Respondents showed considerable variability in the proportionality of reporting, with 46% of respondents reporting fat intake disproportionately. A summary index of proportionality is described and the relation between disproportionate reporting and misclassification is discussed. Disproportionate reporting of macronutrients limits the ability of food frequency questionnaires to rank individuals similarly to a reference method even if estimates are adjusted for total calorie intake.

Adult↗

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↗

PubMed2005

Dietary intake assessments in epidemiology: can we know what we are measuring?

PURPOSE: We discuss fundamental assumptions required for the validation of dietary questionnaire assessments, using latent variable models. METHODS: We discuss some methodological aspects of latent variable modelling in terms of the triad method, for comparisons between questionnaire assessments, food consumption records and/or biomarkers. In particular, we address the recent proposal (Frazer et al) that biases due to correlated random errors between different measurement types could be overcome by using two biomarkers as instrumental variables, as recently proposed by. RESULTS: The instrumental variable approach proposed by Frazer et al, but also other validation models proposed so far, require the assumption that questionnaire measurements are specific for a defined type of food or nutrient intake (latent variable of interest) - that is, conditionally on the true intake levels of this food or nutrient the questionnaire measurements should have no association with intakes of any other dietary component. CONCLUSION: More methodological research is needed on the design of multivariate validation studies that might allow the examination of measurement specificity in practice.

Data Collection↗

PubMed2004

Methods to assess intended effects of drug treatment in observational studies are reviewed.

BACKGROUND AND OBJECTIVE: To review methods that seek to adjust for confounding in observational studies when assessing intended drug effects. METHODS: We reviewed the statistical, economical and medical literature on the development, comparison and use of methods adjusting for confounding. RESULTS: In addition to standard statistical techniques of (logistic) regression and Cox proportional hazards regression, alternative methods have been proposed to adjust for confounding in observational studies. A first group of methods focus on the main problem of nonrandomization by balancing treatment groups on observed covariates: selection, matching, stratification, multivariate confounder score, and propensity score methods, of which the latter can be combined with stratification or various matching methods. Another group of methods look for variables to be used like randomization in order to adjust also for unobserved covariates: instrumental variable methods, two-stage least squares, and grouped-treatment approach. Identifying these variables is difficult, however, and assumptions are strong. Sensitivity analyses are useful tools in assessing the robustness and plausibility of the estimated treatment effects to variations in assumptions about unmeasured confounders. CONCLUSION: In most studies regression-like techniques are routinely used for adjustment for confounding, although alternative methods are available. More complete empirical evaluations comparing these methods in different situations are needed.

Confounding Factors, Epidemiologic↗

PubMed1999

A method for proficiency testing of small peer groups in the College of American Pathologists Coagulation Surveys.

OBJECTIVE: To develop a grading scheme for the proficiency testing of small peer groups of fewer than 10 members for the prothrombin time (PT) and activated partial thromboplastin time (APTT). METHODS: A modified target value for small peer groups was derived based on the assumption that measurement variability in the PT and APTT is more greatly influenced by variations in reagents than in instruments. Criteria for grading were established by statistical simulation to achieve misclassification errors of less than 5% for both incorrectly passing and failing participants. College of American Pathologists Coagulation Survey data were analyzed to determine the number of additional laboratories graded using the proposed scheme, as well as the failure rates among participants in the small peer groups. RESULTS: The modified target value for small peer groups is a weighted average between the mean of the peer group and the mean of all participants using the same reagent (reagent group). Peer groups with as few as 4 members can be graded provided that specific criteria are satisfied: there must be at least 5 peer groups for the same reagent, at least 3 of these 5 peer groups must have more than 3 members, and the coefficient of variation for the reagent group must be less than 10%. This proposed grading scheme decreased the number of ungraded laboratories by 44% to 46% for the PT and 42% to 55% for the APTT. The percentage of failing grades among participants in the small peer groups ranged from 1.3% to 4.1% for the PT and 1.4% to 7.2% for the APTT. These failure rates were 2.8- to 13.0-fold higher than the failure rates in large peer groups (P < or = .05). CONCLUSIONS: The proposed small peer group grading scheme can improve the effectiveness of College of American Pathologists proficiency testing for the PT and APTT and may also be generally applicable to other test methods and analytes.

Anticoagulants↗