Causal Evaluation of Laboratory Markers in Type 2 Diabetes on Cancer and Vascular Diseases Using Various Mendelian Randomization Tools

Multiple studies have demonstrated the effects of type 2 diabetes (T2D) on various human diseases; however, most of these were observational epidemiological studies that suffered from many potential biases including reported confounding and reverse causations. In this article, we investigated whether cancer and vascular disease can be affected by T2D-related traits, including fasting plasma glucose (FPG), 2-h postprandial glucose (2h-PG), and glycated hemoglobin A1c (HbA1c) levels, by using Mendelian randomization (MR). The summary statistics for FPG, 2h-PG, and HbA1c level were obtained through meta-analyses of large-scale genome-wide association studies that included data from 133,010 nondiabetic individuals from collaborating Meta-analysis of Glucose and Insulin Related Traits Consortium studies. Thereafter, based on the statistical assumptions for MR analyses, the most reliable approaches including inverse-variance-weighted (IVW), MR-Egger, MR-Egger with a simulation extrapolation (SIMEX), weighted median, and MR-pleiotropy residual sum and outlier (MR-PRESSO) methods were applied to identify traits affected by FPG, 2h-PG, and HbAlc. We found that coronary artery disease is affected by FPG, as per the IVW [log odds ratio (logOR): 0.21; P = 0.012], MR-Egger (SIMEX) (logOR: 0.22; P = 0.014), MR-PRESSO (logOR: 0.18; P = 0.045), and weighted median (logOR: 0.29; P < 0.001) methods but not as per the MR-Egger (logOR: 0.13; P = 0.426) approach. Furthermore, low-density lipoprotein cholesterol levels are affected by HbA1c, as per the IVW [beta (B): 0.23; P = 0.015), MR-Egger (B: 0.45; P = 0.046), MR-Egger (SIMEX) (B: 0.27; P = 0.007), MR-PRESSO (B; 0.14; P = 0.010), and the weighted median (B: 0.15; P = 0.012] methods. Further studies of the associated biological mechanisms are required to validate and understand the disease-specific differences identified in the TD2-related causal effects of each trait.

Omitted variable bias in studies of plant interactions

Models of plant-plant interactions underpin our understanding of species coexistence, invasive plant impacts, and plant community responses to climate change. In recent studies, models of competitive interactions failed predictive tests, thereby casting doubt on results of many past studies. We believe these model failures owe at least partly to heterogeneity in unmodeled factors (e.g., nutrients, soil pathogens) that affect both target plants and neighboring competitors. Such heterogeneity is ubiquitous, and models that do not account for it will suffer omitted variable bias. We used instrumental variables analysis to test for and correct omitted variable bias in studies that followed common protocols for measuring plant competition. In an observational study, omitted variables caused competition to seem like mutualism. In a quasi-experiment that partially controlled competitor abundances with seeding, omitted variables caused competition to seem about 35% weaker than it really was, even though the experiment occurred in an abandoned agricultural field where environmental heterogeneity was expected to be relatively low. Despite decades of research, consistently accurate estimates of competitive interactions remain elusive. The most foolproof way around this problem is true experiments that avoid omitted variable bias by completely controlling competitor abundances, but such experiments are rare.

The Effect of Combination Antiretroviral Therapy Use Among HIV Positive Children on the Hazard of AIDS Using Calendar Year as an Instrumental Variable

BACKGROUND

Instrumental variable (IV) analyses are a common causal inference technique used in the absence of randomized data. Combination Antiretroviral Therapy (cART) was first introduced in 1996 and calendar periods have been used as a proxy for cART use. However, cART use misclassification can bias IV analyses.

OBJECTIVE

We aim to highlight the differences in the effects of antiretroviral therapy on clinical outcomes between the applications of traditional and adapted IV analysis techniques.

METHODS

This study includes children with perinatal human immunodeficiency virus (HIV-1) infection followed from 1988 to 2009. We describe an application of traditional and adapted IV analysis techniques. Noncompliance adjustments were applied to correct the misclassification of cART-use. Weighting the inverse probability of calendar era, the selected covariates were performed to control for variables that may be related to both the IV and outcome.

RESULTS

During 48,380 person-days, 78 HIV-positive children progressed to an initial stage-3- defining diagnosis or death. The Intention to Treat (ITT) rate ratio (RR) of stage-3-defining diagnosis or death comparing the pre-cART and cART eras was estimated at 2·67 (95% confidence interval (CI): 1·.47, 4·84). The IV estimator was used to adjust for cART use misclassification, yielding an IV RR of 5·42 (95% CI: 2·99, 9·83). Weighting analyses did not markedly alter the results.

CONCLUSION

cART use decreased progression to stage-3-defining diagnosis or death. The use of noncompliance adjustments for cART misclassification in IV analyses may provide more robust evidence of cART's effectiveness than traditional ITT analysis.

Poverty, Pregnancy, and Birth Outcomes: A Study of the Earned Income Tax Credit

BACKGROUND

Economic interventions are increasingly recognised as a mechanism to address perinatal health outcomes among disadvantaged groups. In the US, the earned income tax credit (EITC) is the largest poverty alleviation programme. Little is known about its effects on perinatal health among recipients and their children. We exploit quasi-random variation in the size of EITC payments to examine the effects of income on perinatal health.

METHODS

The study sample includes women surveyed in the 1979 National Longitudinal Survey of Youth (n = 2985) and their children born during 1986-2000 (n = 4683). Outcome variables include utilisation of prenatal and postnatal care, use of alcohol and tobacco during pregnancy, term birth, birthweight, and breast-feeding status. We first examine the health effects of both household income and EITC payment size using multivariable linear regressions. We then employ instrumental variables analysis to estimate the causal effect of income on perinatal health, using EITC payment size as an instrument for household income.

RESULTS

We find that EITC payment size is associated with better levels of several indicators of perinatal health. Instrumental variables analysis, however, does not reveal a causal association between household income and these health measures.

CONCLUSIONS

Our findings suggest that associations between income and perinatal health may be confounded by unobserved characteristics, but that EITC income improves perinatal health. Future studies should continue to explore the impacts of economic interventions on perinatal health outcomes, and investigate how different forms of income transfers may have different impacts.