Paediatric obesity appears to lower the risk of diabetes if selection bias is ignored

The impact of sex, body mass index and chronic kidney disease on outcomes following percutaneous coronary intervention

OBJECTIVE

This study sought to evaluate 1) the relationship between body mass index (BMI), chronic kidney disease (CKD) and bleeding complications in patients undergoing percutaneous coronary intervention (PCI); and 2) whether CKD modified the effect of BMI on major bleeding and major adverse cardiac and cerebrovascular events (MACCE).

BACKGROUND

The interaction of CKD, sex and BMI in patients undergoing PCI is unclear.

METHODS

Between 2010 and 2018, a total of 31,116 patients underwent PCI at six New York metropolitan area hospitals. Bleeding complications were classified by the Bleeding Academic Research Consortium (BARC). Major bleeding was defined as BARC≥3. MACCE was the composite of in-hospital death; myocardial infarction; cerebrovascular events and major bleeding complications. Interaction on multiplicative scales was assessed adjusting for other factors. A three-way multiplicative interaction between BMI, CKD and sex were considered and evaluated for both endpoints of primary interest (BARC≥3 and MACCE).

RESULTS

Patients with BARC≥3 bleeding were older (p < 0.0001) and more likely female (p < 0.0001). A 3-way interaction existed between sex, BMI, and CKD on BARC≥3 (p = 0.02). Specifically, the effect of CKD status on odds of BARC≥3 depended on BMI group among males, whereas BMI did modify the relationship between CKD and BARC≥3 among females; after stratification by sex, a significant interaction between BMI and CKD was present in females (p = 0.03) but not in males (p = 0.43). Among females without CKD, normal BMI patients had the greatest odds of BARC≥3 compared to obese or overweight females. Contrasted to females without CKD, among females with CKD there was no significant increased odds of BARC≥3 in normal BMI patients compared to obese or overweight females. However, overweight females with CKD had a significantly increased odds of BARC≥3 compared to obese females with CKD. Furthermore, obese females with CKD had significantly greater BARC≥3 odds compared to obese females without CKD. Similarly, overweight females with CKD had an increased odds of BARC≥3 compared to overweight females without CKD. No significant interactions were found for the odds of MACCE.

CONCLUSION

In patients undergoing PCI, there was evidence of a significant and complex 3-way interaction between BMI, CKD and sex for major bleeding events. The predicted probability of major bleeding was greater for females than for male patients, and for both sexes, greater among those with CKD, but BMI group influences these probabilities. Obese females with kidney disease had the lowest incidence of bleeding complications when compared with overweight or normal weight female patients undergoing PCI. This interaction was not seen in the male group or for MACCE. Furthermore, age, cardiogenic shock, STEMI and use of IABP/Impella were each independently associated with odds of major bleeding (among both males and females) and MACCE.

Big data in autism research: Methodological challenges and solutions

While the concept of big data has emerged over the past decade as a hot topic in nearly all areas of scientific inquiry, it has rarely been discussed in the context of autism research. In this commentary we describe aspects of big data that are relevant to autism research and methodological issues such as confounding and data error that can hamper scientific investigation. Although big data studies can have transformative impact, bigger is not always better, and big data require the same methodological considerations and interdisciplinary collaboration as "small data" to extract useful scientific insight.

Relation of Body Mass Index to Outcomes in Acute Coronary Syndrome

We assessed the association of BMI with all-cause and cardiovascular (CV) mortality in a contemporary acute coronary syndrome cohort. Patients from the Australian Cooperative National Registry of Acute Coronary Care, Guideline Adherence and Clinical Events and Global Registry of Acute Coronary Events between 2009 and 2019, were divided into BMI subgroups (underweight: <18.5, healthy: 18.5 to 24.9, overweight: 25 to 29.9, obese: 30 to 39.9, extremely obese: >40). Logistic regression was used to determine the association between BMI group and outcomes of all cause and CV death in hospital, and at 6 months. 8,503 patients were identified, mean age 64 ± 13, 72% male. The BMI breakdown was: underweight- 95, healthy- 2,140, overweight- 3,258, obese- 2,653, extremely obese- 357. Obese patients were younger (66 ± 12 vs 67 ± 13), with more hypertension, diabetes, and dyslipidemia vs healthy (all p < 0.05). Obese had lower hospital mortality than healthy: all-cause: 1% versus 4%, aOR (95% CI): 0.49(0.27, 0.87); CV: 1% versus 3%, 0.51(0.27, 0.96). At 6-month underweight had higher mortality than healthy: all-cause: 11% versus 4%, 2.69(1.26, 5.76); CV: 7% versus 1%, 3.54(1.19, 10.54); whereas obese had lower mortality: all-cause: 1% versus 4%, 0.48(0.29, 0.77); CV: 0.4% versus 1%, 0.42(0.19, 0.93). When BMI was plotted as a continuous variable against outcome a U-shaped relationship was demonstrated, with highest event rates in the most obese (>60). In conclusion, BMI is associated with mortality following an acute coronary syndrome. Obese patients had the best outcomes, suggesting persistence of the obesity paradox. However, there was a threshold effect, and favorable outcomes did not extend to the most obese.

Reflection on modern methods: selection bias-a review of recent developments

Selection bias remains a more difficult bias to understand than confounding or measurement error. Past definitions have not always been illuminating and a simple method (such as the change-in-estimate method for confounding) has not been available to determine its presence and magnitude in the study sample. A better understanding of the nature of the bias has led to the definition of endogenous selection bias. It is the result of conditioning on a collider variable, itself caused by two other variables; the latter variables become spuriously associated. Conditioning on a variable in the analysis that is a collider or on an indicator of sample selection has the same effect. Note that selection bias is possible even in the absence of a collider, but in the presence of endogenous selection bias, the concern is whether it is possible to identify a causal effect in the sample. Conditions have been outlined to determine that. However, even if conditions are met to identify a causal effect in the study sample, its generalization to a defined target population is not a given.We discuss the concept of endogeneity and the sources of endogenous selection bias in observational studies. We then briefly address the terms generalizability, target population (or alternative formulations) and transportability. We outline the explicit conditions to identify causal effects in studies affected by selection bias: they involve exchangeability between exposed and unexposed and exchangeability between sampled and unsampled. We briefly describe methods to generalize estimated causal effects to the target population. The latter usually require data from the target population. Finally we discuss sensitivity analyses; some are limited to providing an indication of the presence and direction of the bias and others can provide corrected estimates with user-supplied selection bias parameters.