Examkin

Question

Difficulty: MediumBias, Confounding, and Validity

Epidemiologists are designing an observational cohort study to explore the relationship between frequent consumption of ultra-processed foods and the incidence of chronic kidney disease (CKD) in adults over age 50. The investigators recognize that preexisting hypertension is strongly associated with both poor dietary habits and an increased risk of CKD. They are concerned that this variable could distort the true association between the dietary exposure and the outcome.

Which of the following strategies are valid methodological or statistical approaches to control for this specific variable? (Select all that apply)

  1. Restricting study enrollment exclusively to participants without a baseline history of hypertensionAnswer
  2. Including hypertension as an independent covariate in a multivariable regression model during data analysisAnswer
  3. C
    Utilizing an intention-to-treat framework to manage participants who alter their dietary habits during the follow-up period
  4. D
    Calculating a pooled, adjusted risk estimate even if the relative risk of CKD significantly differs between hypertensive and normotensive strata
  5. E
    Determining the positive predictive value of the dietary assessment tool to ensure it accurately identifies at-risk individuals

Answer

The correct approaches to control for confounding in this scenario are restricting the study population to individuals without hypertension and adjusting for hypertension using a multivariable regression model.
Confounding occurs when an extraneous variable is associated with both the exposure and the outcome, distorting their true relationship. This distortion can be controlled during the study design phase through restriction (e.g., only enrolling normotensive individuals), which prevents the confounder from varying within the study population. Alternatively, during the data analysis phase, confounding can be controlled by including the variable as a covariate in a multivariable regression model, which mathematically adjusts the exposure-outcome association for the influence of the confounder.

Step-by-Step Solution

1
Identify the methodological problem described in the scenario.
The scenario describes hypertension as a potential confounder because it is associated with both the exposure (ultra-processed diet) and the outcome (CKD).
Recognizing the epidemiological role of the variable is required to select appropriate statistical and methodological control strategies.
2
Evaluate each option as a valid method to control for confounding in either the design or analysis phase.
Restriction (a design-phase method) and multivariable regression (an analysis-phase method) are recognized and valid methods to effectively control for confounding.
Restriction eliminates confounding by ensuring all subjects have the same baseline level of the confounder. Regression mathematically adjusts for the confounder's effect during data analysis.
3
Identify why the remaining options are incorrect epidemiological approaches for this specific problem.
Intention-to-treat is used for preserving randomization in trials, pooling heterogeneous strata ignores effect modification, and positive predictive value evaluates diagnostic tests.
These techniques serve entirely different statistical purposes and do not mitigate the distortion caused by confounding variables.

Key Concept

Methods for controlling confounding variables in observational studies (e.g., Restriction, Matching, Stratification, Multivariable Regression).
Rate this question

Topics

Bias, Confounding, and Validity

More questions from this topic