Carryover Effects in Sibling Comparison Designs

Abstract

A convenient way of dealing with confounding is the sibling comparison design, where the outcome in exposed individuals is compared with the outcome in their unexposed siblings. The standard analysis of sibling comparison designs assumes that the exposure and outcome of an individual do not affect the exposure and outcome of his/her siblings, sometimes referred to as an absence of sibling carryover or contagion effects. Unfortunately, there are many situations where carryover effects are likely to be present. In this article, we explore the consequences of carryover effects for sibling comparison designs. We show, using causal diagrams, when and why carryover effects lead to bias, and we investigate the sign and magnitude of this bias under various scenarios.

Mild Traumatic Brain Injury and Criminal Charges and Convictions in Mid and Late Adolescence

“… Second, outcome-to-exposure spillover, whereby the criminal justice involvement of one sibling causes the other to suffer an increased or decreased risk of mTBI, would bias our results in an indeterminate direction. This possibility, however, is ruled out by design in our twin fixed-effect models because outcomes are ascertained 5 or more years after treatments, so one twin’s outcome cannot affect the other’s exposure. Similarly, outcome-to-exposure spillover between siblings is essentially ruled out in sibling fixed-effects models because siblings in the 1995-2000 cohorts can be at most 6 years apart in age and only 3% of siblings differ by 5 years or more years in age (mean birth spacing is 2.6 years).…”

Mild Traumatic Brain Injury and Criminal Charges and Convictions in Mid and Late Adolescence

“… Second, outcome-to-exposure spillover, whereby the criminal justice involvement of one sibling causes the other to suffer an increased or decreased risk of mTBI, would bias our results in an indeterminate direction. This possibility, however, is ruled out by design in our twin fixed-effect models because outcomes are ascertained 5 or more years after treatments, so one twin’s outcome cannot affect the other’s exposure. Similarly, outcome-to-exposure spillover between siblings is essentially ruled out in sibling fixed-effects models because siblings in the 1995-2000 cohorts can be at most 6 years apart in age and only 3% of siblings differ by 5 years or more years in age (mean birth spacing is 2.6 years).…”

Re-examining the link between childhood maltreatment and substance use disorder: a prospective, genetically informative study

“…Carryover effect for exposures and/ or outcomes, typically occur either from older to younger, or from younger to older siblings. 24 We found no significant differences for SUD risks when the older versus when the younger child was exposed to CM (p=0.31).…”

Educational inequality in cardiovascular diseases: a sibling approach

“…Finally, we find it less likely that one sibling's education would influence co-siblings' education or risk of CVD and thus that our study should be subject to carryover effects. [13] In conclusion, this study showed that family factors explained the associations between education and CVD, IHD and stroke to varying degrees, being least pronounced for CVD and most pronounced for IHD -especially before the age of 45 years. Thus, family factors should be given most attention in interventions aimed at reducing inequalities in IHD.…”

Mild Traumatic Brain Injury and Criminal Charges and Convictions in Mid and Late Adolescence

“… Second, outcome-to-exposure spillover, whereby the criminal justice involvement of one sibling causes the other to suffer an increased or decreased risk of mTBI, would bias our results in an indeterminate direction. This possibility, however, is ruled out by design in our twin fixed-effect models because outcomes are ascertained 5 or more years after treatments, so one twin’s outcome cannot affect the other’s exposure. Similarly, outcome-to-exposure spillover between siblings is essentially ruled out in sibling fixed-effects models because siblings in the 1995-2000 cohorts can be at most 6 years apart in age and only 3% of siblings differ by 5 years or more years in age (mean birth spacing is 2.6 years).…”

Re-examining the link between childhood maltreatment and substance use disorder: a prospective, genetically informative study

“…Carryover effect for exposures and/ or outcomes, typically occur either from older to younger, or from younger to older siblings. 24 We found no significant differences for SUD risks when the older versus when the younger child was exposed to CM (p=0.31).…”

Educational inequality in cardiovascular diseases: a sibling approach

“…Finally, we find it less likely that one sibling's education would influence co-siblings' education or risk of CVD and thus that our study should be subject to carryover effects. [13] In conclusion, this study showed that family factors explained the associations between education and CVD, IHD and stroke to varying degrees, being least pronounced for CVD and most pronounced for IHD -especially before the age of 45 years. Thus, family factors should be given most attention in interventions aimed at reducing inequalities in IHD.…”

Mild Traumatic Brain Injury and Criminal Charges and Convictions in Mid and Late Adolescence

“… Second, outcome-to-exposure spillover, whereby the criminal justice involvement of one sibling causes the other to suffer an increased or decreased risk of mTBI, would bias our results in an indeterminate direction. This possibility, however, is ruled out by design in our twin fixed-effect models because outcomes are ascertained 5 or more years after treatments, so one twin’s outcome cannot affect the other’s exposure. Similarly, outcome-to-exposure spillover between siblings is essentially ruled out in sibling fixed-effects models because siblings in the 1995-2000 cohorts can be at most 6 years apart in age and only 3% of siblings differ by 5 years or more years in age (mean birth spacing is 2.6 years).…”

Re-examining the link between childhood maltreatment and substance use disorder: a prospective, genetically informative study

“…Carryover effect for exposures and/ or outcomes, typically occur either from older to younger, or from younger to older siblings. 24 We found no significant differences for SUD risks when the older versus when the younger child was exposed to CM (p=0.31).…”

Educational inequality in cardiovascular diseases: a sibling approach

“…Finally, we find it less likely that one sibling's education would influence co-siblings' education or risk of CVD and thus that our study should be subject to carryover effects. [13] In conclusion, this study showed that family factors explained the associations between education and CVD, IHD and stroke to varying degrees, being least pronounced for CVD and most pronounced for IHD -especially before the age of 45 years. Thus, family factors should be given most attention in interventions aimed at reducing inequalities in IHD.…”