Risk factors relate to the variability of health outcomes as well as the mean: a GAMLSS tutorial

  1. David Bann  Is a corresponding author
  2. Liam Wright
  3. Tim J Cole
  1. University College London, United Kingdom

Abstract

Background: Risk factors or interventions may affect the variability as well as the mean of health outcomes. Understanding this can aid aetiological understanding and public health translation, in that interventions which shift the outcome mean and reduce variability are typically preferable to those which affect only the mean. However, most commonly used statistical tools do not test for differences in variability. Tools that do have few epidemiological applications to date, and fewer applications still have attempted to explain their resulting findings. We thus provide a tutorial for investigating this using GAMLSS (Generalised Additive Models for Location, Scale and Shape).

Methods: The 1970 British birth cohort study was used, with body mass index (BMI; N=6,007) and mental wellbeing (Warwick-Edinburgh Mental Wellbeing Scale; N=7,104) measured in midlife (42-46 years) as outcomes. We used GAMLSS to investigate how multiple risk factors (sex, childhood social class and midlife physical inactivity) related to differences in health outcome mean and variability.

Results: Risk factors were related to sizable differences in outcome variability-for example males had marginally higher mean BMI yet 28% lower variability; lower social class and physical inactivity were each associated with higher mean and higher variability (6.1% and 13.5% higher variability, respectively). For mental wellbeing, gender was not associated with the mean while males had lower variability (-3.9%); lower social class and physical inactivity were each associated with lower mean yet higher variability (7.2% and 10.9% higher variability, respectively).

Conclusions: The results highlight how GAMLSS can be used to investigate how risk factors or interventions may influence the variability in health outcomes. This underutilised approach to the analysis of continuously distributed outcomes may have broader utility in epidemiologic, medical, and psychological sciences. A tutorial and replication syntax is provided online to facilitate this (https://osf.io/5tvz6/).

Funding: DB is supported by the Economic and Social Research Council (grant number ES/M001660/1), The Academy of Medical Sciences / Wellcome Trust ('Springboard Health of the Public in 2040' award: HOP001/1025); DB and LW are supported by the Medical Research Council (MR/V002147/1). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability

All data are available to download from the UK Data Archive: https://beta.ukdataservice.ac.uk/datacatalogue/series/series?id=200001

Article and author information

Author details

  1. David Bann

    Centre for Longitudinal Studies, Social Research Institute, University College London, london, United Kingdom
    For correspondence
    david.bann@ucl.ac.uk
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6454-626X
  2. Liam Wright

    Centre for Longitudinal Studies, Social Research Institute, University College London, london, United Kingdom
    Competing interests
    The authors declare that no competing interests exist.
  3. Tim J Cole

    Great Ormond Street Institute of Child Health, University College London, london, United Kingdom
    Competing interests
    The authors declare that no competing interests exist.

Funding

Medical Research Council (MR/V002147/1)

  • David Bann
  • Liam Wright

Economic and Social Research Council (ES/M001660/1)

  • Liam Wright

Wellcome Trust (HOP001/1025)

  • David Bann

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Reviewing Editor

  1. Belinda Nicolau, McGill University, Canada

Ethics

Human subjects: This paper uses secondary data analysis using data from a cohort study which has been followed-up since birth in 1970. Cohort members provided informed consent, and the study received full ethical approval - most recently from the NRES Committee South East Coast-Brighton and Sussex.

Version history

  1. Preprint posted: March 31, 2021 (view preprint)
  2. Received: July 20, 2021
  3. Accepted: January 4, 2022
  4. Accepted Manuscript published: January 5, 2022 (version 1)
  5. Version of Record published: January 26, 2022 (version 2)

Copyright

© 2022, Bann et al.

This article is distributed under the terms of the Creative Commons Attribution License permitting unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,415
    views
  • 189
    downloads
  • 7
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. David Bann
  2. Liam Wright
  3. Tim J Cole
(2022)
Risk factors relate to the variability of health outcomes as well as the mean: a GAMLSS tutorial
eLife 11:e72357.
https://doi.org/10.7554/eLife.72357

Share this article

https://doi.org/10.7554/eLife.72357

Further reading

    1. Epidemiology and Global Health
    Caroline Krag, Maria Saur Svane ... Tinne Laurberg
    Research Article

    Background:

    Comorbidity with type 2 diabetes (T2D) results in worsening of cancer-specific and overall prognosis in colorectal cancer (CRC) patients. The treatment of CRC per se may be diabetogenic. We assessed the impact of different types of surgical cancer resections and oncological treatment on risk of T2D development in CRC patients.

    Methods:

    We developed a population-based cohort study including all Danish CRC patients, who had undergone CRC surgery between 2001 and 2018. Using nationwide register data, we identified and followed patients from date of surgery and until new onset of T2D, death, or end of follow-up.

    Results:

    In total, 46,373 CRC patients were included and divided into six groups according to type of surgical resection: 10,566 Right-No-Chemo (23%), 4645 Right-Chemo (10%), 10,151 Left-No-Chemo (22%), 5257 Left-Chemo (11%), 9618 Rectal-No-Chemo (21%), and 6136 Rectal-Chemo (13%). During 245,466 person-years of follow-up, 2556 patients developed T2D. The incidence rate (IR) of T2D was highest in the Left-Chemo group 11.3 (95% CI: 10.4–12.2) per 1000 person-years and lowest in the Rectal-No-Chemo group 9.6 (95% CI: 8.8–10.4). Between-group unadjusted hazard ratio (HR) of developing T2D was similar and non-significant. In the adjusted analysis, Rectal-No-Chemo was associated with lower T2D risk (HR 0.86 [95% CI 0.75–0.98]) compared to Right-No-Chemo.

    For all six groups, an increased level of body mass index (BMI) resulted in a nearly twofold increased risk of developing T2D.

    Conclusions:

    This study suggests that postoperative T2D screening should be prioritised in CRC survivors with overweight/obesity regardless of type of CRC treatment applied.

    Funding:

    The Novo Nordisk Foundation (NNF17SA0031406); TrygFonden (101390; 20045; 125132).

    1. Epidemiology and Global Health
    Xiaoxin Yu, Roger S Zoh ... David B Allison
    Review Article

    We discuss 12 misperceptions, misstatements, or mistakes concerning the use of covariates in observational or nonrandomized research. Additionally, we offer advice to help investigators, editors, reviewers, and readers make more informed decisions about conducting and interpreting research where the influence of covariates may be at issue. We primarily address misperceptions in the context of statistical management of the covariates through various forms of modeling, although we also emphasize design and model or variable selection. Other approaches to addressing the effects of covariates, including matching, have logical extensions from what we discuss here but are not dwelled upon heavily. The misperceptions, misstatements, or mistakes we discuss include accurate representation of covariates, effects of measurement error, overreliance on covariate categorization, underestimation of power loss when controlling for covariates, misinterpretation of significance in statistical models, and misconceptions about confounding variables, selecting on a collider, and p value interpretations in covariate-inclusive analyses. This condensed overview serves to correct common errors and improve research quality in general and in nutrition research specifically.