Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Mortality risk relationship using standard categorized BMI or knee-height based BMI – does the overweight/lower mortality paradox hold true?

Gavriilidou, Nivetha Natarajan LU orcid ; Pihlsgård, Mats LU ; Elmståhl, Sölve LU and Ekström, Henrik LU (2024) In Aging clinical and experimental research 36(1).
Abstract

Background: The body mass index (BMI) is prone to misclassification of obesity due to age-related height loss and resulting measurement errors. Knee-height based BMI (KH-BMI) has not been previously studied in relation to mortality risk in older adults. Aim: To evaluate the age- and sex-specific mortality risk relationship using classic BMI and knee height predicted BMI (KH-BMI) overweight and obesity in a 15-year follow-up study including older Swedish adults aged 60–93 years. Methods: A 15-year follow-up study among 2,786 individuals aged ≥ 60 years. Height, weight and KH were measured. KH-predicted height was estimated using formulated gender-specific equations. Classic BMI and KH-BMI (kg/m2) were calculated. Mortality... (More)

Background: The body mass index (BMI) is prone to misclassification of obesity due to age-related height loss and resulting measurement errors. Knee-height based BMI (KH-BMI) has not been previously studied in relation to mortality risk in older adults. Aim: To evaluate the age- and sex-specific mortality risk relationship using classic BMI and knee height predicted BMI (KH-BMI) overweight and obesity in a 15-year follow-up study including older Swedish adults aged 60–93 years. Methods: A 15-year follow-up study among 2,786 individuals aged ≥ 60 years. Height, weight and KH were measured. KH-predicted height was estimated using formulated gender-specific equations. Classic BMI and KH-BMI (kg/m2) were calculated. Mortality data was obtained from the Swedish death registry. Questionnaires were used to collect data on obesity-related lifestyle factors and comorbidities. Results: Cox regression revealed that using the classic BMI, when comparing with the normal/underweight reference group, there was a mortality risk among overweight men (HR = 0.67, 0.52–0.87), overweight women (HR = 0.79, 0.65–0.97), and obese men (HR = 0.60, 0.41–0.89) aged ≥ 80 years old. Using the KH-BMI, only overweight men and overweight women aged ≥ 80 years had a lower mortality risk, men (HR = 0.71, 0.55–0.92); women (HR = 0.77, 0.62–0.95) after adjusting for obesity-related lifestyle factors and comorbidities. Discussion: There is evidence that obesity is overestimated by the BMI, in comparison with the KH-BMI classification. In terms of mortality risk and after adjusting for height, there remains a paradoxical protective association between overweight and mortality. Conclusion: Regardless of classic BMI or KH-BMI estimation, overweight men and women aged ≥ 80 years had a lower mortality risk compared to normal/underweight men and women ≥ 80 years.

(Less)
Please use this url to cite or link to this publication:
author
; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Classic BMI, Knee-height based BMI, Mortality, Older adults, Population study
in
Aging clinical and experimental research
volume
36
issue
1
article number
88
publisher
Kurtis
external identifiers
  • pmid:38587702
  • scopus:85189782343
ISSN
1594-0667
DOI
10.1007/s40520-024-02742-6
language
English
LU publication?
yes
id
497b576f-2267-4a80-8edd-2fffa11d3de0
date added to LUP
2024-04-23 13:10:33
date last changed
2024-06-18 18:08:39
@article{497b576f-2267-4a80-8edd-2fffa11d3de0,
  abstract     = {{<p>Background: The body mass index (BMI) is prone to misclassification of obesity due to age-related height loss and resulting measurement errors. Knee-height based BMI (KH-BMI) has not been previously studied in relation to mortality risk in older adults. Aim: To evaluate the age- and sex-specific mortality risk relationship using classic BMI and knee height predicted BMI (KH-BMI) overweight and obesity in a 15-year follow-up study including older Swedish adults aged 60–93 years. Methods: A 15-year follow-up study among 2,786 individuals aged ≥ 60 years. Height, weight and KH were measured. KH-predicted height was estimated using formulated gender-specific equations. Classic BMI and KH-BMI (kg/m<sup>2</sup>) were calculated. Mortality data was obtained from the Swedish death registry. Questionnaires were used to collect data on obesity-related lifestyle factors and comorbidities. Results: Cox regression revealed that using the classic BMI, when comparing with the normal/underweight reference group, there was a mortality risk among overweight men (HR = 0.67, 0.52–0.87), overweight women (HR = 0.79, 0.65–0.97), and obese men (HR = 0.60, 0.41–0.89) aged ≥ 80 years old. Using the KH-BMI, only overweight men and overweight women aged ≥ 80 years had a lower mortality risk, men (HR = 0.71, 0.55–0.92); women (HR = 0.77, 0.62–0.95) after adjusting for obesity-related lifestyle factors and comorbidities. Discussion: There is evidence that obesity is overestimated by the BMI, in comparison with the KH-BMI classification. In terms of mortality risk and after adjusting for height, there remains a paradoxical protective association between overweight and mortality. Conclusion: Regardless of classic BMI or KH-BMI estimation, overweight men and women aged ≥ 80 years had a lower mortality risk compared to normal/underweight men and women ≥ 80 years.</p>}},
  author       = {{Gavriilidou, Nivetha Natarajan and Pihlsgård, Mats and Elmståhl, Sölve and Ekström, Henrik}},
  issn         = {{1594-0667}},
  keywords     = {{Classic BMI; Knee-height based BMI; Mortality; Older adults; Population study}},
  language     = {{eng}},
  number       = {{1}},
  publisher    = {{Kurtis}},
  series       = {{Aging clinical and experimental research}},
  title        = {{Mortality risk relationship using standard categorized BMI or knee-height based BMI – does the overweight/lower mortality paradox hold true?}},
  url          = {{http://dx.doi.org/10.1007/s40520-024-02742-6}},
  doi          = {{10.1007/s40520-024-02742-6}},
  volume       = {{36}},
  year         = {{2024}},
}