Sex-specific associations between body mass index and death before life expectancy: a comparative study from the USA and Sweden

ABSTRACT

Background: Understanding the impact of obesity on premature mortality is critical, as obesity has become a global health issue.

Objective: To contrast the relationship between body mass index (BMI) and premature death (all-cause; circulatory causes) in New York State (USA) and Northern Sweden.

Methods: Baseline data were obtained between 1989 and 1999 via questionnaires (USA) and health exams (Sweden), with mortality data from health departments, public sources (USA) and the Swedish Death Register. Premature death was death before life expectancy based on sex and year of birth. Within country and sex, time to premature death was compared across BMI groups (18.5–24.9 kg/m² (reference), 25–29.9 kg/m², 30.0–34.9 kg/m², ≥35.0 kg/m²) using Proportional Hazards regression. Absolute risk (deaths/100,000 person-years) was compared for the same stratifications among nonsmokers.

Results: 60,600 Swedish (47.8% male) and 31,198 US subjects (47.7% male) were included. Swedish males with BMI≥30 had increased hazards (HR) of all-cause premature death relative to BMI 18.5–24.9 (BMI 30–34.9, HR = 1.71 (95% CI: 1.44, 2.02); BMI≥35, HR = 2.89 (2.16, 3.88)). BMI≥25 had increased hazards of premature circulatory death (BMI 25–29.9, HR = 1.66 (1.32, 2.08); BMI 30–34.9, HR = 3.02 (2.26, 4.03); BMI≥35, HR = 4.91 (3.05, 7.90)). Among US males, only BMI≥35 had increased hazards of all-cause death (HR = 1.63 (1.25, 2.14)), while BMI 30–34.9 (HR = 1.83 (1.20, 2.79)) and BMI≥35 (HR = 3.18 (1.96, 5.15)) had increased hazards for circulatory death. Swedish females showed elevated hazards with BMI≥30 for all-cause (BMI 30–34.9, HR = 1.42 (1.18, 1.71) and BMI≥35, HR = 1.61 (1.21, 2.15) and with BMI≥35 (HR = 3.11 (1.72, 5.63)) for circulatory death. For US women, increased hazards were observed among BMI≥35 (HR = 2.10 (1.60, 2.76) for all-cause and circulatory HR = 3.04 (1.75, 5.30)). Swedish males with BMI≥35 had the highest absolute risk of premature death (762/100,000 person-years).

Conclusions: This study demonstrates a markedly increased risk of premature death associated with increasing BMI among Swedish males, a pattern not duplicated among females.

KEYWORDS:

• Obesity
• premature mortality
• longitudinal studies
• all-cause mortality
• circulatory disease mortality

Responsible Editor Peter Byass, Umeå University, Sweden

Responsible Editor Peter Byass, Umeå University, Sweden

Acknowledgments

The authors are grateful for the assistance of Göran Lönnberg and Nicole Krupa in preparing and updating the data sets for analysis.

Disclosure statement

No potential conflict of interest was reported by the authors.

Ethics and consent

All study procedures were in accordance with the ethical standards of the appropriate institutional research committees and with the 1964 Helsinki declaration and its later amendments. The Swedish study was approved by the Umeå University ethical review board (approval numbers Dnr 08-113M and 2014-196-32M); the US study was approved by the Mary Imogene Bassett Hospital Institutional Review Board (approval numbers 491 and 925). For the Swedish study, written informed consent was not collected; however, consent was implied via participation in the VIP. Participants were informed that their VIP exam data would be linked to mortality information for research purposes, and were offered the opportunity to withdraw. For the US study, signed informed consent was waived; consent was implied via return of the Bassett Health Census questionnaires.

Paper context

Few studies have contrasted obesity-linked mortality across individual countries. We found clear vulnerabilities among severely obese men in Sweden, up to a fivefold increased risk of premature death versus normal weight. In US men, the risk was much more moderate. No such disparity was found among females. Research to explain the mechanism of these gender-specific differences between countries may inform prevention initiatives globally. This, in turn, would help to mitigate the obesity-related risk of premature death.

Additional information

Funding

The data analyzed herein were produced by studies funded by the Västerbotten County Council and partially funded by the New York State Department of Health.

Notes on contributors

Melissa Scribani

Mrs. Scribani conducted all final statistical analyses, performed a literature review, drafted all sections of the manuscript, and revised and submitted the manuscript. Dr. Norberg assisted with the study design and procurement of Swedish data, contributed to the methods and discussion sections of the manuscript, and provided critical review and revision of the manuscript. Dr. Lindvall provided information on Swedish obesity epidemiology and performed critical review and revision of the manuscript. Dr. Weinehall assisted with study design and played an integral role in the procurement/updates of the Swedish data set, and performed critical review and revision of the manuscript. Dr. Sorensen provided critical review and revision of the manuscript. Dr. Jenkins conceived the study, supervised data collection of US health surveys and mortality data, compiled the combined data sets, performed preliminary statistical analysis, assisted in drafting the methods, results, and discussion sections of the manuscript, and provided critical review and revision of the manuscript.