Body Mass Index Is an Important Predictor for Suicide: Results from a Systematic Review and Meta‐Analysis

Abstract

Public health concerns for the independent management of obesity and suicidal behavior are rising. Emerging evidence suggests body weight plays an important role in quantifying the risk of suicide. In light of these findings, we aimed to clarify the association between body mass index (BMI) and suicidal behavior by systematically reviewing and evaluating the literature. Studies were identified by searching MEDLINE, EMBASE, PsycINFO, and CINAHL from inception to January 2015, supplemented by hand and grey literature searches. Study screening, data extraction, and risk of bias assessment were conducted in duplicate. We included 38 observational studies. Meta-analyses supported an inverse association between BMI and completed suicide. Pooled summary estimates demonstrated that underweight was significantly associated with an increased risk of completed suicide (HR = 1.21, 95% CI 1.07 to 1.36, p = .002), and obesity (HR = 0.71, 95% CI 0.56 to 0.89, p = .003) and overweight (HR = 0.78, 95% CI 0.75 to 0.82, p < .0001) were significantly associated with a decreased risk of completed suicide relative to normal weight. A qualitative summary of the literature demonstrated conflicting evidence regarding the association between BMI and attempted suicide and revealed no association between BMI and suicidal ideation. BMI may be used to aid the assessment of suicide risk, especially that of completed suicide. However, unmeasured confounders and systematic biases of individual studies limit the quality of evidence.

Low digit ratio (2D:4D) in male suicide victims

“…We calculated repeated-measures analyses of variance (ANOVA) with repeated measurements on the factor “hand” (right, left) and separate ANOVAs with R2D:4D, L2D:4D, and Dr-l as dependent variables. Status of alcohol dependence and weight status were included in the statistical models, because these factors are associated with both 2D:4D (Fink et al 2006; Han et al 2016; Kornhuber et al 2011; Oyeyemi et al 2014) and suicide rate (Perera et al 2016; WHO 2014). The variables of interest (“group,” suicide versus non-suicide; “method of suicide,” chemical versus physical) and the potential confounders (“quantification method,” scan versus photograph; “status of alcohol dependence,” alcohol-dependent versus not alcohol-dependent versus unknown; “weight status,” normal weight ≤25 mg/m 2 versus overweight >25 kg/m 2 ) were included in the models as fixed factors.…”

Identifying Potential Causal Risk Factors for Self-Harm: A Polygenic Risk Scoring and Mendelian Randomisation Approach

“…45 Beyond psychiatric conditions, cognitive traits, physical traits, and personality traits were not found to be associated with self-harm using PRS approach, although previous observational findings found significant phenotypic associations for these three domains. 10–12 The absence of significant findings in this case is unlikely to be solely due to lack of power, given that GWAS for some of these traits are more powerful than GWAS for psychiatric conditions (e.g. BMI and education attainment).…”

“…BMI). 12 Second, we selected GWAS which only included participants of European ancestry and did not include UK Biobank participants (to avoid overlapping between discovery sample size and target sample). Finally, we excluded GWAS with effective sample sizes less than N = 15,000 to limit the use of underpowered PRS.…”

“…A range of individual vulnerabilities and traits can potentially lead to self-harm, such as psychiatric illnesses, 6 substance use, 7–9 cognitive abilities, 10 personality traits 11 and physical traits. 12 Although associations between these risk factors and self-harm have been shown in numerous observational studies, causality is difficult to infer reliably. Genetically informed designs can help in strengthening causal inference.…”

“…Firstly, these studies focused on PRS for psychiatric disorders or psychiatric symptoms, and did not include potential risk factors from other domains, such as substance use 7–9 , cognitive abilities, 10 personality traits 11 and physical traits. 12 Secondly, with two exceptions, 21,25 none of the studies had investigated SSH and NSSH simultaneously. Thirdly, these studies have a mixture of clinical and non-clinical samples with varying sample sizes ranging from 224 individuals 23 to 10,408 individuals, 18 making any comparison difficult.…”