Association between body mass index and suicidal behaviors: a systematic review protocol

Suicide is a significant problem throughout the world, impacting society from a public health, community, individual, and often a family perspective. Every year, over 800,000 lives are claimed by suicide [1]. Suicide is a complex phenomenon and requires evidence-based guidelines designed to recognize and mitigate modifiable suicide risk factors. A previous suicide attempt is considered a significant predictor of completed suicide in the general population [2,3]. Attempted suicide occurs 10 to 20 times more frequently than completed suicide [4]. Suicidal behaviors entail a complex set of ideas, thoughts, plans, and actions, which imply intention to end one’s life [4]. There are several well-established risk factors that collectively affect an individual’s vulnerability to suicidal behavior, including psychiatric and substance use disorders, lack of social support, family environment (for example, disrupted marital status), the occurrence of a major life stressor (for example, unemployment, grief), and various demographic factors, such as higher risk of attempted suicide among females and a greater risk of suicide completion among males [4-7]. Novel potential risk factors including biological markers, such as serum cholesterol [8], and physical measurements, such as body weight, have yet to be fully elucidated in their potential impact on suicide. The present systematic review will focus on the association between body mass index (BMI) and suicidal behavior, including completed suicide. The number of obese and overweight people worldwide has dramatically increased over the past three decades, with the number of overweight and obese adults and children rising from approximately 857 million individuals in 1980 to 2.1 billion in 2013 [9]. Excess body mass is a known risk factor of chronic illness, including cardiovascular disease, type 2 diabetes, osteoarthritis, and hypertension [10,11]. Furthermore, being overweight or obese is associated with social stigma among certain cultures, and negative attitudes towards obese individuals are evidently present in North American society [12-14]. As the prevalence of obesity rises globally [9], it is important to investigate the psychosocial burden that adiposity may have on affected individuals. Thus, an emerging body of research has attempted to investigate the relationship between BMI and suicidal behavior. Collectively, these investigations have recruited samples across all weight groups, from underweight to morbidly obese individuals. The majority of such studies have used BMI as a measure of relative body weight and as a marker of adiposity [15-21].

Considering the psychosocial stigma [14] and health related complications [10] of obesity, several systematic reviews and meta-analyses have suggested a positive association between obesity and depression [22-24]. Furthermore, about 60% of suicides occur among individuals with mood disorders [4], and 10% to 20% of patients hospitalized for depression commit suicide [25]. If obesity is indeed positively associated with depression, one might expect that a greater number of completed suicides are occurring among overweight and obese individuals. Surprisingly, however, several retrospective cohort studies have reported an inverse association between completed suicide and BMI, often in a dose-response type pattern, with the risk of suicide diminishing per unit increase in BMI [15-18,21,26]. The inverse association between suicide and BMI has been replicated in different countries using adult samples, albeit using predominantly male and Caucasian cohorts, including Swedish male conscripts [15], United States’ (US) male healthcare professionals [21], Norwegian men and women from a community cohort [18], a cohort of Danish men using conscription data [19], and a US general population cohort [17]. An American ecological study [27] reported that statewide obesity rates were inversely correlated with rates of completed suicide, and another ecological study conducted in the United Kingdom [28] similarly reported this inverse relationship in elderly females but found no such association among elderly males. Newer studies have since replicated an inverse relationship between BMI and completed suicide using longitudinal cohort designs [29,30], as well as with a case-control post-mortem investigation [31]. However, some studies have reported either no association [32-34], a J-shaped association [35], or a positive association between BMI and competed suicide [20,36]. A recent case-control post-mortem investigation [33] conducted in Australia found no association between BMI and completed suicide, failing to replicate the results of the aforementioned post-mortem design [31] out of Germany. Two recent longitudinal studies using Taiwanese [35] and German [36] cohorts reported a J-shaped and a positive relationship (respectively) between BMI and suicide. In summary, while many of the aforementioned studies point towards an inverse relationship between BMI and risk of completed suicide, the relationship remains equivocal and may depend on demographics such as sex or ethnicity. Moreover, the potential explanations for this apparent paradoxical inverse relationship remains unclear, with authors suggesting several potential mediating or correlative factors including impulsivity and case-fatality in relation to the chosen method of suicide [26,37], serum levels of cholesterol and its relation to central serotonin [8], leptin and leptin resistance [38,39], an insulin resistance based model [40], and dietary factors such as essential fatty acid intake [41].

The relationship between BMI and attempted suicide is also inconsistently reported with fewer studies compared to completed suicide, and the results are mixed [19,29,34,42-51]. Cohort studies have demonstrated a decrease in risk for attempted suicide associated with increasing BMI [19,29,34,45], while other studies have found a positive association between overweight and obese and suicide attempt [42,43,47,48,51]. Many of the studies examining BMI and attempted suicide are cross-sectional in design [42,44,46-49], thus precluding the ability to infer temporal or causal pathways. Furthermore, sex may complicate the relationship between suicide attempts and BMI, as sex was an apparent effect modifier in some reports [42]. Additionally, suicidal ideation may have a different association with body weight. Only a few studies have examined the relationship between relative body weight and suicidal ideation [42,47,48,50,52], limiting our ability to draw any conclusions. Two previous reviews have been published regarding the relationship between body weight and suicidal behavior [53,54]. One review provided a systematic review of the literature [53]; the other provided a narrative summary [54]. Both reviews cited that the collective studies were too few in number and heterogeneous to conduct a meta-analysis. Each review conducted a literature search up to 2011, and Klinitzke [53] did not include any studies published prior to January 2000. These reviews limited their searches to papers published in English, and Zhang [54] limited their search to PubMed only, precluding the inclusion of relevant articles from other databases. Zhang [54] did not present reasons for exclusion at each level of screening. Zhang reported on both adolescent and adult populations, and they classified some studies as using adult populations despite such studies including individuals as young as 15 [55,56]. The inclusion of both adolescents and adults is viewed critically due to differences in psychosocial development and differences in risk factors for suicide.

Since the conduct of these reviews, several observational studies have been published that have investigated body weight in relation to suicidal behavior [29-31,33-36,48-52], and an updated systematic review is required to contextually interpret the evidence. In addition, few studies included women in their samples when investigating BMI and suicidal behavior, and therefore, a more inclusive review will also help in understanding the association between BMI and suicidal behavior and the potential variation by sex.