- Like Mother, Like Daughter: a reflection on a study - September 10, 2018
- Lightly Salted: The Food Industry and Public Health - November 20, 2016
- The Blame Game: The Sugar Industry and Coronary Heart Disease - September 22, 2016
It is well understood in the medical community that overweight and obesity are growing problems both in the US and worldwide. According to the World Health Organization (WHO), there are currently 1.3 billion adults worldwide who are overweight (Body Mass Index (BMI) between 25-30 kg/m2) and an additional 600 million who are obese (BMI ≥ 30 kg/m2). By comparison, about 800 million people are malnourished worldwide, according to the World Hunger Education Service. While still staggering, the malnourished population is shrinking, and the overweight and obese populations are growing. Regardless of our geographical location or medical specialty, we will be responsible for providing clinical care to this growing segment of the population as future health care providers.
We know that obesity is linked to chronic diseases such as diabetes, hypertension, and cardiovascular disease, but is there a link to overall mortality? This past July, The Global BMI Mortality Collaboration published a global meta-analysis of 239 studies from 1970 to 2015 examining the relationship between body mass index and all cause mortality. The study was stratified by geographical region and used strict exclusion criteria to exclude patients with a past or current history of smoking and those currently suffering from chronic disease, as these factors can make an individual lose weight and confound the relationship between BMI and mortality. Additionally, the authors excluded data from the first 5 years of each study in the meta-analysis to exclude any patients who may have died from underlying conditions. The final analysis represents data from 3,951,455 adult participants in 189 studies.
Lowest all-cause mortality was seen in those with BMI between 20-25 kg/m2. The WHO defines normal weight as patients who have a BMI between 18.5 kg/m2 to 25 kg/m2. A higher risk of mortality was seen in the populations both below and above this range. Interestingly, higher mortality was also seen in individuals with a BMI in the range of 18.5-20 kg/m2, even though the WHO considers this population in the lower end of normal weight. Dr. Shilpa Bhupathiraju, research scientist at Harvard T.H. Chan School of Public Health and co-lead author of the study, postulated, “Although our analysis excluded those with chronic diseases at baseline and never smokers, it is possible that the 18.5-20 kg/m2 BMI group has individuals with underlying chronic diseases, especially those with a long latency, that can result in low body weight and a higher risk of death.”
Study authors found that BMI above 25 kg/m2 was associated with a higher risk of mortality, as compared to those populations with BMI between 20-25 kg/m2. Risk of mortality increased almost linearly with increasing BMI above 25 kg/m2. This trend was seen across all geographical regions. Furthermore, a higher BMI was associated with a much higher mortality among men than women. Extrapolating that the observed relationships were largely causal, study authors estimated that if overweight and obese populations were to have normal BMI (WHO-defined), about one in five premature deaths in North America would be avoided, as well as one in six in Australia and New Zealand, one in seven in Europe, and one in 20 in East Asia.
While not altogether surprising, these results contradict another meta-analysis, published in the Journal of the American Medical Association in 2013. Flegal et al. found that, relative to normal weight, overweight (25-30 kg/m2) was actually associated with lower overall mortality, and grade 1 obesity (30-35 kg/m2) was associated with no increased risk.
How did two large-scale studies asking the same questions come to such drastically different results?
The Global BMI Mortality Collaboration used more rigorous exclusion criteria to avoid serious methodological biases that can invalidate a study’s findings. They were able to do so because they used individual participant level data. Flegal et al., on the other hand, used only published data. The Global BMI Mortality Collaboration study included only patients who had never smoked or had any major chronic disease at baseline. Weight loss due to underlying illness or smoking can deceptively suggest comparable or even higher mortality in normal weight individuals than overweight or obese people.
Complicated statistical analyses, hazard ratios, and variance estimates aside, it is clear that overweight and obesity are associated with increased morbidity. Now it appears they are associated with increased mortality as well. While the interaction between BMI and mortality is clear, what exactly is the relationship? Why is it different between men and women? Does the location of excess fat play a role in this relationship? Men tend to carry excess weight around the abdomen, while women tend to carry it on legs and hips. Is it possible that the greater mortality risk among obese men is connected to greater abdominal obesity?
This study’s population was restricted to adults, but childhood obesity is a growing problem in the United States and weight status could be a useful tool to predict long-term health outcomes. “Several well conducted studies,” according to Dr. Bhupathiraju, “have repeatedly shown that childhood overweight and obesity is associated with a higher risk for chronic diseases, such as cardiovascular disease, type 2 diabetes, and many cancers, in adulthood, all of which increase the risk of mortality.”
A growing percentage of our population is overweight or obese, and a growing number of the health problems we as clinicians will see will be related to body weight. But what does this mean for us as health care providers? What does the role of clinician look like for a patient population struggling more and more with body weight? BMI is certainly not a perfect measure, but it is a very good gauge of a person’s level of body fat, one notable exception being for athletes who are muscular. But, the majority of the general population is not athletic and BMI can be an excellent conversation-starter for clinicians and students who might feel uncomfortable discussing weight-related issues.
It is important to discuss other factors such as body composition, food choices, and physical activity that play a crucial role in the obesity-mortality relationship. Perhaps BMI is one tool for clinicians to use to assess the health of their patients among many. Body weight is multifactorial, complex, and patient-specific. As clinicians, we can use BMI as one way to start these complex conversations about weight to ensure the best possible health of our patients.
Ultimately, these one-on-one conversations with patients will be crucial, but population-wide interventions to prevent and/or reverse the obesity epidemic will be crucial to ending the obesity epidemic. Dr. Bhupathiraju noted that for community wide interventions to be successful they “need to be designed such that they achieve changes in both eating behaviors and amount of physical activity.” Of course such interventions need to be robust and encompass all aspects of patients lives. It is hard to impact the trajectory of a patient’s chronic disease in a singular office visit, and as physicians there are many opportunities for us to affect the health of our patients. Dr. Bhupathiraju stressed that, “interventions that improve the surrounding environment, such as increasing accessibility to healthy foods, taxing fast foods and sugar sweetened beverages, and limiting portion sizes have all been shown to have an effect on body weight…Policy changes that could improve physical activity, such as increasing green space, bike paths, and walk-ability can also impact body weight.”
As physicians, we are in a unique position to impact policy and patient behavior. BMI can be a useful tool to start the conversation with patients and open opportunities for community-wide conversations about obesity and public health.