Adults with excess weight or obesity, but not with overweight, report greater pain intensities than individuals with normal weight: a systematic review and meta-analysis

Context: Over 1.9 billion adult people have overweight or obesity. Considered as a chronic disease itself, obesity is associated with several comorbidities. Chronic pain affects approximately 60 million people and its connection with obesity has been displayed in several studies. However, controversial results showing both lower and higher pain thresholds in subjects with obesity compared to individuals with normal weight and the different parameters used to define such association (e.g., pain severity, frequency or duration) make it hard to draw straight forward conclusions in the matter. The objective of this article is to examine the relationship between overweight and obesity (classified with BMI as recommended by WHO) and self-perceived pain intensity in adults. Methods: A literature search was conducted following PRISMA guidelines using the databases CINAHL, Cochrane Library, EMBASE, PEDro, PubMed, Scopus and Web of Science to identify original studies that provide BMI values and their associated pain intensity assessed by self-report scales. Self-report pain scores were normalized and pooled within meta-analyses. The Cochrane’s Q test and I index were used to clarify the amount of heterogeneity; meta-regression was performed to explore the relationship between each outcome and the risk of bias. Results: Of 2194 studies, 31 eligible studies were identified and appraised, 22 of which provided data for a quantitative analysis. The results herein suggested that adults with excess weight (BMI ≥ 25.0) or obesity (BMI ≥ 30.0) but not with overweight (pre-obesity) alone (BMI 25.0–29.9), are more likely to report greater intensities of pain than individuals of normal weight (BMI 18.5–24.9). Subgroup analyses regarding the pathology of the patients showed no statistically significant differences between groups. Also, influence of age in the effect size, evaluated by meta-regression, was only observed in one of the four analyses. Furthermore, the robustness of the findi