Temporal Associations Among Body Mass Index, Fasting Insulin, and Systemic Inflammation: A Systematic Review and Meta-analysis

IMPORTANCE: Obesity is associated with a number of noncommunicable chronic diseases and is purported to cause premature death. OBJECTIVE: To summarize evidence on the temporality of the association between higher body mass index (BMI) and 2 potential mediators: chronic inflammation and hyperinsulinemia. DATA SOURCES: MEDLINE (1946 to August 20, 2019) and Embase (from 1974 to August 19, 2019) were searched, although only studies published in 2018 were included because of a high volume of results. The data analysis was conducted between January 2020 and October 2020. STUDY SELECTION AND MEASURES: Longitudinal studies and randomized clinical trials that measured fasting insulin level and/or an inflammation marker and BMI with at least 3 commensurate time points were selected. DATA EXTRACTION AND SYNTHESIS: Slopes of these markers were calculated between time points and standardized. Standardized slopes were meta-regressed in later periods (period 2) with standardized slopes in earlier periods (period 1). Evidence-based items potentially indicating risk of bias were assessed. RESULTS: Of 1865 records, 60 eligible studies with 112 cohorts of 5603 participants were identified. Most standardized slopes were negative, meaning that participants in most studies experienced decreases in BMI, fasting insulin level, and C-reactive protein level. The association between period 1 fasting insulin level and period 2 BMI was positive and significant (β = 0.26; 95% CI, 0.13-0.38; I(2) = 79%): for every unit of SD change in period 1 insulin level, there was an ensuing associated change in 0.26 units of SD in period 2 BMI. The association of period 1 fasting insulin level with period 2 BMI remained significant when period 1 C-reactive protein level was added to the model (β = 0.57; 95% CI, 0.27-0.86). In this bivariable model, period 1 C-reactive protein level was not significantly associated with period 2 BMI (β = –0.07; 95% CI, –0.42 to 0.29; I(2) = 81%). CONCLUSIONS AND RELEVANCE: In this meta-analysis, the finding of temporal sequencing (in which changes in fasting insulin level precede changes in weight) is not consistent with the assertion that obesity causes noncommunicable chronic diseases and premature death by increasing levels of fasting insulin.