Metabolic syndrome and dietary patterns: a systematic review and meta-analysis of observational studies—reply

Abstract

We thank Fabiani and colleagues for the attention given to our article “Metabolic syndrome and dietary patterns: a systematic review and meta-analysis of observational studies” [1]. We have conducted an accurate review of the manuscript and there are issues argued by Fabiani and colleagues on which we would like to comment. The authors mention that there are oversights in the inclusion of risk values associated with the prudent and the western dietary patterns in some papers due to being calculated considering a different pattern as reference (OR = 1). Specifically, the dietary patterns included in the articles from Sun 2014; Berg 2008; Kimokoti 2012; Song 2012; Leite 2009 and Duffey 2012 [2–7]. To assess the robustness of combined effect estimates, we performed a sensitivity analysis. The results show no evident changes in the risk after excluding the articles from Sun 2014; Berg 2008; Song 2012 and Leite 2009 [2, 3, 5, 6] either in the prudent/healthy dietary pattern [OR 0.83 (0.75, 0.91; I2 = 74.7%)] or in the western/ unhealthy dietary pattern [OR 1.21 (1.10, 1.32; I2 = 69.9%)] compared to the original results [1] (Table 1). The authors also mention that there were inconsistencies in the data included from the articles of Bell 2015 [8] and Arisawa 2014 [9]. After excluding these two articles [8, 9], there were no evident changes in the risk values in the prudent/healthy dietary pattern [OR 0.82 (0.75, 0.90; I2 = 73.7%)] or in the western/unhealthy dietary pattern [OR 1.28 (1.16, 1.41; I2 = 71.3%)] (Table 1). Furthermore, we also performed sensitivity analysis excluding the articles from Sun 2014; Berg 2008; Song 2012, Leite 2009 [2, 3, 5, 6] and the ones from Bell, 2015 [8] and Arisawa, 2014 [9] obtaining no evident changes in the risk either in the prudent/healthy dietary pattern [OR 0.82 (0.74, 0.91; I2 = 76.5%)] or in the western/ unhealthy dietary pattern [OR 1.23 (1.13, 1.35; I2 = 70.3%)] (Table 1). In addition, the authors mention on their letter that there were inconsistencies in the data inclusion from the studies of Hong 2012; Suliga 2015 and Choi 2015 [10–12]. Therefore, we have conducted a new meta-analysis of the western/ unhealthy dietary pattern correcting the data from the studies. Overall, the pooled OR (95% CI) for metabolic syndrome in a comparison of the highest to the lowest category of western/unhealthy dietary patterns in cross-sectional studies was 1.24 (95% CI 1.14, 1.34; P for heterogeneity = 0.0; and I2 = 67.8%) (Fig. 1). Potential sources of heterogeneity, such as sex (P = 0.701), geographic area (P = 0.491), a posteriori approach (P = 0.772), adjustment for key confounders (P = 0.690), metabolic syndrome definition (P = 0.310), age (P = 0.686) or health status (P = 0.395) produced only minor, non-significant differences (Table 2). In conclusion, the results, despite having excluded some of the articles previously included in our work and as we already described in our publication, show that an unhealthy/ western dietary pattern is associated with an increased risk of metabolic syndrome and that following a prudent/healthy dietary pattern is associated with a lower prevalence of metabolic syndrome, in cross-sectional studies. Consequently, the central findings of the original article remain unaffected. This reply refers to the letter available at https ://doi.org/10.1007/ s0039 4-019-02081 -7.