In this study, it was discovered that high levels of WHtR increased the hazard of cardiovascular and all-cause mortality in the general population by 23% and 39%, respectively. It also increases the possibility of death due to cancer by 21% in people with/without primary cancer.
Except for three outcomes which had low level of certainty, the grade of evidence was moderate for other outcomes based on the GRADE criteria. These results are due to several reasons. First, the risk of bias was checked for each study separately, and no serious issues were found. Second, the narrow confidence intervals of the meta-analysis results suggest that the study's imprecision was within an acceptable range. Moreover, the high observed heterogeneity leaded to the reduction of the consistency level, and the overall certainty of the findings. In the case of publication bias, it was found to have no impact on most of the outcomes. Finally, in accordance with the present results, previous studies have also demonstrated that different anthropometrics, particularly indices of central fatness, have a high predictive potential for mortality. For example, in a recent systematic review (68), the association of several indices of central fatness with mortality was examined. They indicated that the hazard of all-cause mortality rises by 24% per 0.1 unit increase in WHtR. This result is in line with other indices of central obesity since they also reported that high waist circumference and waist-to-hip ratio levels are related to 11% and 20% of increased mortality rates, respectively. Overall, they concluded that measures of central obesity are positively and significantly associated with a higher all-cause mortality hazard. This finding broadly supports the work of our study in this area linking WHtR with the hazard of mortality. Another study confirming the association between indices of central fatness and mortality was reported by Cameron et al. (2013) (41), in which they suggested that the waist-to-hip ratio is a potent predictor in the case of all-cause mortality with 53% of excess hazard of death.
Regarding other anthropometrics, there exist some measures showing promising potential for predicting mortality; however, they are not commonly used. For instance, it has been revealed that the A-body shape index (ABSI) can be a reliable predictor for all-cause mortality, and high levels of this index can lead to 15% and 9% more mortality risk in men and women (69). These results seem to be consistent with another research by Jayedi et al. (2020) (68), which found that higher levels of ABSI are related to 15% more mortality in the general population (both sexes). Altogether, these findings suggest that central obesity is associated with an excess hazard of death (70–74).
Central obesity and elevated hazard of mortality
Although the reason why central obesity leads to elevated rates of mortality is not totally clear, studies suggest various mechanisms for each individual disease. For instance, a strong relationship between central obesity and an elevated risk of death due to CVD has been reported in the literature. One strong piece of evidence in this area is reported by Cameron et al. (2013), in which they reported that central fatness (defined as high levels of waist-to-hip ratio) is related to 106% excess hazard of death in people with CVDs. Although the reason for this relation is not completely understood, this observed increase in mortality hazard could be attributed to the worsening of CVDs or the presence of comorbidities. That is because central fatness is believed to be a potent risk factor for high blood pressure (75,76), myocardial infarction (77), left ventricular dysfunction (70), and coronary artery disease (71–73). Given that the reasons for the relationship between central fatness and elevated mortality risk may not be restricted to the mentioned mechanisms, future studies on the current topic are therefore recommended.
Too much about CVDs, several reports have shown that mortality rates in patients suffering from non-CVDs can also be increased by central fatness (69,73,74,78,79). Studies suggest various explanations regarding why central obesity leads to higher risks of death due to non-CVDs such as cancer. For example, in a systematic review by Ashwell et al. (2012) (77), the authors declared that it seems possible that these results are due to the high inflammatory activity of the visceral fat accumulated within the abdominal cavity. This finding is consistent with that of Oliveros et al. (2014) (73), in which it is shown that plasma inflammatory interleukin-1α, interleukin-1β, interleukin-6, interleukin-8, and tumor necrosis factor-α were significantly higher in people with normal-weight but with central obesity. Although this suggestion may not be a direct risk factor for mortality among cancer patients, it can take place as an indirect risk factor leading to the worsening of the disease, which can finally cause death. Moreover, studies have reported other possible explanations for the worsening of cancer patients' health by central obesity. Some of these suggestions include dyslipidemia, dysglycemia, lower metabolic rate, lower oxygen consumption, decreased insulin sensitivity, and finally, metabolic syndrome (73,75).
The harms of these unpleasant consequences of central obesity are not restricted to cancer patients. They can also lead to an elevated hazard of mortality among other non-CVD patients as well. Speaking of other non-CVDs, an interesting suggestion regarding the cause of elevated mortality hazard among people is correlated to asthma. It has been suggested that central obesity can cause mechanical mechanisms facilitating the development of asthma symptoms in susceptible patients, such as reduced lung volumes, airway narrowing, reduced functional residual capacity, and reduced lung tidal volume (78,80). Finally, despite these promising results exist, plenty of questions remain. Therefore, further research should be undertaken to investigate other possible mechanisms of the relationship between central obesity and the increased hazard of death in non-CVD patients.
Altogether, these results can provide a convincing justification for the superiority of abdominal obesity measurements, which reflect central obesity, in predicting mortality hazards. Among measures of central obesity, WHtR is proven to be a reliable one (81). It is also proved to be cheaper, easier to measure and calculate, and more sensitive (as an early warning of health risks) than others (77,82). Another advantage of WHtR against other measures of central fatness is that WHtR is independent of age, gender, and ethnicity (83,84). Hence, adding the importance of elevated mortality hazard in central obese people to the suggested superiority of WHtR for predicting central fatness, we were encouraged to examine the association between high levels of WHtR and mortality hazard.
Strengths
Our systematic review has several strengths. First, a comprehensive systematic search was conducted, and there were no restrictions regarding the cause of mortality. Hence, we examined the association between WHtR and all reported mortality causes. Moreover, all the data used for the analysis of this study are extracted from cohort and prospective cohort studies with a good quality score based on the NOS system. Additionally, the research protocol was registered to PROSPERO beforehand and then followed. Finally, on top of all-cause mortality, we reported specified meta-analysis results for the mortality caused by CVDs and cancer.
Limitations
However, the limitations of the chosen references restricted our study for the following reasons: First, articles used for the meta-analysis part of this study considered different categories and cutoffs for classifying WHtR values. Second, although we only imported the adjusted values (full models) of WHtR, the diversity of confounders considered in each reference may be a source of confounding bias. Third, however we stratified our meta-analysis for CVD and cancer causes of mortality, we could not differentiate between various diseases among CVDs or report for some other common diseases like diabetes. Furthermore, there was high heterogeneity among studies which could reduce the reliability of the findings. Finally, data used for this meta-analysis were extracted only from observational studies. Observational studies only demonstrate associations (not causal relationships) and have a lower level of evidence compared to interventional studies.