Abstract
Dear Editor,
I read with interest the recent article by Al-Shami et al. 1 examining metabolic syndrome (MetS) prevalence among Jordanian adults following COVID-19. While the study addresses an important topic, several major methodological and interpretative issues warrant clarification.
First, the study is described as “population-based,” although participants were recruited from outpatient clinics and caregivers in selected hospitals. This represents a facility-based sample rather than a population-based design, and such mislabeling has important implications for interpretation. Closely related, the sampling strategy introduces a strong risk of selection bias, as individuals attending healthcare facilities are more likely to have underlying health conditions or risk factors, potentially inflating prevalence estimates. It is well established that non-representative samples may limit generalizability and bias epidemiological estimates.
Second, there is a serious inconsistency in the inclusion and exclusion criteria. The manuscript indicates inclusion of individuals with chronic conditions such as diabetes, hypertension, and dyslipidemia, while also stating that individuals taking medications for chronic diseases were excluded. Given that medication use is integral to standard diagnostic definitions for MetS components, this contradiction raises concerns about case definition and potential misclassification. 2
Third, the analytical approach does not adequately address the structure of the sample. Despite a clear imbalance in key characteristics, including sex distribution, no weighting or adjustment strategy was applied to improve representativeness. In addition, the handling of missing data is insufficiently detailed, and inconsistencies in reported totals across variables suggest that missingness may be greater than stated. Furthermore, the timing of measurements is unclear, particularly given the use of both direct anthropometric assessments and retrospective clinical records, which may not reflect concurrent metabolic status.
Fourth, interpretation of subgroup findings requires caution. In Table 3 in the original manuscript, the reported percentages appear to reflect the distribution of MetS cases within the sample rather than true within-group prevalence, which may lead to misinterpretation of the magnitude of associations.
Fifth, the comparison of prevalence across years presented in Figure 1 in the original manuscript is problematic. The figure juxtaposes estimates from different studies with varying methodologies, sampling frames, and population structures, yet presents them as temporal trends. Without harmonized data or comparable designs, such comparisons may be misleading.
Finally, there is a persistent inconsistency between the reported stability in crude prevalence and the interpretation of increasing burden based on age-standardized estimates. While standardization facilitates comparisons across populations, it does not establish temporal trends in the absence of comparable samples, and thus, the conclusions should be interpreted more cautiously.
In light of these issues, I recommend that the findings be reframed as a descriptive assessment of MetS burden within a facility-based sample, rather than as evidence of population-level trends following COVID-19. Addressing these concerns would strengthen the methodological rigor and interpretability of the study.
