INTRODUCTION

Background

Metabolic syndrome (MS) can be defined as a cluster of metabolic abnormalities in which an individual presents with at least two of the following conditions: visceral obesity, insulin resistance, dyslipidemia and hypertension. There is increasing evidence that it is a risk factor for type 2 diabetes mellitus, liver cancer and cardiovascular diseases (Kumar, Abbas and Aster, 2013; Hammer and McPhee, 2018). Major factors associated with the distribution of MS are high caloric low fiber diets and sedentary lifestyles. The invention and scale-up of mechanized transportation options has also fueled this phenomenon as well as the ever-increasing prevalence of non-communicable diseases (NCDs). Coronavirus disease 2019 (COVID-19) is an infectious disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) which was first identified in Wuhan, China. Many of the factors associated with MS are also associated with COVID-19 (Saklayen, 2018). In fact it has been suggested that the two health concerns potentially share pathogenic pathways. Systematic reviews use clear, rigorous, structured methods and techniques to assemble and integrate the results of studies that are concerned with a specified topic. They often involve statistical analyses that include meta-analyses, combining p-values, vote counting, calculating range and distribution. Meta-analysis refers to the application of statistical techniques to combine the results of different studies on a specified topic. Though meta-analyses are often performed as a component of systematic analyses, they can also be conducted a stand-alone exercises (Page et al., 2021).

Research question

What was the prevalence of MS among individuals that were diagnosed with COVID-19?

Aim

To estimate the prevalence of MS among individuals that were diagnosed with COVID-19

METHODS

Study design

This was a cross-sectional study that applied a systematic review and meta-analysis (Celentano and Szklo, 2019). The estimate of interest was the prevalence. Though none of the studies reported standard errors of the estimates, these were calculated using the following formula (Bobbitt, 2021):

Where:

P = Proportion experiencing outcome of interest (metabolic syndrome).

n = Sample size = Number of COVID-19 patients enrolled into study.

se = Standard error of the estimate.

Sampling design

The target group were individuals that were diagnosed with COVID-19 after the year 2020.

1.1.1      Inclusion Criteria

Studies that met the flowing criteria were considered for selection into the study:

1. Target population was individuals diagnosed with COVID-19
2. Prevalence reported or can be calculated from study data

• Employed a cohort study design

1.1.2      Exclusion criteria

Studies meeting the following criteria were dropped:

1. Employed a case-control, case-series, experimental or cross-sectional design
2. Study initiated before October 2019

• Unclear study design.

1. Could not access full text.
2. Multiple publications for same population.

Article selection and review process

Several online databases, namely, GoogleScholar, PubMed, ScienceDirect and Research4Life to identify journal articles that satisfied the inclusion criteria. The flow diagram in figure 2.1 illustrates the steps taken in identifying and selecting studies for inclusion in the meta-analysis.

Figure 0.1: Identification and selection process

Statistical analysis

All statistical analysis was performed with Stata version 16 [StataCorp 4905 Lakeway Drive, College Station, TX 77845, USA] and assumed a significance level of 5% (Acock, 2014). The data was declared as meta-analysis data using the “met set” command which utilizes precomputed estimates.

Descriptive statistics

Summary statistics were generated and illustrated in the forest-plots in figures 3.1.

Inferential statistics

Restricted maximum likelihood (REML) estimation was applied, to minimize bias in the estimates, and a random-effects model was used to account for potential clustering in the data (Everitt and Hothorn, 2006; Kirkwood and Sterne, 2003).

RESULTS

Descriptive statistics

Figure 3.1 illustrates how the study-specific prevalence estimates are spread around and the overall prevalence estimate (≈39.97) with their associated confidence intervals. The difference between the calculated overall prevalence estimate (39.97) and zero was statistically significant and therefore unlikely to be the result of chance finding (p-value <0.00001). It is also notable that the test for homogeneity suggests that the selected studies were statistically heterogeneous (p-value <0.00001). The calculated percentage of between study variability (I²) was equal to 100%.

Figure 0.1: Forest plot showing distribution of study-specific prevalence values around the overall prevalence.

Inferential statistics

Table 3.1 displays the results of application of the random-effects modelling technique which indicate that the effect size (prevalence) decreases by about 0.00122 with each unit increase in sample size. However, we cannot rule out chance finding as the result is not significant (p-value = 0.0540).

Table 0.1: Results of random-effects maximum likelihood estimation

Limitations

There are very few published studies that report the prevalence of MS in patients diagnosed with COVID-19 infection. The few studies that have been published apply different study designs which results in methodological heterogeneity. There were also a number of clinically heterogeneous studies that were concerned with the occurrence of MS in the context of diseases other than COVID-19.

CONCLUSION

The estimated prevalence of metabolic syndrome among COVID-19 patients the 10 review studies was quite high at approximately 39.97 (95% CI [26.92, 53.02]). This appears to agree with suggestions that these conditions may have some common pathways in their pathogenesis. However, it should be noted that the number of studies reviewed was quite limited as there are few published studies that report the prevalence of MS in COVID-19 patients. Additionally, methodological heterogeneity may have contributed to the very high statistical heterogeneity (I² = 100.00%).

Though the result was not statistically significant, application of random-effects modelling technique suggested that prevalence decreases with increasing sample size (Coef = -0.00122).

ANNEXES

1. Included Studies List

2. STATA 17 – Codebook Output

REFERENCES

Acock, A. C. (2014) A Gentle Introduction to Stata. 4th ed. College Station: Stata Press. Available

at: papers3://publication/uuid/F353E55B-ED7F-4916-98D6-17CA98E62042.

Bobbitt, Z. (2021) Statology. Available at: https://www.statology.org/ (Accessed: October 24,

2022).

Celentano, D. D. and Szklo, M. (2019) Gordis Epidemiology. 6th ed. Philadelphia: Elsevier, Inc.

Everitt, B. S. and Hothorn, T. (2006) A handbook of statistical analyses using r. Boca Raton:

Chapman & Hall/CRC.

Hammer, G. D. and McPhee, S. J. (eds.) (2018) Pathophysiology of disease: an introduction to

clinical medicine. 8th ed. New York: McGraw Hill Education.

Kirkwood, B. R. and Sterne, J. A. C. (2003) Essential Medical Statistics. 2nd ed. Oxford:

Blackwell Science Ltd.

Kumar, V., Abbas, A. K. and Aster, J. C. (2013) Robbins Basic Pathology. 9th ed. Edited by V. Kumar, A. K. Abbas, and J. C. Aster. Philadelhia: Elsevier Saunders.

Page, M. J. et al. (2021) “The PRISMA 2020 statement: an updated guideline for reporting

systematic reviews,” BMJ, 372(71). doi: 10.1136/BMJ.N71.

Saklayen, M. G. (2018) “The Global Epidemic of the Metabolic Syndrome,” Current Hypertension

Reports, 20(12). doi: 10.1007/S11906-018-0812-Z.