The prevalence of type 2 diabetes and prediabetes among armed forces personnel: a systematic review and meta-analysis

Background: Previous studies of diabetes among armed forces personnel did not assess a comprehensive disease prevalence. We conducted a systematic review and meta-analysis to estimate the prevalence of type 2 diabetes and prediabetes among the military personnel and police ocers. Methods: We carried out a systematic search on electronic databases from January 2000 to July 2020. All studies with a report of the prevalence of diabetes and/or prediabetes among armed forces personnel were selected. However, studies with inappropriate population, estimation of incidence, and non-English language were excluded. The Cochran test and I2 statistics were used to determine the heterogeneity of the studies. Random- or xed-effects model, along with subgroup analysis on diabetes or prediabetes as well as the component of armed forces, was carried out to determine the overall prevalence. All analyses were conducted using STATA software. Results: Among 858 citations, 30 studies were entered into the meta-analysis, involving 1879170 individuals. The meta-analysis estimated the prevalence of type 2 diabetes among the military personnel (9.15%; CI: 5.35-12.96) and police ocers (9.58%; 95% condence intervals [CI]: 5.98-13.18). The overall prevalence of prediabetes among the military personnel and the police ocers were 7.32% (CI: 4.22-10.42) and 6.39% (CI: 4.54-8.23), respectively. Conclusion: The prevalence of diabetes and prediabetes among armed forces were consistent with global reports among general population. These results will inform policy-makers and healthcare providers to nd solutions, and help to take action to reduce diabetes risk factors on a wider scale, particularly among armed forces personnel.


Methods
The current study is a systematic review and meta-analysis conducted to estimate the prevalence of prediabetes and diabetes among armed forces personnel in 2020. The study carried out based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [14].

Literature Search
A comprehensive literature search of published studies was conducted using Medline/PubMed, Scopus, ProQuest, and Web of Knowledge, as well as Google Scholar in July 2020. The medical subject headings (MeSH) keywords included "army", "military", "soldiers", "police", "policemen", "cop", "conscript" "diabetes", "prediabetes", "metabolic syndrome", and "prevalence". The keywords dashboard is represented in Supplementary le 1. All articles list were imported to EndNote X7 (Thomson Reuters, Carlsbad, CA, USA) library, and the duplicates were automatically removed. Moreover, the references list of included studies were manually searched to ensure satisfying coverage.

Selection criteria
The primary screening was performed according to the title and abstract, by two independent researchers (MF and SH).
After that, data extraction and quality assessment were done by reading the full-text of the remaining articles. In the case of con ict in any of the secondary stages, the corresponding author (MS) made the nal decision.
The primary inclusion criteria were original studies with a report of the prevalence of diabetes and/or prediabetes among inservice armed forces personnel. The selected studies consist of cross-sectional studies as well as cohort studies reported the prevalence in the baseline. Studies with an inappropriate population (retired personnel, veterans, and outlier ages), estimation of incidence, and non-English language, as well as conference abstracts, poster papers, and editorial, were excluded.

Data extraction
A spreadsheet was prepared using all targeted statistics. This checklist included the name of the rst author, year of publishing, time period, location of study, number of participants, method of diagnosis, and frequency (with lower and upper 95% con dence intervals) for diabetes and prediabetes.

Quality assessment
The Joanna Briggs Institute (JBI) checklist was applied in order to evaluate the quality of the included studies [15]. This appraisal is aimed to estimate the methodological quality and to assess the possibility of bias in the design, conduct, and analysis of studies. The included studies were scored out of nine using this assessment tool, and categorized as satisfactory (score = 7-9), good (score = 4-6), and low (score = 1-3) quality.

Meta-regression analysis
The meta-regression was utilized due to the high level of heterogeneity of the included studies. One of these subgroup analyses was the country's HDI as a relative index of life expectancy and quality as well as literacy and education. The other subgroups consisted of the mean age of participants, sample size, and publication year of study.

Meta-analysis
The Cochran test (with a signi cance level of <0.1) in combination with I 2 statistics (with a signi cance level of >50%) was used to determine the heterogeneity of the studies. Thereafter, the random-effects model with the inverse-variance method was applied in the presence of heterogeneity, and the xed-effects model was used in the absence of heterogeneity. The subgroup analysis was performed in the case of heterogeneity based on the type of disease (diabetes or prediabetes), the component of armed forces personnel (military personnel or police o cers), the countries' HDI, gender, and geographical regions. All analyses were conducted using STATA software (version 13).

Sensitivity analysis
In order to evaluate the quality and consistency of the ndings and to assess the robustness of the obtained results, sensitivity analysis was conducted through deleting each study separately.

Assessing risk of bias
The publication bias was evaluated using the Egger test [16].

Results of the search
The primary stage of searches yielded 858 citations. Additionally, 44 studies also identi ed through gray literature. After removing duplicates, 657 studies selected to investigate the titles and abstracts, and 108 articles remained for the full-text review. Of these, 46 studies ful lled the inclusion criteria. Finally, 30 articles, involving 1879170 individuals, were extracted for the meta-analysis. The reasons for exclusion of 645 papers were irrelevant title (n = 567), studies with inappropriate population (n = 12), not reporting prevalence of either diabetes or prediabetes (n = 11), reporting incidence of either diabetes or prediabetes (n = 2), review or meta-analysis study (n = 6), non-English papers (n = 10), and editorial, conference abstracts, or poster papers (n = 7). The owchart of the included studies is shown in Fig. 1.

Quality assessment
Based on the JBI checklist, included studies categorized as satisfactory (n = 21), good (n = 9), and low (n = 0) quality. The results of the quality assessment are summarized in Table 1 and Supplementary le 1.

Heterogeneity
Considerable inter-study heterogeneity was observed based on the Cochran test and the I 2 index. The results of heterogeneity and the models of analysis are expressed in Table 2.

Results of the meta-regression
The Results of meta-regression demonstrated an association between publication year of study and the prevalence of diabetes among military personnel (Reg Coef = 0.062, p = 0.069). However, this was not signi cantly associated with HDI score (Reg Coef = -8.11, p = 0.441), sample size (Reg Coef = 0.001, p = 0.935), and mean age of the participants (Reg Coef = 0.024, p = 0.879).
On the other hand, the obtained results showed an association between the mean age of the participants and the prevalence of diabetes among police o cers (Reg Coef = 0.064, p = 0.089). Such a meaningful nding was not observed for HDI score (Reg Coef = 2.21, p = 0.182), sample size (Reg Coef = -0.001, p = 0.593), and publication year of study (Reg Coef = -0.001, p = 0.919).

Results of the meta-analysis
Prevalence of diabetes The analysis of pooled data indicated the overall prevalence of diabetes among the military personnel was 9.15% (95%  Table 2.

Sensitivity analysis
The ndings were reanalyzed by deleting each study separately. The results demonstrated the studies with small sample size, and the outlier studies had no signi cant effect on the nal nding.

Publication bias
Results of Egger's tests demonstrated a lack of publication bias (p >0.10).

Discussion
To the best of authors' knowledge, this is the rst meta-analysis conducted on the prevalence of diabetes and prediabetes among armed forces personnel. Several investigations have reported the prevalence of diabetes and prediabetes among different components of armed forces. However, the result of these studies has demonstrated a range of variability in the prevalence. The reported prevalence of diabetes in armed forces was varied from 0.0% to 34.7% or even higher, depending on different diagnostic criteria, age, HDI of countries, special diet, various comorbidities, and other environmental factors. The present meta-analysis also indicated that different types of armed forces might play a signi cant role in the prevalence of dysglycemia.
The obtained results demonstrated that the prevalence of diabetes among the military personnel (9.15%) is relatively lower compared to that of the police o cers (9.58%). The highest reported prevalence of diabetes (>30%) was also related to studies carried out among police o cers [7,17,20,23]. The overall prevalence of diabetes among military personnel, not police o cers, is under the global prevalence of diabetes (9.30%) [44]. This prevalence is uneven since policemen are expected to be in good shape with regular physical activities. Several previous studies conducted among police o cers demonstrated the high prevalence of overweight and obesity made them susceptible to diabetes and heart disease [45,46]. These risk factors of dysglycemia also were associated with a mean age of study population, and elderly participants were more prone to develop diabetes [46].
The most frequent and highly rated stressors might be the other main reasons for the high prevalence of many mental and non-communicable diseases in this occupation group [47,48]. Moreover, ergonomic problems, long work hours, and a particular nutrition regimen should not be underestimated [49,50]. Recent comprehensive studies also showed the prevalence of metabolic syndrome, and its component is high among policemen relative to the general population [9,20].
However, the overall prevalence of diabetes among military personnel was lower than that of other components of armed forces. Several studies reported that diabetes only affected less than two percent of this occupational group [31,35,39]. In addition to a more appropriate lifestyle, there are some other factors contributing to the relatively low prevalence of diabetes in the military unit. Men are initially recruited at younger ages and undergo regular medical checkups, physical training, and calorie-controlled diets. Then, those individuals with underlying medical conditions, including central obesity, are exempted from these services or replaced in public services [9,51]. Hence, some included military personnel in the present investigation might have a better health-related quality of life compared to the police o cers or the general population [51].
The results also showed that the prevalence of diabetes among this population was not associated with the year of study.
There are several other studies that indicated the prevalence of diabetes is increasing, passing the time [52,53]. These are not only related to the increasing trend of physical inactivity and consumption of unhealthy diet but also progression in diagnostic techniques as well as effective strategies for detection and follow-up people with dysglycemia [54]. However, this kind of association was not observed in the present study.
Our ndings demonstrated that the prevalence of prediabetes among the military personnel and the police o cers were 7.32% and 6.39%, respectively. A body of evidence has demonstrated a range of prevalence of dysglycemia in armed forces according to different biochemical criteria. Some studies reported a prevalence of impaired fasting glucose about 30-40%, particularly among police o cers [17,20,34,[55][56][57]. However, the overall prevalence of prediabetes in included studies is under the estimated prevalence of the adult population (7.30%) around the world [58]. However, the fasting blood glucose should be screened regularly among armed forces personnel, as an affordable and invaluable biochemical factor, and impaired fasting glucose would be better to consider as a critical index for early detection and prevention of at-risk population.

Strengths and limitations
The main strength of the current study is that this is the rst analysis that showed the prevalence of diabetes and prediabetes among armed forces personnel. This study can supply reliable baseline information and may guide other scholars to design and conduct novel researches.
However, several limitations re ected by some factors of data are con rmed. Su cient information was not available to conduct separate or subgroup analyses in terms of assessment of all age groups. The lack of data regarding lifestyle, physical activity, nutritional habits, and comorbid factors of the participants, which could explain the estimated prevalence of diabetes, is another main limitation. Because of the high level of heterogeneity of the included studies, it was only attempted to analyze the results according to countries' HDI.

Conclusion
In summary, this meta-analysis clearly demonstrated that the prevalence of diabetes and prediabetes among all types of armed forces personnel was consistent with global reports of these diseases among the general population. These results will inform policy-makers and healthcare providers to nd solutions and help to take action to reduce diabetes risk factors on a wider scale, particularly among armed forces personnel. Future large-scale investigations studying the prevalence of diabetes and its associated factors among armed forces personnel would also help to estimate the more exact prevalence of dysglycemia, and help to take effective strategies in prevention, early detection, and management of diabetes among this occupational group.
Abbreviations CI: 95% con dence interval; IDF: International Diabetes Federation; HDI: human development index; PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses; MeSH: medical subject headings; JBI: Joanna Briggs Institute.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
All data is presented within the manuscript le.

Competing interests
The authors declare that they have no competing interests.

Funding
Not applicable.

Authors' contributions
All authors were responsible for conceiving and designing the protocol. MF, SH, and MS designed the study. MF and MT did the literature search and, together with SH, selected the studies, extracted the relevant data. SH and MS synthesized the data. MF wrote the rst draft of the Manuscript. MS and SA provided critical guidance on the analysis and overall direction of the study. All authors critically revised successive drafts of the paper and approved the nal version. [31] personnel 54 27 Tharkar, 2008 [20]