Sex Differences in Osteomyelitis of the Foot in Persons With Diabetes Mellitus: A Meta-Analysis

Osteomyelitis of the foot is a common complication of infection in persons with diabetes mellitus. Usually starting in a wound and the soft tissues, it often involves cortical bone and/or bone marrow.¹ Approximately 20% of infected foot ulcers in patients with diabetes are complicated by osteomyelitis.² Osteomyelitis is the main risk factor of amputation in patients with diabetic foot disease.³ In a retrospective comparative case series, Wukich et al⁴ showed that the total amputation rate of patients with diabetic foot disease and osteomyelitis was 5.6 times higher than that of patients without osteomyelitis.

Many risk factors of osteomyelitis of the foot in persons with diabetes mellitus have been investigated. Mutluoglu et al,⁵ who performed a retrospective study by reviewing the records of patients with diabetes who were admitted to a medical center with infected foot wounds over a 2-year period, reported that risk factors for osteomyelitis in a multivariate model included the following: length of hospital stay, total duration of antibiotic therapy, duration of intravenous antibiotic therapy, and minor amputation. Patient sex was not a risk factor (male: female odds ratio [OR] = 1.17; 95% confidence interval [CI], 0.45-2.95). The prospective study conducted by Lavery et al,⁶ which included 1666 persons with diabetes and foot wounds who presented to a managed care diabetes disease management program, also showed that male sex was not a risk factor for osteomyelitis (pooled OR = 1.25; 95% CI, 0.63-2.53). In addition, multidrug-resistant organism infection, previous antibiotic therapy, and wound area > 4 cm² were shown to be risk factors regardless of sex in the investigation conducted by Huang et al,⁷ which included medical records of 372 patients with diabetes and infected foot wounds who were admitted to the hospital between January 2011 and December 2014. However, Schwegler et al⁸ conducted a prospective study in 20 patients with diabetic foot ulcers and found a higher incidence of osteomyelitis in men than in women (pooled OR = 2.14; 95% CI, 0.30-15.36). Tang et al³ published a meta-analysis that investigated sex differences in amputation risk in patients with diabetes and found that the risk of amputation in men was about 1.5 times higher than in women (OR = 1.67; 95% CI, 1.307-2.149).

The purpose of this meta-analysis was to examine if sex was a risk factor for developing osteomyelitis in patients with diabetes mellitus and foot disease.

Database and study selection. To identify all studies, electronic databases including the China National Knowledge Infrastructure and PubMed were searched systematically from their inception to May 2020. The search terms were “diabetic foot” and “osteomyelitis.” No limitations were imposed regarding language or publication status. A total of 541 records were identified through PubMed, and 391 through China National Knowledge Infrastructure. Additional records were identified through meeting abstracts (n = 5). The grey literature, which includes conferences and workshop proceedings as well as ongoing trials, also was reviewed (n = 2).

Inclusion criteria were 1) patients with diabetes and foot disease or ulcers, 2) patients diagnosed with osteomyelitis of the foot, and 3) ORs of osteomyelitis risk stratified by sex were provided or could be calculated from the reported variables. Exclusion criteria included 1) osteomyelitis not caused by or related to foot disease or ulcers, 2) preclinical studies, and 3) reviews, case reports, and case series. Full-text reviews were performed by 2 independent reviewers based on these criteria.

Data extraction and qualitative assessment. A self-designed abstraction form was used to retrieve data including first author, publication year, study design, hemoglobin A1C (HbA1c) value (%), and diagnostic criteria of osteomyelitis. All data were independently extracted by 2 reviewers to identify potentially eligible studies. Any discrepancy was resolved through discussion between the reviewers.

The quality of studies was assessed according to the Newcastle–Ottawa Scale (NOS). Selection, comparability, and exposure are the 3 dimensions in the NOS. The NOS score ranges from 0 to 9 stars. A maximum of 1 star can be awarded to each numbered item in the selection and exposure categories; a maximum of 2 stars can be awarded for comparability. Studies scoring 7 or more (corresponding to 78% of the maximum score) were regarded as having good quality.

Statistical analysis. All statistical analyses were performed with Stata15.0 (Stata Corp, College Station, TX). Either a fixed or a random effects model was applied, according to the heterogeneity test, which was explored by inconsistency (I²) statistics (P < .5, for I² value of 50% or more represented substantial heterogeneity). Heterogeneity was judged to be low (< 25%), moderate (25–75%), or high (> 75%). ORs and 95% CIs were estimated with a fixed model or a random effects model and were graphically displayed in forest plots. Subgroup analysis was performed for 2 study designs. Publication bias was investigated based on funnel plots.⁹ Sensitivity analyses were conducted to evaluate the consistency of the primary results after omitting 1 study at a time to examine its influence on the pooled ORs. Meta-regression was conducted to examine the relationships among the incidence of osteomyelitis and OR.¹⁰

Study descriptions. Of the 939 original articles that were retrieved from the databases, the authors identified 9 studies^{1, 5-8, 11-14} that met the inclusion criteria; these 9 studies encompassed 2583 patients. The selection process is shown in Figure 1.¹⁵ Of the 9 studies, 4 were conducted in China and 1 each were conducted in the United States, the United States and the United Kingdom, Hungary, Italy, and Switzerland. Seven (7) studies (89.7% of patients) were retrospective cohort studies, and 2 studies (10.3% of patients) were prospective cohort studies. There were 663 persons (25.7%) who had osteomyelitis secondary to foot ulcers. Characteristics of the studies, such as design, country, participants, diagnostic criteria of osteomyelitis, and diabetes duration, are presented in Table 1. HbA1c (%) of patients was 8.50 to 9.27. Patient age ranged from 58.4 to 71.0 years, with a mean age of 65.2 years.

Meta-analysis of pooled ORs and sensitivity analyses. Nine (9) studies provided enough information for pooled ORs. Data were pooled by the use of the fixed effects model because the heterogeneity test was insignificant (I² = 0.00% in all comparisons). When comparing male to female as the reference group, the pooled OR was 1.14 (95% CI, 0.94-1.38; P =.766), which was not statistically significant (Figure 2). There was no evidence of publication bias based on funnel (Figure 3).

Sensitivity analysis tested the stability of the results. The included studies were removed one by one to observe the change of the combined effect after each study was removed. Pooled estimates and conclusions were consistent with the primary analysis (Figure 4).

The authors compared the incidence of osteomyelitis to ORs by regression analysis, with the graph (t = -0.61, P > .561) suggesting that ORs were not associated with the incidence of osteomyelitis (Figure 5). In addition, ORs were not affected by the year of the research studies, according to the graph of regression analysis (P > 0.203; t = -0.14) (Figure 6).

Risk of bias assessment of included studies. The quality rating of the included studies ranged from 6 to 8 stars on the NOS. All 9 studies had a low risk of bias for the following parameters: population, outcomes measured, follow-up time, and appropriate statistical analysis. All studies had a control group.

This meta-analysis found no sex differences in osteomyelitis in patients with diabetes mellitus and diabetic foot disease, with males and females having an almost equivalent possibility of osteomyelitis (pooled OR = 1.14; 95% CI, 0.94-1.38; P = .766). In the studies, 32.03% of men and 30.0% of women had osteomyelitis.

The results of the current study are consistent with the results of Mutluoglu et al⁵ and Lavery et al.⁶ Mutluoglu et al,⁵ who performed a 2-year retrospective study by reviewing the records of persons with diabetes mellitus hospitalized with a foot ulcer, found that sex was not a risk factor for osteomyelitis. The prospective study conducted by Lavery et al,⁶ which included 1666 persons with diabetes who presented to a managed care diabetes disease management program, also showed that sex differences were not a risk factor for osteomyelitis. In a prospective observation study in 20 patients with diabetes and foot ulcers, Schwegler et al⁸ found that osteomyelitis was more prevalent in male compared with female patients (pooled OR = 2.14; 95% CI, 0.30-15.36).

It is not entirely clear why the risk of osteomyelitis in this population does not differ between males and females. According to Albright and Fleischer,¹⁶ patients who received a preventive foot examination within the past year were 33% less likely to be hospitalized within that year. In 2012, Pscherer et al¹⁷ found that more women made more appointments for preventive care and were more active in self-care practices than men. Laclé and Valero-Juan¹⁸ believed that men had barriers associated with access to health care, specifically men’s reluctance to seek medical care if they did not perceive themselves to be critically ill. However, results of a descriptive cross-sectional study by Hirpha et al¹⁹ found that 73% of patients inspected their feet more than once per day and 19.5% inspected their feet daily, and no significant differences in self-care practices were seen between women and men. Similarly, a cross-sectional study by AlOwais and Shido²⁰ showed that more than 65% of patients were certain about their knowledge regarding foot self-care, and the scores of a foot self-care knowledge test did not differ by participant sex.

Senneville²¹ suggests that the health care system should develop a primary health care strategy to help ensure that patients with diabetes learn good foot self-care. In addition, Meizheng et al²² suggested that clinicians should inform patients to be alert to the occurrence of osteomyelitis and other adverse phenomena, and strive to achieve early detection and timely treatment.²² The International Working Group on the Diabetic Foot also provides a variety of evidence-based prevention suggestions, which can help reduce the risk of foot ulcers.²³ When physical examination findings are combined with results from plain radiographs, magnetic resonance imaging, and serum inflammatory markers,²⁴ both the accuracy and reliability of early stage diagnosis increase.²⁵

There are a number of limitations in this meta-analysis. First, 7 of the 9 studies included were retrospective investigations, which have inherent limitations compared with prospective studies. Second, studies with small sample sizes were included. Third, the 9 studies had different diagnostic criteria for osteomyelitis, because current criteria for diagnosis are not universally agreed upon. This may have affected the results.

A meta-analysis was conducted to examine if sex was a risk factor for developing osteomyelitis in patients with diabetes mellitus and foot disease. Analysis of the data included in 9 studies (2583 patients) did not show a significant difference. Persons with diabetes mellitus are at risk of diabetic foot disease and osteomyelitis and should receive comprehensive follow-up care, including education about foot self-care practices and early detection of complications.

Dr. Chen is a professor and Ms. Zhang, Ms. Wang, Ms. Zhao, and Mr. Li are undergraduate students, Nantong University, Nantong City, Jiangsu Province, China. Address all correspondence to Hong-Lin Chen, PhD, School of Public Health, Nantong University, Nantong City, Jiangsu Province, China; email: 3214460881@qq.com.