Abstract
Background.
Invasive group A Streptococcus (iGAS) infections cause significant morbidity and mortality worldwide. We analyzed whether obesity and diabetes were associated with iGAS infections and worse outcomes among an adult US population.
Methods.
We determined the incidence of iGAS infections using 2010–2012 cases in adults aged ≥18 years from Active Bacterial Core surveillance (ABCs), a population-based surveillance system, as the numerator. For the denominator, we used ABCs catchment area population estimates from the 2011 to 2012 Behavioral Risk Factor Surveillance System (BRFSS) survey. The relative risk (RR) of iGAS was determined by obesity and diabetes status after adjusting for age group, gender, race, and other underlying conditions through binomial logistic regression. Multivariable logistic regression was used to determine whether obesity or diabetes was associated with increased odds of death due to iGAS compared to normal weight and nondiabetic patients, respectively.
Results.
Between 2010 and 2012, 2927 iGAS cases were identified. Diabetes was associated with an increased risk of iGAS in all racial groups (adjusted risk ratio [aRR] ranged from 2.71 to 5.08). Grade 3 obesity (body mass index [BMI] ≥40) was associated with an increased risk of iGAS for whites (aRR = 3.47; 95% confidence interval [CI], 3.00–4.01). Grades 1–2 (BMI = 30.0–<40.0) and grade 3 obesity were associated with an increased odds of death (odds ratio [OR] = 1.55, [95% CI, 1.05, 2.29] and OR = 1.62 [95% CI, 1.01, 2.61], respectively) when compared to normal weight patients.
Conclusions.
These results may help target vaccines against GAS that are currently under development. Efforts to develop enhanced treatment regimens for iGAS may improve prognoses for obese patients.
Keywords: group A Streptococcus infections, obesity, diabetes
Invasive group A Streptococcus (iGAS) infections cause significant morbidity and mortality in the form of severe skin and soft tissue infections (SSTIs), bacteremic pneumonia, necrotizing fasciitis (NF), streptococcal toxic shock syndrome (STSS), bloodstream infections, and sepsis [1]. In 2013, an estimated 11 500 cases and 1100 deaths due to iGAS infections occurred in the United States [2]. Vaccines against invasive and noninvasive GAS infections are under development [3, 4], but currently there are limited strategies for disease prevention. Identifying factors that put people at risk for iGAS or impact their clinical prognosis is important for clinicians who must promptly recognize and treat these severe infections and for public health officials working on prevention efforts.
The proportion of obese adults in the United States increased substantially during the 1980s and 1990s and rose slightly in the last decade [5, 6]. In 2012, approximately 35% of adults were considered obese [7]. The proportion of adults diagnosed with diabetes has risen steadily from approximately 3% to approximately 8% between the 1980s and 2012 [8]. Obesity and diabetes have been linked to increased risk of surgical-site and other SSTIs, respiratory infections, bloodstream infections, and sepsis—all of which are common manifestations of iGAS [9–21]. Because obesity is a risk factor for diabetes [22], a common mechanism may at least partially explain why both are associated with an increased risk of these types of infections. There have been no known published reports looking at the impact of both obesity and diabetes on the incidence and outcomes of severe GAS infections. We used population-based surveillance for iGAS that includes individual-level data on underlying conditions in conjunction with population-based estimates of the frequency of obesity and diabetes to calculate incidence rates and to analyze whether obesity and diabetes are risk factors for iGAS infections and poor outcomes.
METHODS
Surveillance Population
We identified nonpregnant adults (≥18 years old) with iGAS infections between 1 January 2010 and 31 December 2012 through the Active Bacterial Core surveillance (ABCs) system. ABCs, a part of the Centers for Disease Control and Prevention’s (CDC) Emerging Infections Program (EIP), conducts active, laboratory- and population-based surveillance for iGAS at 10 sites located throughout the United States, covering approximately 32 million persons (approximately 10% of the US population). ABCs methods have been previously described [23]. Briefly, ABCs staff routinely contacts all microbiology laboratories that serve patients residing in the ABCs catchment area. A case is defined as isolation of GAS from a normally sterile site or from a wound in a patient with NF or STSS who is also a resident of the surveillance area. The ABCs areas included in this analysis were: California (3 county San Francisco Bay area); Colorado (5 county Denver area); Connecticut; Georgia (20 county Atlanta area); Maryland (6 county metropolitan Baltimore area); Minnesota; New Mexico; New York (15 county Rochester and Albany areas for the years 2011 and 2012 only); Oregon (3 county Portland area); and Tennessee (20 urban counties). Patient demographics (eg, age, gender, race), height, weight, and clinical data (eg, underlying conditions, infection type, outcome) were obtained from medical record reviews based on a standardized case report form. Race groups were divided into white, black, and other (includes American Indian/Alaska Natives and Asian/Other Pacific Islanders) races. We could not further categorize persons into Hispanic and non-Hispanic ethnicity because of a high degree of missing data. All available isolates were collected and emm typed at the CDC’s Streptococcus Laboratory [24].
Determination of Obesity and Diabetes Status for iGAS Cases
We used body mass index (BMI) to determine obesity status for iGAS patients. For cases that had height and weight recorded in the medical record, BMI was calculated by the standard formula (BMI=weight (kilograms)/height (meters2)). For each patient missing height and weight or whose calculated BMI was thought to be implausible (≤12 or >100), BMI was imputed 30 times using a regression model that included demographics (age, race, gender, insurance status), year, location, presence of underlying conditions, clinical syndrome, and height (if available) and weight (if available). Thirty imputations were selected based on the average amount of missing data [25, 26]. Patients were categorized into standard BMI categories: underweight (BMI < 18.5), normal weight (BMI 18.5–<25.0), overweight (25.0–<30.0), grades 1–2 obesity (BMI: 30.0–<40.0), and grade 3 obesity (BMI: ≥40.0) [27]. A person was considered to have diabetes if it was recorded in the medical record by a healthcare provider. All types of diabetes were included except gestational diabetes.
Incidence and Relative Risk Calculation
We estimated the incidence of iGAS for demographic groups (age, race, gender) and underlying conditions of interest (obesity status, diabetes, and heart disease) using ABCs cases as numerators and ABCs catchment area population estimates from the Behavioral Risk Factor Surveillance System (BRFSS) surveys as denominators. BRFSS [28] is a cross-sectional cellular and landline telephone survey of noninstitutionalized adults administered by state health departments using a standardized questionnaire developed in conjunction with the CDC. Data collected include self-reported height and weight and self-reported chronic health conditions such as diabetes. The survey is designed to determine the state-specific prevalence of certain health conditions and risk behaviors. We used small area estimation for BRFSS denominators by a method previously described to determine the prevalence of health conditions in the ABCs catchment area which includes entire states or only certain counties within states as delineated above [29]. To account for bias known to exist from self-reported heights and weights, we derived ratios that compared national estimates of calculated BMIs from the 2011 to 2012 BRFSS self-reported heights and weights to calculated BMIs from the 2011 to 2012 National Health and Nutrition Examination Survey (NHANES), which obtains measured heights and weights, stratified by age group (18–49 year olds, 50–64 year olds, and 65 years and older), race (white, black, and other) and gender [30]. The ratios were used to adjust individual BMIs used in the denominators of the incidence estimates. We used multilevel binomial logistic regression with county-level random effects to estimate the relative risk (RR) with 95% confidence intervals (CIs) of iGAS by BMI category and diabetes status after controlling for gender, age group, county, the presence of heart disease, and the presence of one or more other underlying conditions thought to be risk factors for iGAS (includes asthma/chronic obstructive pulmonary disease [COPD], current smoking, cancer, chronic kidney disease, and stroke). In sum, 95% CIs that did not include 1.0 were considered significant relative risks. Data were stratified by race to account for differences across racial groups. The RR was approximated from the odd ratios (ORs) because iGAS is a relatively rare disease.
Analysis of Outcomes
Multivariable logistic regression was used to calculate ORs with 95% CIs for intensive care unit (ICU) admission and death by BMI category and diabetes status. Besides BMI category and diabetes status, other factors included in the model were age group, gender, race, state of residence, presence of specific underlying conditions (skin conditions, heart disease, current smoker, asthma/COPD, immunosuppression, malignancy and alcohol abuse), emm type and syndrome (eg, SSTI, pneumonia, NF, meningitis, STSS/septic shock). SSTIs included cellulitis, erysipelas, wound infections, phlebitis, lymphangitis, lymphadenitis, and gangrene but excluded NF. Patients who had multiple syndromes were classified into the syndrome (“primary syndrome”) with the highest case fatality. Death was considered GAS-associated if it occurred during the hospitalization for iGAS. If the outcome status was unknown during initial medical record review, site staff reviewed state vital records to determine if the patient died during their hospitalization for iGAS. Factors associated (P < .20) with obesity and diabetes and that were associated with the severe outcomes (ICU admission or death) were included in the initial multivariable logistic regression model. Factors were dropped from the model if they were not significantly (P < .05) associated with the outcomes and did not significantly change the results for the main factors being evaluated (ie, obesity and diabetes). Two-way interactions were also considered. The 95% CIs that did not include 1.0 were considered significant.
Human Subject Considerations
These activities were considered part of public health surveillance and determined to be “non-research” by CDC’s Institutional Review Board (IRB). Where required, IRB approval was obtained at ABCs site health departments or academic institutions.
RESULTS
Between 2010 and 2012, 2927 iGAS cases were identified through ABCs. The median age of patients was 56 years, and they were fairly equally distributed among the age groups (Table 1). A majority of patients (55%) were male and most (76%) were white. BMI was multiply imputed for 516 (18%) patients. Thirty-nine percent of patients were obese (grades 1–3) with about one-third of those being extremely (grade 3) obese (Table 1). Most (87%) patients had at least one underlying condition—the most common being diabetes, skin conditions, current smoking, and heart disease. The most common primary syndromes associated with iGAS were SSTIs, bacteremia without a focus, STSS/septic shock, and pneumonia. Isolates were available for emm typing from approximately 76% of patients (Table 1).
Table 1.
Characteristics of Patients With Invasive Group A Streptococcus Infection: Active Bacterial Core surveillance (2010–2012)
| Characteristic | Number (%) N = 2927 |
|---|---|
| Age group in years: | |
| 18–49 | 1102 (37.6) |
| 50–64 | 852 (29.1) |
| ≥65 | 973 (33.2) |
| Male gender | 1611 (55.0) |
| Race | |
| White | 2222 (75.9) |
| Black | 432 (14.8) |
| Other racea | 273 (9.3) |
| Body mass index (BMI) category | |
| Underweight (<18.5) | 131 (4.5) |
| Normal weight (18.5–<25.0) | 836 (28.6) |
| Overweight (25.0–<30.0) | 818 (27.9) |
| Obesity grade 1–2 (30.0–<40.0) | 743 (25.4) |
| Obesity grade 3 (≥40) | 399 (13.6) |
| Underlying conditions | |
| Diabetes | 859 (29.3) |
| Skin conditionsb | 836 (28.6) |
| Current smoker | 548 (18.7) |
| Heart disease | 544 (18.6) |
| Chronic renal disease/dialysis | 375 (12.8) |
| Immunosuppressionc | 310 (10.6) |
| Malignancyd | 303 (10.4) |
| Alcohol abuse | 277 (9.5) |
| Chronic obstructive pulmonary disease (COPD)/asthma | 272 (9.3) |
| Any underlying conditione | 2539 (86.7) |
| Primary syndromef | |
| Skin/soft tissue infectiong | 976 (33.3) |
| Bacteremia without a focus | 629 (21.5) |
| STSS/septic shock | 533 (18.2) |
| Pneumonia | 352 (12.0) |
| Necrotizing fasciitis | 123 (4.2) |
| Meningitis/other central nervous system infection | 29 (1.0) |
| Other syndrome | 314 (10.7) |
| emm typesh | |
| emm1 | 480 (16.4) |
| emm89 | 222 (7.6) |
| emm12 | 195 (6.7) |
| emm28 | 140 (4.8) |
| emm3 | 139 (4.7) |
| Other emm types | 1062 (36.3) |
| emm type unknown | 689 (23.5) |
Abbreviation: STSS, streptococcal toxic shock syndrome.
Includes American Indian/Alaska Natives (6.0%) and Asian/Other Pacific Islanders (3.3%).
Skin condition includes chronic skin breakdown or recent (within 7 days of culture) varicella infection, penetrating or blunt trauma, surgical wound or burn.
Includes patients with immunoglobulin deficiencies, nephrotic syndrome, organ transplantation, receipt of immunosuppression therapy, AIDS and asplenia.
Excludes malignancies of the skin.
Includes conditions listed and other underlying conditions.
Patients with more than one syndrome were categorized into the most severe one based on the likelihood of death for patients with a single syndrome.
Skin/soft tissue infection includes cellulitis, erysipelas, wound infections, phlebitis, lymphangitis, lymphadenitis and gangrene but excludes necrotizing fasciitis.
Five most common emm types are listed individually.
Approximately 24.8 million adults resided in the catchment area. The crude incidence of iGAS was higher in grade 3 obese patients compared to normal and overweight patients across all race groups (Table 2). The highest incidence of iGAS occurred in underweight patients. Patients with diabetes had higher incidences of iGAS compared to nondiabetic patients overall and across all race groups (Table 2).
Table 2.
Incidence (per 100 000) of Invasive Group A Streptococcus Infection by Age, Gender, Body Mass Index Category, Underlying Conditions and Race: Active Bacterial Core Surveillance (2010–2012)
| Characteristic | Incidence for Whites (95% CI) | Incidence for Blacks (95% CI) | Incidence for Other Racea (95% CI) | Incidence for All Races (95% CI) |
|---|---|---|---|---|
| Age group in years | ||||
| 18–49 | 2.4 (2.2, 2.5) | 3.2 (2.8, 3.7) | 2.6 (2.2, 3.1) | 2.5 (2.4, 2.7) |
| 50–64 | 4.3 (3.9, 4.6) | 5.2 (4.4, 6.2) | 5.6 (4.5, 7.0) | 4.5 (4.2, 4.8) |
| ≥65 | 8.2 (7.7, 8.8) | 6.8 (5.4, 8.5) | 8.7 (6.8, 11.1) | 8.1 (7.6, 8.6) |
| Gender | ||||
| Female | 3.4 (3.2, 3.6) | 3.6 (3.1, 4.1) | 3.4 (2.9, 4.1) | 3.4 (3.2, 3.6) |
| Male | 4.5 (4.2, 4.7) | 4.6 (4.1, 5.3) | 4.4 (3.7, 5.1) | 4.5 (4.3, 4.7) |
| Body mass index category | ||||
| Normal weight (18.5–<25.0) | 3.4 (3.2, 3.7) | 4.7 (3.9, 5.7) | 2.1 (1.7, 2.7) | 3.4 (3.2, 3.6) |
| Underweight (<18.5) | 17.3 (14.1, 21.2) | 37.8 (26.4, 54.1) | 3.3 (1.6, 6.9) | 15.7 (13.2, 18.6) |
| Overweight (25.0–<30.0) | 3.1 (2.9, 3.4) | 2.9 (2.4, 3.6) | 3.6 (2.9, 4.6) | 3.1 (2.9, 3.4) |
| Obese grade 1–2 (30.0–<40.0) | 3.8 (3.5, 4.2) | 3.3 (2.7, 3.9) | 7.0 (5.6, 8.7) | 3.9 (3.6, 4.2) |
| Obese grade 3 (≥40) | 11.4 (10.1, 12.7) | 6.3 (4.9, 8.0) | 18.7 (13.7, 25.5) | 10.4 (9.4, 11.5) |
| Diabetes | ||||
| Absent | 3.1 (2.9, 3.2) | 3.3 (2.9, 3.6) | 2.5 (2.2, 2.9) | 3.0 (2.9, 3.2) |
| Present | 13.8 (12.8, 15.0) | 9.8 (8.2, 11.6) | 19.9 (16.5, 24.0) | 13.5 (12.7, 14.5) |
| Heart disease | ||||
| Absent | 3.6 (3.4, 3.8) | 3.7 (3.4, 4.1) | 3.5 (3.1, 4.0) | 3.6 (3.5, 3.8) |
| Present | 8.8 (7.9, 9.9) | 9.7 (7.5, 12.7) | 12.2 (8.7, 17.2) | 9.2 (8.3, 10.1) |
Abbreviation: CI, confidence interval.
Other race includes American Indian/Alaska Natives (6.0%) and Asian/Other Pacific Islanders (3.3%).
After adjusting for age, gender, obesity status, heart disease, and the presence of one or more other underlying conditions, the RR of iGAS was significantly higher in diabetic patients than nondiabetic patients with the point estimate ranging from 2.71 in blacks, 3.37 in whites, and 5.08 in other races (Table 3). Grade 3 obesity compared to normal weight was associated with an increased risk of iGAS in whites and other races. Rates of iGAS were lower for overweight persons compared to normal weight persons (in both whites and blacks). Being underweight compared to being normal weight was associated with a higher incidence of iGAS for both whites and blacks but not for other races. Age ≥65 years (except for blacks), male gender, and heart disease (except for other races) were also associated with higher rates of iGAS infections compared to younger ages (18–49 years), being female, and the absence of heart disease, respectively (Table 3).
Table 3.
Adjusted Relative Risk of Invasive Group A Streptococcus Infection by Age, Gender, Body Mass Index Category, Underlying Conditions and Race: Active Bacterial Core Surveillance (2010–2012)
| Characteristic | Relative Riska for Whites (95% CI) | Relative Riska for Blacks (95% CI) | Relative Riska for Other Raceb (95% CI) |
|---|---|---|---|
| Age group | |||
| 18–49 years-old | reference | Reference | reference |
| 50–64 years-old | 1.52 (1.36,1.70) | 1.21 (.95,1.54) | 1.39 (1.02,1.89) |
| ≥65 years-old | 2.69 (2.41,3.01) | 1.29 (.95,1.75) | 1.86 (1.30,2.65) |
| Gender | |||
| Female | reference | reference | reference |
| Male | 1.41 (1.29,1.54) | 1.33 (1.10,1.62) | 1.35 (1.05,1.74) |
| Body mass index category | |||
| Normal weight (18.5–<25.0) | reference | reference | reference |
| Underweight (<18.5) | 4.82 (3.73,6.24) | 7.32 (4.67,11.5) | 1.57 (.64,3.86) |
| Overweight (25.0–<30.0) | 0.77 (.68,.88) | 0.60 (.45,.81) | 1.23 (.86,1.77) |
| Obese grade 1–2 (30.0–<40.0) | 1.02 (.90,1.16) | 0.65 (.49,.86) | 2.02 (1.41,2.89) |
| Obese grade 3 (≥40) | 3.47 (3.00,4.01) | 1.29 (.93,1.79) | 4.34 (2.76,6.81) |
| Diabetes | |||
| Absent | reference | reference | reference |
| Present | 3.37 (3.05,3.71) | 2.71 (2.15,3.42) | 5.08 (3.81,6.77) |
| Heart disease | |||
| Absent | reference | reference | reference |
| Present | 1.50 (1.32,1.71) | 1.99 (1.45,2.72) | 1.29 (.84,1.98) |
| Any other chronic conditionc | |||
| Absent | reference | reference | reference |
| Present | 1.05 (.96,1.14) | 0.91 (.74,1.1) | 0.66 (.5,.87) |
Abbreviations: BMI, body mass index; CI, confidence interval.
Relative risk is adjusted for all other characteristics listed.
Other race includes American Indian/Alaska Natives (6.0%) and Asian/Other Pacific Islanders (3.3%).
Other chronic conditions include asthma/chronic obstructive pulmonary disease, current smoking, cancer (excluding nonmelanoma skin cancer) chronic kidney disease and stroke, and exclude obesity (BMI ≥ 30), diabetes and heart disease.
On univariate analysis, factors associated (P < .20) with death included older age group (>50 years), state of residence, higher BMI category, heart disease, chronic renal disease/dialysis, malignancy, alcohol abuse, or STSS/septic shock as was having emm type 1 or 3. Factors associated (P < .20) with ICU admission included all those associated with death as well as having COPD/asthma, pneumonia or NF.
On multivariable analysis, grades 1–2 obesity and grade 3 obesity were associated with an increased odds of death compared with normal weight (Table 4). Diabetes was not associated with an increased odds of death. There was no interaction between diabetes and obesity and death in the multivariable model. The same percentage (9%) of normal weight patients both with and without diabetes died (P = .93), and almost the same percentage of obese patients both with (11%) and without diabetes (12%) died (P = .17). Other factors that were independently associated with death included older age, heart disease, malignancy, and alcohol abuse. Among the primary syndromes, the odds of death was elevated among patients who had STSS/septic shock compared to those without those syndromes. Patients with SSTIs had a reduced odds of death compared to those who did not have SSTIs. When subset to patients with SSTIs, patients with grade 3 obesity had an increased odds of death compared to normal weight patients (OR = 9.21; 95% CI, 1.06–79.64), but diabetic patients did not have an increased odds of death compared to nondiabetic patients (OR = 0.58, 95%CI, .17–1.94). There was no increase odds of death based on obesity or diabetes status for those who had syndromes other than SSTIs.
Table 4.
Adjusted Odds Ratios for Characteristics Associated With Intensive Care Unit (ICU) Admission and Death for Patients with invasive Group A Streptococcus infection: Active Bacterial Core surveillance (2010–2012)
| Characteristic | Adjusted ORa for ICU Admission (95% CI) | Adjusted ORa for Death (95% CI) |
|---|---|---|
| Age group | ||
| 18–49 years-old | reference | reference |
| 50–64 years-old | 1.27 (1.01,1.60) | 2.02 (1.41,2.88) |
| ≥65 years-old | 1.31 (1.05,1.63) | 3.38 (2.36,4.83) |
| Race | ||
| White | reference | reference |
| Black | 1.27 (.97,1.66) | 0.84 (.56,1.26) |
| Other | 0.90 (.63,1.29) | 0.95 (.56,1.62) |
| BMI category | ||
| Normal weight (18.5–<25.0) | reference | reference |
| Underweight (<18.5) | 1.10 (.68,1.78) | 1.63 (.84,3.15) |
| Overweight (25.0–<30.0) | 1.36 (1.05,1.75) | 1.38 (.94,2.03) |
| Obese grade 1–2 (30.0–<40.0) | 1.46 (1.12,1.90) | 1.55 (1.05,2.29) |
| Obese grade 3 (≥40) | 2.07 (1.50,2.86) | 1.62 (1.01,2.61) |
| Other conditions | Reference: absence of condition | Reference: absence of condition |
| Diabetes | 1.06 (.86,1.30) | 0.77 (.57,1.04) |
| Heart disease | 1.29 (1.02,1.64) | 1.50 (1.10,2.04) |
| Malignancy | b | 1.50 (1.05,2.17) |
| Alcohol abuse | 2.34 (1.73,3.16) | 1.94 (1.28,2.96) |
| Current smoker | 1.24 (.98,1.56) | 0.65 (.44,.98) |
| Primary syndromec | Reference: absence of syndrome | Reference: absence of syndrome |
| Skin/soft tissue infectiond | 0.53 (.41,.64) | 0.26 (.17,.40) |
| Pneumonia | 1.95 (1.51,2.54) | b |
| Necrotizing fasciitis | 4.13 (2.75,6.19) | b |
| STSS/shock | 17.03 (12.57,23.08) | 4.82 (3.67,6.35) |
Abbreviations: BMI, body mass index; CI, confidence interval; ICU, intensive care unit; OR, odds ratio; STSS, streptococcal toxic shock syndrome.
The models are adjusted for the variables listed unless otherwise indicated and are also adjusted for state of residency.
These factors were not included in the model because they were not significant (P > .20) on univariable analysis for the outcome (ICU admission or death) or did not change the results for the primary factors being considered (BMI category or diabetes status) and were not significant in multivariable analysis (95% confidence limits of the odds ratio included 1.0).
Persons with more than one syndrome were categorized into the most severe type based on the likelihood of death for persons with a syndrome.
Skin/soft tissue infection includes cellulitis, erysipelas, wound infections, phlebitis, lymphangitis, lymphadenitis and gangrene but excludes necrotizing fasciitis.
The odds of ICU admission was higher in obese patients compared to those with normal weight. However, there was no appreciable difference in the mean or median length of stay between patients who had grade 1–2 obesity (mean 10.6 days, median 7 days), grade 3 obesity (mean 10.8 days, median 7 days), were underweight (mean 11.0 days, median 7 days), or were normal weight (mean 10.3, median 8 days). Diabetes was not associated with an increased odds of ICU admission, and length of stays were similar between diabetic (mean 10.9 days, median 8 days) and nondiabetic patients (10.2 and median 7 days). There was no interaction between diabetes and obesity and ICU admission in the multivariable model. About the same percentage of normal weight patients with (38%) and without (37%) diabetes required ICU (P = .16).
The distribution of primary syndromes associated with iGAS infections did not appreciably differ by BMI category or diabetes status, with one important exception. A higher proportion of patients with grade 3 obesity (45.3%) had SSTIs compared to normal weight patients (29.2%, P < .001) as did diabetic patients (37.3%) compared to nondiabetic patients (31.1%, P < .001) (Figure 1). There was a higher odds of SSTIs among grade 3 obese compared to normal weight patients among whites (OR = 1.83; 95% CI, 1.38–2.42), blacks (OR = 2.36; 95% CI, 1.89–4.71), and patients of other races (OR = 4.62, 95% CI, 2.08–10.56). The odds of SSTI was higher among diabetic compared to nondiabetic patients for whites (OR = 1.28, 95% CI, 1.05–1.55). The relationship between SSTIs and diabetes was not statistically significant among blacks (OR = 1.16, 95% CI, .72–1.9) or other races (OR = 1.53, 95% CI, .92–2.56).
Figure 1.
Distribution of select syndromes* among invasive group A Streptococcus patients, by body mass index category and diabetes status: Active Bacterial Core surveillance (2010–2012). *Persons with more than one syndrome were categorized into the most severe type based on the likelihood of death for persons with a single syndrome. **Skin/soft tissue infection (SSTI) includes cellulitis, erysipelas, wound infections, phlebitis, lymphangitis, lymphadenitis and gangrene but excludes necrotizing fasciitis. Abbreviations: STSS, streptococcal toxic shock syndrome; SSTI, skin and soft tissue infection.
DISCUSSION
We found diabetes to be a risk factor for iGAS infections among all race groups and extreme obesity to be a risk factor for iGAS infections among whites and other races. Obesity was also associated with an increased risk of ICU admission and death, although there were no associations between diabetes and poorer outcomes. Understanding who is at increased risk for iGAS infections and poorer outcomes has implications for recognizing severe infections, administering timely and appropriate treatments, and targeting public health prevention efforts.
Obesity and diabetes have been previously shown to be risk factors for syndromes often associated with iGAS, including SSTIs [9–11, 14–15, 16, 18, 20–21], pneumonia [12, 17, 21], bloodstream infections, and sepsis [13, 19, 31]. A case-control study from the late 1990s did show diabetes to be a risk factor for iGAS (relative risk = 2.27, P = .03), but the study did not assess obesity as a potential risk factor [32]. Our results showing an increased risk of iGAS among extremely obese whites and other races and diabetic patients across all races are consistent with these findings. SSTI was the only primary syndrome that was positively associated with obesity and diabetes, although the relationship between diabetes and SSTIs varied by race. Likewise, the increased risk of death was seen in obese patients with SSTIs but was not present for other syndromes. Differences in the occurrence of SSTIs seem to be driving much of the increased risk of iGAS and poorer outcomes. We did not find an overall association between extreme obesity and the incidence of iGAS infections among blacks. Biologic differences in body composition may vary across races and ethnicities, so BMI may not be measuring the same level of adiposity across the different demographic groups [33]. Mechanisms thought to increase an obese person’s susceptibility to severe infections include alterations in immune system functions, impaired skin barrier functions and altered lung mechanics [34–35]. Likewise, diabetes may alter immune system function and wound healing beyond the effects of obesity [36].
Although increases in the adult prevalence of obesity (from approximately 31% to approximately 35%) and diabetes (from approximately 5% to approximately 8%) in the United States have been noted over the past 17 years [6–8], there has been no significant increase in the overall incidence of iGAS as detected by ABCs during the same time period [37]. The prevalence of grade 3 obesity has doubled from 3% to 6%. Absolute changes in the proportions of adults with these conditions were still relatively small, so a major impact on overall iGAS rates is not expected.
This analysis has some limitations. First, we imputed BMI for approximately 18% of cases, but values were based on predictive variables in a regression analysis. BRFSS uses self-reported heights and weights, and we attempted to adjust for known biases in self-reporting. It is possible that both obese and nonobese persons had undiagnosed diabetes, but we found an increased risk among the extremely obese even after controlling for patients whose diabetes was diagnosed. However, we were only able to measure the direct impact of obesity on iGAS infections because we controlled for diabetes in the model. We did not measure the indirect effect that obesity has on iGAS through its association with diabetes [22]. We did not collect measures of glucose control, like hemoglobin A1C, which may affect a diabetic person’s risk of iGAS. Although we found obesity (except in blacks) and diabetes to be risk factors for iGAS after controlling for other factors, it is possible that we did not control for other unknown factors. Finally, the ABCs catchment area includes a more urban population than the United States as a whole, so the findings may not be representative of the entire US population.
There are few strategies for preventing iGAS infections or their severe outcomes. Current approaches include prevention of secondary cases in healthcare and household settings [38]. Curbing obesity and diabetes has been a major focus of the medical and public health communities for decades, and reductions in the prevalence of these conditions may help reduce overall rates of iGAS. Additionally, vaccines against GAS are currently under development [3–4]. These results, if confirmed, may help target vaccines to patients with diabetes or obesity should they become available. However, one also must consider that obese persons may have reduced responses to vaccinations [39]. More needs to be done to understand how obesity may impact outcomes—particularly whether different treatment regimens, such as altered antibiotic dosing, would improve prognoses.
Acknowledgments.
We acknowledge the following persons for their efforts in collecting and compiling ABCs data. Colorado Emerging Infections Program: Deborah Aragon, Shaun Cosgrove, Ken Gershman, Jennifer Sadlowski, Benjamin White. Connecticut Emerging Infections Program: Carmen Marquez, Michelle Wilson. Georgia Emerging Infections Program: Wendy Baughman, Stephanie Thomas, Amy Tunali. New Mexico Emerging Infections Program: Megin Nichols, Joseph Bareta, Karen Scherzinger, Kathy Angeles, Lisa Butler, Sarah Khanlian, Robert Mansmann. Oregon Emerging Infections Program: Jamie Thompson. Tennessee Emerging Infections Program: Brenda Barnes.
Footnotes
Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.O’Loughlin RE, Roberson A, Cieslak PR, et al. The epidemiology of invasive group A streptococcal infection and potential vaccine implications: United States, 2000–2004. Clin Infect Dis 2007; 45:853–62. [DOI] [PubMed] [Google Scholar]
- 2.CDC. Group A Streptococcus. CDC Website; 2013. Available at: http://www.cdc.gov/abcs/reports-findings/surv-reports.html. Accessed 2014. [Google Scholar]
- 3.Moreland NJ, Waddington CS, Williamson DA, et al. Working towards a group A streptococcal vaccine: report of a collaborative Trans-Tasman workshop. Vaccine 2014; 32:3713–20. [DOI] [PubMed] [Google Scholar]
- 4.Song Y, Zhang X, Lu C, Zhang F, Zhu H. Progress in development of Group A Streptococcus vaccines. Curr Pharm Biotechnol 2014; 14:947–50. [DOI] [PubMed] [Google Scholar]
- 5.Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA 2010; 303:235–41. [DOI] [PubMed] [Google Scholar]
- 6.Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999–2000. JAMA 2002; 288:1723–7. [DOI] [PubMed] [Google Scholar]
- 7.Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011–2012. JAMA 2014; 311:806–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Geiss LS, Wang J, Cheng Y, et al. Prevalence and Incidence Trends for Diagnosed Diabetes Among Adults Aged 20 to 79 Years, United States, 1980–2012. JAMA 2014; 312:1218–26. [DOI] [PubMed] [Google Scholar]
- 9.Jain RK, Shukla R, Singh P, Kumar R. Epidemiology and risk factors for surgical site infections in patients requiring orthopedic surgery. Eur J Orthop Surg Traumatol 2014; 8:8. [DOI] [PubMed] [Google Scholar]
- 10.Karppelin M, Siljander T, Vuopio-Varkila J, et al. Factors predisposing to acute and recurrent bacterial non-necrotizing cellulitis in hospitalized patients: a pro-spective case-control study. Clin Microbiol Infect 2010; 16:729–34. [DOI] [PubMed] [Google Scholar]
- 11.Korol E, Johnston K, Waser N, et al. A systematic review of risk factors associated with surgical site infections among surgical patients. PLoS One 2013; 8:e83743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kwong JC, Campitelli MA, Rosella LC. Obesity and respiratory hospitalizations during influenza seasons in Ontario, Canada: a cohort study. Clin Infect Dis 2011; 53:413–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Michalia M, Kompoti M, Koutsikou A, et al. Diabetes mellitus is an independent risk factor for ICU-acquired bloodstream infections. Intensive Care Med 2009; 35:448–54. [DOI] [PubMed] [Google Scholar]
- 14.Muller LM, Gorter KJ, Hak E, et al. Increased risk of common infections in patients with type 1 and type 2 diabetes mellitus. Clin Infect Dis 2005; 41:281–8. [DOI] [PubMed] [Google Scholar]
- 15.Olsen MA, Higham-Kessler J, Yokoe DS, et al. Developing a risk stratification model for surgical site infection after abdominal hysterectomy. Infect Control Hosp Epidemiol 2009; 30:1077–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Segal CG, Waller DK, Tilley B, Piller L, Bilimoria K. An evaluation of differences in risk factors for individual types of surgical site infections after colon surgery. Surgery 2014; 29:00272–4. [DOI] [PubMed] [Google Scholar]
- 17.Thomsen RW, Hundborg HH, Lervang HH, Johnsen SP, Schonheyder HC, Sorensen HT. Risk of community-acquired pneumococcal bacteremia in patients with diabetes: a population-based case-control study. Diabetes Care 2004; 27:1143–7. [DOI] [PubMed] [Google Scholar]
- 18.Tserenpuntsag B, Haley V, Van Antwerpen C, et al. Surgical site infection risk factors identified for patients undergoing colon procedures, New York state 2009–2010. Infect Control Hosp Epidemiol 2014; 35:1006–12. [DOI] [PubMed] [Google Scholar]
- 19.Wang HE, Griffin R, Judd S, Shapiro NI, Safford MM. Obesity and risk of sepsis: a population-based cohort study. Obesity (Silver Spring) 2013; 21:E762–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wukich DK, Crim BE, Frykberg RG, Rosario BL. Neuropathy and poorly controlled diabetes increase the rate of surgical site infection after foot and ankle surgery. J Bone Joint Surg Am 2014; 96:832–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shah BR, Hux JE. Quantifying the risk of infectious diseases for people with diabetes. Diabetes Care 2003; 26:510–3. [DOI] [PubMed] [Google Scholar]
- 22.Nguyen NT, Nguyen XM, Lane J, Wang P. Relationship between obesity and diabetes in a US adult population: findings from the National Health and Nutrition Examination Survey, 1999–2006. Obes Surg 2011; 21:351–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.CDC. Active Bacterial Core surveillance methodology. 2012. Available at: http://www.cdc.gov/abcs/index.html. Accessed 23 May 2013.
- 24.CDC. Introduction to emm typing: M protein gene (emm) typing Streptococcus pyogenes. 2008. Available at: http://www.cdc.gov/streplab/M-ProteinGene-typing.html. Accessed 29 July 2015.
- 25.Graham JW, Olchowski AE, Gilreath TD. How many imputations are really needed? Some practical clarifications of multiple imputation theory. Prev Sci 2007; 8:206–13. [DOI] [PubMed] [Google Scholar]
- 26.White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Statist Med 2011; 30:377–99. [DOI] [PubMed] [Google Scholar]
- 27.National Institutes of Health NH, Lung and Blood Institute. The Practical Guide: Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. 2000.
- 28.CDC. Methodologic changes in the behavioral risk factor surveillance system in 2011 and potential effects on prevalence estimates. MMWR Morb Mortal Wkly Rep 2012; 61:410–3. [PubMed] [Google Scholar]
- 29.Zhang X, Holt JB, Lu H, et al. Multilevel regression and poststratification for small-area estimation of population health outcomes: a case study of chronic obstructive pulmonary disease prevalence using the behavioral risk factor surveillance system. Am J Epidemiol 2014; 179:1025–33. [DOI] [PubMed] [Google Scholar]
- 30.Ezzati M, Martin H, Skjold S, Vander Hoorn S, Murray CJ. Trends in national and state-level obesity in the USA after correction for self-report bias: analysis of health surveys. J R S Med 2006; 99:250–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Thomsen RW, Hundborg HH, Lervang HH, Johnsen SP, Sorensen HT, Schonheyder HC. Diabetes and outcome of community-acquired pneumococcal bacteremia: a 10-year population-based cohort study. Diabetes Care 2004; 27:70–6. [DOI] [PubMed] [Google Scholar]
- 32.Factor SH, Levine OS, Schwartz B, et al. Invasive group A streptococcal disease: risk factors for adults. Emerg Infect Dis 2003; 9:970–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wagner DR, Heyward VH. Measures of body composition in blacks and whites: a comparative review. Am J Clin Nutr 2000; 71:1392–402. [DOI] [PubMed] [Google Scholar]
- 34.Falagas ME, Kompoti M. Obesity and infection. Lancet Infect Dis 2006; 6:438–46. [DOI] [PubMed] [Google Scholar]
- 35.Huttunen R, Syrjanen J. Obesity and the risk and outcome of infection. Int J Obes (Lond) 2013; 37:333–40. [DOI] [PubMed] [Google Scholar]
- 36.Koh GC, Peacock SJ, van der Poll T, Wiersinga WJ. The impact of diabetes on the pathogenesis of sepsis. Eur J Clin Microbiol Infect Dis 2012; 31:379–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.CDC. Active bacterial core surveillance: Surveillance reports. 2014. Available at: http://www.cdc.gov/abcs/reports-findings/surv-reports.html. http://www.cdc.gov/nchs/data/hestat/underweight_adult_07_10/underweight_adult_07_10.pdf. Accessed 17 September 2014.
- 38.CDC. Prevention of invasive group A streptococcal disease among household contacts of case patients and among postpartum and postsurgical patients: recommendations from the Centers for Disease Control and Prevention. Clin Infect Dis 2002; 35:950–9. [DOI] [PubMed] [Google Scholar]
- 39.Dhurandhar NV, Bailey D, Thomas D. Interaction of obesity and infections. Obes Rev 2015; 10:12320. [DOI] [PubMed] [Google Scholar]