GRAPHICAL ABSTRACT

INTRODUCTION

Individuals with diabetes are more susceptible to pneumonia than the general population [1-4]. In the United States, annual hospitalization rates from pneumonia drawn from the National Inpatient Sample and National Health Interview Surveys 2000–2015 were 2.2–2.6 times as high in adults with diabetes compared to those without diabetes [1]. Furthermore, declines in hospitalization rates from pneumonia observed in adults without diabetes were not observed in those with diabetes. Rather, significant increases in hospitalization rates from pneumonia were noted in young adults with diabetes, suggesting the necessity of risk mitigation of pneumonia in adults with diabetes [1]. To reduce the social burden associated with excess mortality from pneumonia in people with diabetes, it may be helpful to identify risk factors associated with death from pneumonia and to provide appropriate preventive strategies to vulnerable groups.

Low household income is a potential factor related to excess mortality from pneumonia in people with diabetes. Although multiple factors unequivocally contribute to outcomes in diabetes, social and environmental factors, collectively considered social determinants of health (SDOH), are key components affecting health and healthcare disparities, accounting for 50% to 60% of health outcomes [5,6]. It is an important public health issue to identify the impact of SDOH in diabetes considering the high prevalence, economic costs, and disproportional population burden of disease [6-8]. In addition to education and occupation, income is a major axis of socioeconomic status (SES), one of the key components of SDOH [6]. Regarding the impact of income on glycemia and outcomes in type 2 diabetes mellitus (T2DM), a meta-analysis reported an inverse association between income and glycated hemoglobin levels in people with T2DM [9], and low family income has been associated with increased risk of diabetes-related mortality [10]. However, to the best of our knowledge, no previous study has specifically examined the relationship between income and mortality from pneumonia in individuals with T2DM, although a previous study suggested that lower income levels are potent risk factors of bacteremic pneumococcal pneumonia in the general population in United States [11].

Furthermore, requiring insulin treatment may indicate more advanced diabetes among people with T2DM in the real world, which might increase vulnerability to death from pneumonia. Previously, we reported that insulin users, compared with nonusers, experienced 2.5-fold increased hazard of all-cause mortality during a mean 7.8 years in a nationwide cohort with recently-diagnosed T2DM [12]. Leutner et al. [13] also reported that stable monotherapy with insulin and insulin in combination with other antidiabetic medications were associated with increased risk of pneumonia in people with T2DM compared to untreated controls.

Therefore, to assess the effects of low household income on the risk of death from pneumonia in people with T2DM, we compared the hazards of death from pneumonia according to cumulative duration in low household income status among adults with T2DM with or without insulin therapy using the Korean National Health Insurance Service (KNHIS) database. In addition, to examine the combined effect of insulin use and cumulative duration in low household income status, hazards of mortality from pneumonia were compared in groups categorized by insulin use and exposure duration to low household income status.

METHODS

Data sources

The KNHIS is the single nationwide insurer operated by the Korean government. It covers all residents in Korea with two major programs: National Health Insurance covering approximately 97% of the population, and Medical Aid (MA) for the lowest income people encompassing the remaining 3% of the population [14]. In 2012, the KNHIS established a public database using medical treatment and health screening records and eligibility data from an existing database system [15]. For this study, we used KNHIS data from January 2002 to December 2018. Anonymized data including demographics, monthly household income, and date of death for the deceased, primary and secondary diagnoses described according to the International Classification of Diseases 10th Revision (ICD-10), prescriptions, procedures, and dates of hospital visits and hospitalizations for all Korean residents are available from the KNHIS datasets. Furthermore, the KNHIS recommends standardized preventive health examinations at least every 2 years by actively running a national health screening program. Health examination results, including demographic information, smoking history, alcohol consumption, physical activity, anthropometric measurements such as height, weight, and waist circumference, blood pressure, and laboratory data, including lipid profiles, fasting plasma glucose level, and estimated glomerular filtration rate are included in the datasets of preventive health examinations. These comprise the largest scale, nationwide cohort database with laboratory test results in Korea [16]. Standardized health examinations are conducted only in hospitals certified by the KNHIS. Details about this database have been provided in previous reports [14,15]. In addition, in an anonymous state, these KNHIS data are linked to data from the Korean Statistical Information Service of Statistics Korea, the official information source on Korean population and mortality records, which enables to determine the causes of death.

The Institutional Review Board (IRB) of Soongsil University approved this study (file number: SSU-202003-HR-201-01). An informed consent exemption was granted by the IRB because the researchers were provided anonymous, de-identified data by the KNHIS.

Study cohort, outcomes, and follow-up

In this nationwide, longitudinal, population-based cohort study, we included individuals with T2DM who underwent at least one health examination between 2009 and 2012 and were aged ≥20 years at the time of this examination (baseline). T2DM was defined as having ≥one claim per year for the prescription of anti-diabetic medication under ICD-10 codes E11–14 or having a fasting plasma glucose ≥126mg/dL, as in previous studies [17-19]. The timepoint of the health examination between 2009 and 2012 was considered as the baseline. We excluded individuals missing the household income variable, those who died within 1 year of baseline, and those with missing data for at least one other variable (Fig. 1).

The endpoint was death from pneumonia (ICD-10 codes J12–18), ascertained using cause of death information from the Korean Statistical Information Service of Statistics Korea. The study population was followed up from baseline (date of health examination between 2009 and 2012) until the date of death from pneumonia or other causes, or December 31, 2018, whichever came first.

Exposure to low household income state

Every year, the KNHIS identifies the household income status of all enrollees and determines the lowest income people (the lowest 3%) to be registered for MA. Individuals were considered low household income status if they were MA beneficiaries in a certain year. They were categorized into three groups according to years exposed to this condition during the 5 years prior to the baseline (0, 1–4, and 5 years).

Measurements and definitions

Information on smoking history, alcohol consumption, and regular exercise was collected from questionnaires. Alcohol consumption [20], regular exercise [16], hypertension [17,21], dyslipidemia [17,21], chronic kidney disease (CKD) [17,22], and insulin use [12,23] were defined according to previous studies, and these definitions are summarized in Supplementary Table 1. Chronic lower respiratory diseases and lung diseases due to external agents (pneumoconiosis, hypersensitivity pneumonitis due to organic dust, and airway disease due to specific organic dust) were defined based on the corresponding ICD-10 diagnostic codes (Supplementary Table 1). The definition of body mass index (BMI) is also provided in Supplementary Table 1. Venous samples for blood tests, including plasma glucose and lipid profiles, were drawn after an overnight fast.

Statistical analysis

SAS software version 9.3 (SAS Institute, Cary, NC, USA) was used for statistical analyses. Two-sided P values <0.05 were considered significant. The baseline characteristics of the study population are presented according to the three groups categorized by the exposed years to low household income state during the 5 years before baseline. We presented continuous variables with normal distributions as mean±standard deviation, and categorical variables as frequency and percentage.

The incidence rates of death from pneumonia during follow-up were derived as number of incident cases divided by total follow-up duration (person-years). The cumulative incidence rate of outcome was compared according to three groups categorized by exposure duration to low household income status using Kaplan-Meier curves; differences among the three groups were evaluated using a log-rank test. We applied multivariable Cox regression analysis to calculate the hazard ratios (HRs) and 95% confidence intervals (CIs) for the incidence of the outcome measure according to the three groups of exposures to low household income state. Model 1 was adjusted for age (as a continuous variable) and sex. Model 2 was adjusted for smoking history, alcohol consumption, regular exercise, and presence of CKD in addition to the potential confounders included in Model 1, while Model 3 was adjusted for the potential confounders included in Model 2, fasting plasma glucose, and diabetes duration (≥5 years vs. <5 years). Model 4 was adjusted for the potential confounders included in Model 3, and presence of chronic lower respiratory disease and lung diseases due to external agents. Potential multi-collinearity was evaluated using variance inflation factor (VIF), and VIFs of all variables were lower than 2 (Supplementary Table 2). In constructing these regression models, various risk factors of pneumonia or death from pneumonia were included as potential confounders based on previous literature [24-31]. The proportional hazard assumption of the Cox models was assessed from examinations of Schoenfeld residuals.

The HRs (95% CIs) for outcome incidence in subjects with 1 or more years of exposure to low household income state were compared with those in subjects with no exposure (reference) during the previous 5 years from baseline in subgroups. These subgroups were categorized by age (<65 years vs. ≥65 years), sex, current smoking, heavy alcohol consumption, regular exercise, presence of obesity (defined as a BMI ≥25 kg/m² referring to the obesity guidelines for the Korean population [32]), insulin use, and diabetes duration (≥5 years vs. <5 years). The potential effect modification by factors determining the subgroups was assessed, and P for interaction was calculated.

To examine the potential effect modification by insulin use in more detail, the HRs (95% CIs) of outcome incidence according to three groups of exposure to low household income state was evaluated after stratifying the study population by the insulin use, and P for interaction was calculated. Furthermore, the study population was categorized into six groups based on insulin use and years exposed to low household income state during the 5 years before baseline. HRs (95% CIs) for outcome incidence were compared according to these six groups to evaluate the combined effect of insulin use and cumulative duration in low household income state.

Sensitivity analyses

To account for competing risk by death from all-other causes (all-causes except for pneumonia), we plotted a cumulative incidence function for death from pneumonia, and estimated subdistribution HRs with the Fine and Gray method for outcome incidence [33] after adjusting for the same potential confounders used in the main analysis. Additional sensitivity analysis was conducted after excluding individuals who developed outcomes within 2 years after baseline.

RESULTS

Baseline characteristics of the study population

A total of 2,503,581 individuals with T2DM aged ≥20 years was included in the study (Fig. 1). Among them, 2,436,505 (97.32%) had never been exposed to low household income state during the previous 5 years before baseline while 20,922 (0.83%) had been consistently exposed to low household income state throughout the whole 5 years prior to the baseline. The baseline characteristics of the study population are presented according to the three groups categorized by the duration in low household income state (Table 1). Individuals without exposure to low household income state were more likely to be male and moderate or heavy drinkers, and to exercise regularly, whereas those with exposure had a higher prevalence of hypertension and CKD, and a higher proportion of nondrinkers. As the duration in low household income state advances, trends toward increase were observed in proportions of insulin users, longer-standing T2DM (diabetes duration ≥5 years), and prevalence of dyslipidemia and chronic lower respiratory diseases, and trends of decrease were noted in proportions of individuals aged <40 years and total and low-density lipoprotein cholesterol, in conjunction with the lipid-lowering agent use for hyperlipidemia. The mean age and proportion of individuals aged ≥65 years were the highest in the group with 1 to 4 years of exposure to low household income state.

Death from pneumonia during follow-up according to duration in low household income state

During a median follow-up of 7.18 years (17,248,731.21 person-years), 12,245 deaths from pneumonia occurred in the entire cohort. The cumulative incidence of the outcome measure is shown according to the three groups categorized by durations in low household income state using Kaplan-Meier curves (Fig. 2). The cumulative incidence of outcome was the highest in individuals with 1 to 4 years of exposure to low household income state, who had the highest mean age and proportion of those aged ≥65 years. However, after adjusting for potential confounders including age and sex, HRs (95% CIs) for outcome incidence were the highest in individuals who had been consistently in a low household income state for 5 years (Table 2). Using individuals without exposure as the reference, those who had been exposed to low household income state had significantly higher hazards of death from pneumonia during follow-up in a dose-response manner (HR, 1.726; 95% CI, 1.568 to 1.899 in those who had been exposed for 1–4 years and HR, 4.686; 95% CI, 3.948 to 5.562 in those who had been consistently exposed for 5 years) in the fully-adjusted model 4.

Subgroup analyses

The HRs (95% CIs) of outcome in subjects with exposure to low household income were analyzed in subgroups stratified by age (<65 years vs. ≥65 years), sex, current smoking, heavy alcohol consumption, regular exercise, presence of obesity, insulin use, and diabetes duration (≥5 years vs. <5 years), and compared with those of subjects with no exposure (Fig. 3). Individuals with exposure to low household income state had higher hazard of outcomes in all subgroups compared with those without exposure. However, the association was more prominent in people aged <65 years, males, current smokers (P for interaction <0.0001 for all of the aforementioned), and heavy alcohol consumers (P for interaction=0.0039). No other significant effect modifications were noted.

Stratified analyses according to insulin use

When the main analysis was rerun after stratifying the study population according to the insulin use, individuals who had been exposed to low household income state had significantly higher hazards of death from pneumonia during follow-up compared to those without exposure regardless of insulin use (Supplementary Table 3). No significant effect modification by insulin use was observed in models 1–4 (P for interaction 0.7968 in model 4).

Death from pneumonia during follow-up according to the insulin use and duration in low household income state

The HRs (95% CIs) for outcome incidence were compared among six groups categorized by insulin use and years exposed to low household income state during the 5 years before baseline (exposed duration 0, 1–4, and 5 years) (Supplementary Table 4). When insulin non-users without exposure to low household income state were used as a reference, insulin users with 5 years of exposure to low household income state exhibited the highest hazard of death from pneumonia during follow-up (HR, 7.193; 95% CI, 5.266 to 9.825), followed by insulin non-users with 5 years of exposure (HR, 4.586; 95% CI, 3.741 to 5.621) and insulin users with 1–4 years of exposure (HR, 2.887; 95% CI, 2.355 to 3.539). Insulin users without exposure to low household income state (HR, 1.758; 95% CI, 1.675 to 1.846) and insulin non-users with 1–4 years of exposures (HR, 1.706; 95% CI, 1.531 to 1.900) also had higher hazards of outcome than the reference group.

Sensitivity analyses

When deaths from all other causes were accounted for as competing events, the corresponding cumulative incidence function and subdistribution HRs for death from pneumonia did not change from the main findings (Supplementary Table 5). The main results were consistent when individuals who developed outcomes within 2 years of follow-up were excluded (Supplementary Table 6).

DISCUSSION

In this large-scale nationwide longitudinal study including 2,503,581 adults with T2DM, cumulative exposures to low household income state were associated with increased risks of death from pneumonia in a dose-response manner. These associations were significant even after adjusting for risk factors of pneumonia or death from pneumonia including preexisting pulmonary diseases, lifestyle factors, presence of CKD, and category of diabetes duration. Subdistribution HRs accounting for mortality from all other causes as a competing event and sensitivity analyses excluding individuals who developed outcome within 2 years of follow-up also yielded consistent results. Findings were consistent in subgroup analyses stratified by age, sex, current smoking, heavy alcohol consumption, regular exercise, presence of obesity, insulin use, and diabetes duration, while the impact of low household income was more prominent in individuals aged <65 years, males, current smokers, and heavy alcohol consumers. When the study population was categorized based on insulin use and duration in low household income state, insulin users with the highest and most sustained exposures to low household income state were associated with the highest risk of death from pneumonia.

To the best of our knowledge, this is the first study to examine the impact of low household income on the risk of death from pneumonia in people with T2DM. Recently, SDOH including SES drew attention as a key contributor of diabetes outcomes [6], and lower SES was suggested as a determining factor of poor glycemia, quality of life, complications including cardiovascular disease, retinopathy, end-stage renal disease, and amputation, and early mortality in people with T2DM [34]. Among the components of SES, lower income has been related to higher glycated hemoglobin values [9], and increased risk of diabetes-related mortality [10], in people with T2DM. Furthermore, low SES, defined based on household income, insurance type, and residence, was a strong predictor of amputation among Koreans with diabetic foot ulcers [35]. Our findings are consistent with previous reports on the associations between low income and other adverse outcomes in T2DM and indicate that low household income in T2DM may predict higher risk of death from pneumonia as well.

The excess hazard of death from pneumonia from low household income state among adults with T2DM was more pronounced in individuals aged <65 years, males, current smokers, and heavy alcohol consumers in subgroup analyses. This suggests that the impact of cumulative exposures to low household income state in terms of the risk of death from pneumonia in T2DM may be more prominent among these subpopulations. Although exact mechanisms cannot be clarified in the current study, we considered several possible explanations. First, unfavorable changes in health behaviors resulting from chronic stress associated with low household income may have occurred more actively in younger people, current smokers, or heavy alcohol consumers. Maladaptive health behaviors including current smoking and heavy drinking may lead to cumulative biopsychosocial vulnerability [5,34], and low household income might have greater impacts on self-management behaviors and subsequently on general health among current smokers and heavy alcohol consumers. Second, with respect to interaction by age group, from a statistical point of view, older subjects (aged ≥65 years) may be at higher risk for death from pneumonia in general, possibly attenuating the excess effect of low household income. Third, regarding the interaction by sex, men might be more vulnerable to low income than women. In previous studies among the general populations of Spain and the United States, the gaps in life expectancy and mortality by SES were generally greater among men than women [36,37]. Similarly, the influence of income trajectories on mortality and hospitalizations was larger in men than in women in a nationwide general population study in Denmark [38].

Interestingly, proportions of insulin users at baseline tended to increase as duration in low household income state advances while insulin users with the highest cumulative exposures to low household income state had the highest hazard of mortality from pneumonia. These findings suggest that among adults with T2DM, insulin users exposed to low household income state are a medically vulnerable group that requires monitoring and systematic interventions to prevent excess mortality from pneumonia. In a real-world cohort study from our group [12], an excess hazard of all-cause mortality was noted in insulin users compared to non-users during a mean 7.8 years among individuals with recently-diagnosed T2DM. These findings in real-world studies are irrespective of results from randomized trial [39] that demonstrated no significant increase in early mortality due to the insulin treatment itself. Insulin use in the real world may be closely related to the disease progression in T2DM. Thus, increased risk of death from pneumonia in insulin users are likely not to originate from the effect of insulin itself. Rather, other indications of insulin treatment including progressed diabetes or disease control states may have affected. Insulin use along with low household income states may be used as factors to detect vulnerable populations to death from pneumonia among people with T2DM in the real-world.

Limitations of this study should be acknowledged. First, considering the observational nature of the study, it is inevitably limited in ability to clarify causal relationships. Thus, a reverse causality effect that deterioration of health increasing susceptibility to more severe pneumonia might have led to repeated exposures to low household income state should also be considered. However, to minimize this possibility, exposure to low household income state was examined during the 5 years prior to baseline, as opposed to after baseline. Furthermore, individuals who died within 1 year after baseline were excluded, and additional sensitivity analysis excluding those who developed outcomes within 2 years of follow-up also demonstrated consistent findings. Second, as an observational study, effects of unmeasured confounding may remain despite adjustments for measured potential confounders. For instance, in the KNHIS datasets, which was used for the current study, indicators of SES other than income status (such as education) are unavailable, and thus, we could not include them as potential confounders. In addition, due to the unavailability of past data, diabetes duration was treated only as a categorical variable (≥5 years vs. <5 years). Third, extrapolation of our findings to people with different ethnicities, those without T2DM, or those with type 1 diabetes mellitus should be cautious since our study population included only Korean adults with T2DM. Nevertheless, our study has several strengths including a large sample size (n=2,503,581), use of a representative nationwide cohort database managed by the Korean government that includes diverse lifestyle variables, and laboratory data from a huge population. Thus, adjustments for various potential confounders were possible. Furthermore, we examined low household income state in terms of repeated exposure during the 5 years before baseline, not simply as a cross-sectional status at a single timepoint, thus enabling evaluations on a dose-response relationship and providing novelty. Finally, we obtained consistent results from subgroup analyses and sensitivity analyses.

In this large real-world nationwide population-based study, cumulative exposures to low household income state were associated with an increased risk of mortality from pneumonia among adults with T2DM in a graded dose-response manner. Particularly, insulin users with the highest cumulative exposures to low household income state were most vulnerable to death from pneumonia. These findings advance the argument that for adults with T2DM, strategies to prevent lethal pneumonia including monitoring systems, vaccination, screening, and care for underlying comorbid respiratory diseases, and appropriate glycemic control should be offered more intensively to the low-income population, especially those using insulin, and policy-level interventions and investment to mitigate diabetes-related excess mortality from pneumonia and disparities should be promoted.

SUPPLEMENTARY MATERIALS

NOTES

CONFLICTS OF INTEREST

Kyu Yeon Hur was editorial board member of the Diabetes & Metabolism Journal from 2020 to 2021. Sang-Man Jin has been associate editor of the Diabetes & Metabolism Journal since 2022. They were not involved in the review process of this article. Otherwise, there was no conflict of interest.

AUTHOR CONTRIBUTIONS

Conception or design: Y.B.L., S.H.P., K.H., J.H.K.

Acquisition, analysis, or interpretation of data: K.L., B.K., K.H.

Drafting the work or revising: Y.B.L., J.H.K.

Final approval of the manuscript: Y.B.L., S.H.P., K.L., B.K., S.Y.K., J.P., G.K., S.M.J., K.Y.H., K.H., J.H.K.

FUNDING

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. 2019R1I1A1A0106118813) to Prof. Kyungdo Han. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Acknowledgements

This work was performed using data from the Korean National Health Insurance Service (KNHIS). We used the National Health Information Database constructed by the KNHIS, and the study results do not necessarily represent the opinion of the KNHIS.

Fig. 1.

Flow diagram of the study population.

Fig. 2.

Cumulative incidence of death from pneumonia according to exposure duration to low household income state. The dashed lines represent the 95% confidence intervals.

Fig. 3.

Adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for death from pneumonia in subjects with versus without exposure to low household income state according to subgroups. Adjusted for age, sex, smoking history, alcohol consumption, regular exercise, presence of chronic kidney disease, fasting plasma glucose, and diabetes duration (≥5 years vs. <5 years).

REFERENCES

• 1. Harding JL, Benoit SR, Gregg EW, Pavkov ME, Perreault L. Trends in rates of infections requiring hospitalization among adults with versus without diabetes in the U.S., 2000-2015. Diabetes Care 2020;43:106-16.ArticlePubMedPDF
• 2. Kornum JB, Thomsen RW, Riis A, Lervang HH, Schonheyder HC, Sorensen HT. Diabetes, glycemic control, and risk of hospitalization with pneumonia: a population-based case-control study. Diabetes Care 2008;31:1541-5.PubMedPMC
• 3. Smith SA, Poland GA; American Diabetes Association. Influenza and pneumococcal immunization in diabetes. Diabetes Care 2004;27 Suppl 1:S111-3.PubMed
• 4. Valdez R, Narayan KM, Geiss LS, Engelgau MM. Impact of diabetes mellitus on mortality associated with pneumonia and influenza among non-Hispanic black and white US adults. Am J Public Health 1999;89:1715-21.ArticlePubMedPMC
• 5. Ogunwole SM, Golden SH. Social determinants of health and structural inequities: root causes of diabetes disparities. Diabetes Care 2021;44:11-3.ArticlePubMedPDF
• 6. Hill-Briggs F, Adler NE, Berkowitz SA, Chin MH, Gary-Webb TL, Navas-Acien A, et al. Social determinants of health and diabetes: a scientific review. Diabetes Care 2020;44:258-79.ArticlePubMedPMCPDF
• 7. Haire-Joshu D, Hill-Briggs F. The next generation of diabetes translation: a path to health equity. Annu Rev Public Health 2019;40:391-410.ArticlePubMed
• 8. Golden SH, Maruthur N, Mathioudakis N, Spanakis E, Rubin D, Zilbermint M, et al. The case for diabetes population health improvement: evidence-based programming for population outcomes in diabetes. Curr Diab Rep 2017;17:51.ArticlePubMedPMCPDF
• 9. Bijlsma-Rutte A, Rutters F, Elders PJ, Bot SD, Nijpels G. Socioeconomic status and HbA1c in type 2 diabetes: a systematic review and meta-analysis. Diabetes Metab Res Rev 2018;34:e3008.ArticlePubMedPDF
• 10. Saydah S, Lochner K. Socioeconomic status and risk of diabetes-related mortality in the U.S. Public Health Rep 2010;125:377-88.ArticlePubMedPMCPDF
• 11. Flory JH, Joffe M, Fishman NO, Edelstein PH, Metlay JP. Socioeconomic risk factors for bacteraemic pneumococcal pneumonia in adults. Epidemiol Infect 2009;137:717-26.ArticlePubMedPMC
• 12. Lee YB, Han K, Kim B, Choi MS, Park J, Kim M, et al. Risk of early mortality and cardiovascular disease according to the presence of recently diagnosed diabetes and requirement for insulin treatment: a nationwide study. J Diabetes Investig 2021;12:1855-63.ArticlePubMedPMCPDF
• 13. Leutner M, Kaleta M, Bellach L, Kautzky A, Thurner S, Klimek P, et al. Insulin as monotherapy and in combination with other glucose-lowering drugs is related to increased risk of diagnosis of pneumonia: a longitudinal assessment over two years. J Pers Med 2021;11:984.ArticlePubMedPMC
• 14. Lee YH, Han K, Ko SH, Ko KS, Lee KU; Taskforce Team of Diabetes Fact Sheet of the Korean Diabetes Association. Data analytic process of a nationwide population-based study using National Health Information Database established by National Health Insurance Service. Diabetes Metab J 2016;40:79-82.ArticlePubMedPMCPDF
• 15. Seong SC, Kim YY, Khang YH, Park JH, Kang HJ, Lee H, et al. Data resource profile: the National Health Information Database of the National Health Insurance Service in South Korea. Int J Epidemiol 2017;46:799-800.PubMed
• 16. Lee YB, Kim DH, Kim SM, Kim NH, Choi KM, Baik SH, et al. Hospitalization for heart failure incidence according to the transition in metabolic health and obesity status: a nationwide population-based study. Cardiovasc Diabetol 2020;19:77.ArticlePubMedPMCPDF
• 17. Lee YB, Han K, Kim B, Lee SE, Jun JE, Ahn J, et al. Risk of early mortality and cardiovascular disease in type 1 diabetes: a comparison with type 2 diabetes, a nationwide study. Cardiovasc Diabetol 2019;18:157.ArticlePubMedPMCPDF
• 18. Lee YB, Han K, Kim B, Jin SM, Lee SE, Jun JE, et al. High proportion of adult cases and prevalence of metabolic syndrome in type 1 diabetes mellitus population in Korea: a nationwide study. Diabetes Metab J 2019;43:76-89.ArticlePubMedPMCPDF
• 19. Noh J, Han KD, Ko SH, Ko KS, Park CY. Trends in the pervasiveness of type 2 diabetes, impaired fasting glucose and comorbidities during an 8-year-follow-up of nationwide Korean population. Sci Rep 2017;7:46656.ArticlePubMedPMCPDF
• 20. Stampfer MJ, Kang JH, Chen J, Cherry R, Grodstein F. Effects of moderate alcohol consumption on cognitive function in women. N Engl J Med 2005;352:245-53.ArticlePubMed
• 21. Kim MK, Han K, Kim HS, Park YM, Kwon HS, Yoon KH, et al. Cholesterol variability and the risk of mortality, myocardial infarction, and stroke: a nationwide population-based study. Eur Heart J 2017;38:3560-6.ArticlePubMedPMC
• 22. Lee YB, Han K, Kim B, Jun JE, Lee SE, Ahn J, et al. Risk of endstage renal disease from chronic kidney disease defined by decreased glomerular filtration rate in type 1 diabetes: a comparison with type 2 diabetes and the effect of metabolic syndrome. Diabetes Metab Res Rev 2019;35:e3197.ArticlePubMedPDF
• 23. Lee YB, Kim B, Park J, Kim M, Choi MS, Kim G, et al. Early mortality and cardiovascular disease, varied association with body mass index and its changes in insulin-treated diabetes: a nationwide study. Int J Obes (Lond) 2021;45:2482-9.ArticlePubMedPDF
• 24. Simou E, Britton J, Leonardi-Bee J. Alcohol and the risk of pneumonia: a systematic review and meta-analysis. BMJ Open 2018;8:e022344.ArticlePubMedPMC
• 25. Song Y, Ren F, Sun D, Wang M, Baker JS, Istvan B, et al. Benefits of exercise on influenza or pneumonia in older adults: a systematic review. Int J Environ Res Public Health 2020;17:2655.ArticlePubMedPMC
• 26. James MT, Quan H, Tonelli M, Manns BJ, Faris P, Laupland KB, et al. CKD and risk of hospitalization and death with pneumonia. Am J Kidney Dis 2009;54:24-32.ArticlePubMed
• 27. Baskaran V, Murray RL, Hunter A, Lim WS, McKeever TM. Effect of tobacco smoking on the risk of developing community acquired pneumonia: a systematic review and meta-analysis. PLoS One 2019;14:e0220204.ArticlePubMedPMC
• 28. Baik I, Curhan GC, Rimm EB, Bendich A, Willett WC, Fawzi WW. A prospective study of age and lifestyle factors in relation to community-acquired pneumonia in US men and women. Arch Intern Med 2000;160:3082-8.ArticlePubMed
• 29. Sarnak MJ, Jaber BL. Pulmonary infectious mortality among patients with end-stage renal disease. Chest 2001;120:1883-7.ArticlePubMed
• 30. Grau I, Ardanuy C, Calatayud L, Schulze MH, Linares J, Pallares R. Smoking and alcohol abuse are the most preventable risk factors for invasive pneumonia and other pneumococcal infections. Int J Infect Dis 2014;25:59-64.ArticlePubMed
• 31. Bello S, Menendez R, Antoni T, Reyes S, Zalacain R, Capelastegui A, et al. Tobacco smoking increases the risk for death from pneumococcal pneumonia. Chest 2014;146:1029-37.ArticlePubMed
• 32. Kim MK, Lee WY, Kang JH, Kang JH, Kim BT, Kim SM, et al. 2014 Clinical practice guidelines for overweight and obesity in Korea. Endocrinol Metab (Seoul) 2014;29:405-9.ArticlePubMedPMC
• 33. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999;94:496-509.Article
• 34. Brown AF, Ettner SL, Piette J, Weinberger M, Gregg E, Shapiro MF, et al. Socioeconomic position and health among persons with diabetes mellitus: a conceptual framework and review of the literature. Epidemiol Rev 2004;26:63-77.ArticlePubMed
• 35. Ha JH, Jin H, Park JU. Association between socioeconomic position and diabetic foot ulcer outcomes: a population-based cohort study in South Korea. BMC Public Health 2021;21:1395.ArticlePubMedPMCPDF
• 36. Bilal U, Cainzos-Achirica M, Cleries M, Santaeugenia S, Corbella X, Comin-Colet J, et al. Socioeconomic status, life expectancy and mortality in a universal healthcare setting: an individual-level analysis of >6 million Catalan residents. Prev Med 2019;123:91-4.ArticlePubMed
• 37. Chetty R, Stepner M, Abraham S, Lin S, Scuderi B, Turner N, et al. The association between income and life expectancy in the United States, 2001-2014. JAMA 2016;315:1750-66.ArticlePubMedPMC
• 38. Ahrenfeldt LJ, Pedersen JK, Thinggaard M, Christensen K, Lindahl-Jacobsen R. Sex differences in health and mortality by income and income changes. J Epidemiol Community Health 2020;74:225-31.ArticlePubMedPMC
• 39. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998;352:837-53.ArticlePubMed

Figure & Data

References

Citations

Citations to this article as recorded by