Trend of incidence rate of age-related diseases: results from the National Health Insurance Service–National Sample Cohort (NHIS-NSC) database in Korea: a cross- sectional study

Abstract

Background

This study aimed to identify and select age-related diseases (ARDs) in Korea, which is about to have a super-aged society, and to elucidate patterns in their incidence rates.

Methods

The National Health Insurance Service–National Sample Cohort, comprising 1 million health insurance and medical benefit beneficiaries in Korea from 2002 to 2019, was utilized. We selected 14 diseases with high disease burden and prevalence among Koreans from the 92 diseases defined in the Global Burden of Diseases, Injuries, and Risk Factors Study as ARDs. The annual incidence rate represented the number of patients newly diagnosed with an ARD each year from 2006 to 2019, excluding those with a history of ARD diagnosis from 2002 to 2005. The incidence rate by age was categorized into 10-year units based on age as of 2019. The number of patients with ARDs in each age group was used as the numerator, and the incidence rate for each age group was calculated with the age group as the denominator.

Results

Regarding the annual incidence rates of ARDs from 2006 to 2019, chronic obstructive pulmonary disease, congestive heart failure, and ischemic heart disease decreased annually, whereas dyslipidemia, chronic kidney disease, cataracts, hearing loss, and Parkinson's disease showed a significant increase. Hypertension, diabetes, cerebrovascular disease, osteoporosis, osteoarthritis, and age-related macular degeneration initially displayed a gradual decrease in incidence but exhibited a tendency to increase after 2015. Concerning age-specific incidence rates of ARDs, two types of curves emerged. The first type, characterized by an exponential increase with age, was exemplified by congestive heart failure. The second type, marked by an exponential increase peaking between ages 60 and 80, followed by stability or decrease, was observed in 13 ARDs, excluding congestive heart failure. However, hypertension, ischemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, and hearing loss in men belonged to the first type.

Conclusions

From an epidemiological perspective, there are similar characteristics in age-specific ARDs that increase with age, reaching a peak followed by a plateau or decrease in Koreans.

Background

The older population is increasing worldwide [1], and Korea has the fastest aging rate among the major countries in the Organization for Economic Co-operation and Development [2]. According to a report by the National Statistical Office, the rate of aging in Korea increased at an average annual rate of 3.3% from 1970 to 2018, and Korea is expected to become a super-aged society by 2025 [3]. An increase in the older population is a significant cause of increased medical expenses and financial burden on health insurance [4]. For example, in 2019, medical expenses for the older population (age ≥ 65 years) in Korea accounted for 41.6% of the total medical expenses and increased by 9.3% over the past decade. In addition, the annual medical cost per older population individual is 4.91 million KRW (~ 3,600 USD), three times the annual medical cost per non-older population individual [5].

It is essential to pay more attention to age-related diseases (ARDs), as a significant portion of medical expenses and healthcare burden will be concentrated in the older population with the rapid advent of super-aging. ARDs generally refer to diseases that increase in incidence with age, including chronic diseases such as hypertension, diabetes, cardiovascular disease, cerebrovascular disease, Alzheimer disease, Parkinson's disease, age-related macular degeneration, osteoarthritis, osteoporosis, and cancer [6,7,8,9,10,11,12]. However, no consistent consensus has defined the disease categories included in ARD. The distinction between normal aging, which occurs naturally with advancing age, and pathological aging remains unclear. As geriatric diseases are a combination of age-related decline in function and disease, the disease categories included in ARD are slightly different depending on the literature [6,7,8,9,10,11,12]. Among these, we focused on 92 ARDs classified by defining ARDs as those with exponentially increasing incidence with age, out of a total of 293 causes of disease from the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2017 [13]. The GBD evaluated the burden of each disease of ARDs using the disability-adjusted life-years (DALYs), and ARDs accounted for 51.3% of the total disease burden globally based on the data from the 2017 GBD and the top 10 diseases that had the largest absolute increases in the number of DALYs between 1990 and 2019, including six diseases corresponding to ARD, such as ischemic heart disease, chronic kidney disease, lung cancer, and hearing loss [14].

ARD-related studies in Korea have usually focused on specific diseases such as macular degeneration, hearing loss, and Alzheimer disease [15,16,17]. In addition, although the Geriatrics Fact Sheet and Burdens of Disease of the Older Population have been published using a Korean national cohort [4, 18], no studies have analyzed ARDs or their incidence according to year and age. Therefore, we aimed to identify and select ARDs in terms of disease burden to provide information for the management of the older population in Korea and to analyze their characteristics by evaluating the incidence rate of each disease.

Methods

Definition of ARD

We defined ARD as a disease with a burden that increases with age and an incidence rate that increases exponentially with age. To determine the disease group belonging to ARDs, we evaluated a list of 293 causes of disease from the GBD using the two-step method described by Chang et al., which excludes diseases that do not have a positive correlation between the incidence rate and age as well as diseases whose incidence does not increase exponentially with age; finally, a total of 92 ARDs were identified [13]. In addition, in 2020, the Ministry of Health and Welfare confirmed the results of a survey on older individuals to examine the causes of diseases with high prevalence in the older population in Korea [18]. In a study on the burden of disease in Koreans, the top five specific causes of DALYs by age were examined, and diseases with high incidence rates were identified [19]. Finally, 14 diseases were selected as ARDs after consultation with researchers and clinicians. Cancer, a representative ARD, was excluded from this study.

Data sources

This study used data from the National Health Insurance Service–National Sample Cohort (NHIS-NSC). This database included data from a sample of 1 million individuals who maintained health insurance and were beneficiaries of medical benefits in Korea for 1 year in 2006 and were followed up from 2002 to 2019. The NHIS-NSC data include sociodemographic information, outpatient and inpatient records, pharmacy claims, health examination results, and data on deaths collected by the National Statistical Office, such that the date and cause of death can be determined. The obtained information was then extracted, summarized, processed, and anonymized to ensure that the subjects cannot be identified [20]. The authors received approval from the Ethics Committee of the National Health Insurance Corporation to use this data. The research was conducted after receiving approval (IRB number: NHIS-2021–1-459) from the Ethics Committee of the National Health Insurance Ilsan Hospital.

Incidence rate of ARDs

The diagnostic codes for the 14 diseases classified as ARDs were extracted from individuals with records of these diseases in the NHIS-NSC data. The diagnostic codes used were based on the 8th revised Korea Standard Disease Classification, and they are summarized in Additional file 1.

The annual incidence rate was derived from patients with new occurrences of each ARD each year for a total of 14 years (from 2006 to 2019), excluding patients with a history of being diagnosed with each ARD from 2002 to 2005. The incidence rate for 2006 to 2019 was calculated using the number of patients with ARDs as the numerator and the number of people who did not develop each ARD in the previous year as the denominator. The incidence rate by age was divided into 10-year increments by age as of 2019. To calculate the incidence rate of each ARD by age group, the number of patients with ARDs in each age group was used as the numerator, and the number of people for each age in 2019 was used as the denominator. SAS version 9.4 (SAS Institute, Cary, NC, USA) was used for data pre-processing and incidence rate calculation.

Results

Disease categories of ARD

The 14 ARDs included hypertension, diabetes, dyslipidemia, cerebrovascular disease, ischemic heart disease, osteoporosis, osteoarthritis, chronic obstructive pulmonary disease, congestive heart failure, chronic kidney disease, cataracts, age-related macular degeneration, hearing loss, and Parkinson's disease.

Incidence rates of ARDs per year

Examination of the annual incidence rate per 100,000 people by disease showed that hypertension gradually decreased from 2006, rebounding from its lowest incidence rate in 2014 and demonstrating a gradual increase in the incidence rate (Fig. 1). Meanwhile, the incidence rate of diabetes gradually decreased and experienced a slight increase from 2018, but the changes did not appear to be statistically significant. The incidence rate of dyslipidemia continued to rise. The incidence rate of cerebrovascular disease increased until 2008, then decreased, and started to gradually increase again from 2016, while that of ischemic heart disease showed a gradual decrease. Osteoporosis also showed a steady decline but showed a tendency to increase after 2016. The incidence of osteoarthritis declined until 2009, followed by a gradual increase. The incidence rate of chronic obstructive pulmonary disease gradually decreased. In the case of congestive heart failure, the incidence rate peaked in 2007, remained relatively stable from 2008 to 2015, started to decrease after 2015, and has since maintained a similar pattern, whereas chronic kidney disease showed a moderate incidence but increased after 2015. In the case of cataracts, the incidence rate continued to increase gradually, and the incidence rate of age-related macular degeneration, which remained stable, increased rapidly after 2014, whereas the incidence rate of hearing loss increased gradually over the years. In the case of Parkinson’s disease, the incidence rate itself was not high and varied by year but showed an overall increasing trend (Table 1).

Fig. 1

Incidence rates of age-related diseases by year per 100,000 persons

Table 1 Incidence rate of age-related disease by year

Incidence rates of ARDs per age group

The age-specific incidence rate of each disease generally increases with age but can be broadly categorized into two types. The first type was characterized by an exponential increase in the incidence rate with age, with congestive heart failure corresponding to this pattern (Table 2, Fig. 2). When analyzed separately by sex, hypertension, ischemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, and hearing loss also showed an increase in men (Additional files 2 and 3). The second type was characterized by an exponential increase in the incidence rate with age, reaching a peak between the ages of 60 and 80 years, after which the incidence rate either remained stable or decreased (Table 2, Fig. 2). In this study, 13 ARDs, excluding congestive heart failure, followed this pattern. When differentiated by sex, all diseases, except congestive heart failure and the five previously mentioned diseases, fell into the second category (Additional files 4–7).

Table 2 Incidence rate of age-related disease by age group

Fig. 2

Incidence rates of age-related diseases by age group per 100,000 persons

Discussion

In this study, we selected 14 diseases as ARDs, based on their high prevalence in older Koreans and their incidence rates, which increase with age. Accordingly, there was a trend toward decreasing annual incidence rates of chronic obstructive pulmonary disease, congestive heart failure, and ischemic heart disease from 2006 to 2019. In contrast, dyslipidemia, chronic kidney disease, cataracts, hearing loss, and Parkinson’s disease showed increased incidence rates during the same period. The incidence rates of hypertension, diabetes, cerebrovascular disease, osteoporosis, osteoarthritis, and age-related macular degeneration, which had been gradually decreasing, showed a tendency to increase after 2015.

Our main finding was that there are two different types of disease incidence rate curves when examining diseases according to age in 10-year increments. The first type included diseases for which the incidence rate increased exponentially until the oldest age, represented solely by congestive heart failure. However, when analyzed by sex, the incidence rates of hypertension, ischemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, and hearing loss in men continued to increase until the oldest age. The other type includes diseases for which the incidence rate increased exponentially and peaked between the ages of 60 and 80 years, after which it plateaued or declined. The majority of diseases fell under this category. Several epidemiological studies have investigated ARDs and their incidence rates. Using a primary database from England, researchers grouped 278 diseases into nine clusters based on the median age of disease onset [21]. They assumed that four clusters containing 207 diseases were related to ARD, showing an increase in incidence rates with age. Although the median age of disease onset decreased from one cluster to another, they consistently exhibited an exponential increase in incidence rates. These clusters included dementia (cluster 1; onset age, 82 years), age-related macular degeneration and heart failure (cluster 2; onset age, 77 years), Parkinson’s disease (cluster 3; onset age, 69 years), and hypertension and osteoarthritis (cluster 4; onset age, 57 years). Another study identified four groups of 92 non-communicable diseases using the 2019 GBD database [22]. They showed that group A diseases had an exponential increase in incidence rates with age; group B diseases had an exponential increase in incidence rates that typically peaked in late life and then declined or plateaued at the oldest ages; groups C and D diseases had an onset in early life with stable or decreased incidence rates in old age. Furthermore, another study concluded that seven diseases were selected as ARD because their doubling time of 8-year Gompertz mortality was the same [23]. In other words, various diseases presumed to belong to ARDs exhibit the characteristic of being grouped based on age-specific incidence rates. This similarity suggests that the general biological aging process dominates the pathogenesis of various diseases, which can be explained by the accumulation of senescent cells and differences in individual susceptibility to diseases [24].

Senescent cells stop dividing in response to various stressors and accumulate in the body with age. They secrete senescence-associated secretory profiles (SASP) to induce inflammation or to reproduce normal cells [25]. According to a study on the relationship between age and senescent cells, the turnover of senescent cells produced and eliminated occurs rapidly during young age; however, with increasing age, the turnover slows, especially the rate of elimination [26], and this was used to develop a statistical probability model for the generation and removal of senescent cells. This is called the saturated-removal (SR) model, and it is confirmed that the accumulation of senescent cells occurs because the generation of senescent cells increases with various stressors as one ages, while the self-removal rate decreases [27]. Nevertheless, because the number of senescent cells differs among individuals, the rate of senescent cell removal varies. Assuming that death occurs when senescent cells exceed the threshold, the SR model can explain the distribution of death time [26, 27]. As aging cells are associated with several ARDs, if ARDs occur when they exceed a specific disease threshold, aging cells secrete SASPs that affect the physiological parameters related to the occurrence of certain diseases, causing the disease to exceed the threshold. Therefore, as the number of aging cells increases exponentially with age, the disease increases exponentially with age [24].

However, while previous studies considered that ARD continues to increase in incidence rate with age, our study showed that most of the ARDs we selected exhibited exponential increases, followed by a peak and a subsequent decrease or plateau. Other studies have also shown that diseases belonging to ARD tend to increase approximately exponentially with age and then decrease in very old age (beyond the peak age), similar to the findings of our study [23, 28]. The decrease in the incidence rate at a very old age can be explained by differences in individual susceptibility to specific diseases [29, 30]. Each population has a different susceptibility to diseases owing to differences in genetic or environmental factors; therefore, the risk of developing a disease may vary. Thus, ARD would occur in individuals with a low threshold for each disease, but it would not occur during normal aging in a population with a high disease threshold. However, most very old people with a low threshold for a disease would have already been afflicted with the disease, and most of the remaining people would have a high threshold for the disease; therefore, the probability of developing a new disease is relatively low, resulting in a decrease in the incidence.

A sub-analysis was performed by dividing the incidence rate of age-specific ARDs by sex. In general, the graph curve of the incidence rate was similar between sexes. However, in the case of hypertension, ischemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, and hearing loss in men, unlike in women or all sexes, the incidence rate tends to increase exponentially with age. For vascular diseases, such as hypertension, ischemic heart disease or cerebrovascular disease, the number of disease occurrences continues to increase in men, peaking at 60 years, and then suddenly decreasing at 80 or 90 years. In contrast, in women, it peaks at 60 years and then gradually decreases thereafter. Previous research has shown that, owing to the protective effect of estrogen, the onset of vascular disease occurs later in life, and the mortality rate usually increases after the age of 55 years in women. However, disease onset occurs relatively earlier in men than in women, and the mortality rate is relatively higher in men than in women. Thus, the difference in mortality rates between women and men decreases as they get older [30]. However, our study did not observe a rapid increase in the number of postmenopausal vascular disease events in women. In addition, because women have a longer life expectancy than men, the remaining female population is relatively large. Thus, the incidence rate in patients older than 80 years decreased, even though the number of vascular disease events did not decrease steeply. In contrast, the incidence rate of vascular disease remains high due to the small number of men surviving beyond the age of 80 years, even though the number of cases in men suddenly decreased after the age of 80 years. This phenomenon could be explained by the hypothesis that there is a cohort of healthy survivors with delayed mortality or a cohort of frail individuals with earlier mortality [31]. In the case of chronic obstructive pulmonary disease, one study reported that female patients with chronic obstructive pulmonary disease were younger, had relatively fewer smokers, and had better lung function; however, the incidence rate of dyspnea was higher, and the chronic obstructive pulmonary disease survival rate in women was higher than that in men [32]. Although smoking status and pulmonary function test results were not confirmed in this study, our study showed that the number of cases among men was almost twice as high as that among women, and the survival rate of women was higher than that of men. Unlike men, the number of survivors among women is greater, and the incidence rate among women older than 80 years has decreased. Lastly, in the case of hearing loss, similar to other diseases, the incidence rate is believed to be higher because of the small number of men surviving beyond 80 years.

This study has some limitations. First, since the collected claims data were used for management purposes for insurance claims and refunds, information such as diagnostic codes might be inaccurate, possibly affecting the incidence rate; second, because of this, the actual disease may have been underestimated or overestimated compared with the number of occurrences. Finally, because the disease category of ARD was selected through an expert meeting based on ARDs proposed by Chang et al. [13] and published data, the DALY on the exponential increase in the incidence rate was calculated and not based on ARD.

Nevertheless, this study is meaningful as it represents a large-scale investigation based on a representative sample cohort of one million Koreans. To the best of our knowledge, it is the first research to explore the incidence rates of ARDs across the age spectrum, from 0 to 90 years, within the Korean population, demonstrating the presence of two distinct types of incidence rate curves for age-specific ARDs.

Conclusion

Our findings showed that the incidence rate of most diseases belonging to ARDs increased exponentially with age and exhibited a consistent pattern of peaking between the ages of 60 and 80 years, followed by a plateau or decrease. However, there were slight differences based on sex among Koreans. Understanding the general characteristics of ARDs and their disease burden could help develop public health policies for healthy aging.