Factors Associated with Heart Disease in Japan: Multivariate Analysis Based on Specific Health Checkups

Abstract

The global population affected by heart failure is projected to reach 30 million. The number of deaths due to heart disease has surged, rising from 2 million in 2000 to 8.9 million in 2019. In Japan, the prevalence of heart failure is rapidly increasing, with the number expected to reach 1.3 million by 2030. Primary prevention is crucial to prevent heart disease. We explored the associations of heart disease incidence with findings from checkups performed a decade ago. A multivariate logistic regression analysis revealed that individuals who reported a history of stroke, history of chronic renal failure, or weight gain ≥ 10 kg since age 20 in the questionnaire-based health checkup are at high risk for developing heart disease. Additionally, those with biochemical test results from 10 years ago indicating the use of antihypertensive drugs, use of insulin injections or hypoglycemic medications, systolic hypertension, and abnormal creatinine levels also exhibited a significantly higher risk of heart disease. Conversely, individuals who identified as female, walked faster than people of the same age, drank alcohol daily, and felt refreshed upon awakening in the questionnaire-based health checkup were protected from heart disease.

1. Introduction

The World Health Organization (WHO) has stated that the largest contributor of death worldwide from 2000 to 2019 was ischemic heart disease, responsible for 16% of all deaths. Since 2000, the number of deaths attributed to this condition has risen sharply, from 2 million to 8.9 million in 2019 [1].

Since 2000, ischemic heart disease has experienced the largest rise in deaths, from 2 million to 8.9 million by 2019. Over the past 20 years, deaths from this disease have surged by more than 2 million, reaching 8.9 million in 2019. It is estimated that ischemic heart disease accounted for 16% of all deaths in 2019.

Globally, 30 million individuals have heart failure. Based on the 2013–2016 National Health and Nutrition Examination Survey, more than 6.2 million individuals in the US have heart failure, a figure expected to rise to 8 million by 2030 [2]. In Japan, heart disease surpassed cerebrovascular disease as the second leading cause of mortality in 1985, responsible for 14.8% of deaths in 2022 [3].

In Japan, the prevalence of heart failure is rapidly increasing. According to the Trends in Newly Onset Heart Failure in Japan, fewer than 50,000 cases were recorded in 1950, but this number has steadily risen, surpassing 100,000 by 1990. The number of new cases is expected to continue growing, exceeding 350,000 by 2030. Consequently, the prevalence of heart failure is estimated to reach 1.3 million by 2030 [4]. The Vital Statistics published by Japan’s Ministry of Health, Labor, and Welfare in 2021 provides a detailed breakdown of heart disease-related deaths (excluding hypertension). It reveals that heart failure is the most common cause of mortality, significantly surpassing other conditions such as ischemic heart diseases, arrhythmias, conduction disorders, and acute myocardial infarction [5].

In Japan, heart disease is the sixth leading cause of the need for nursing care, accounting for 5.1% of all cases. However, when combined with stroke (16.1%), another cardiovascular disease, it constitutes 21% of cases, surpassing dementia at 16.6%, making cardiovascular disease the leading cause of the need for nursing care [6].

Therefore, effective standardized preventive strategies for heart disease are required. Similar to other chronic conditions, identifying risk factors is crucial for developing successful prevention strategies. Disease preventive strategies may be categorized as primary, secondary, and tertiary. To prevent heart disease, society must prioritize both primary and secondary strategies, necessitating longitudinal cohort studies to determine key lifestyle and health-related risk factors.

The Global Cardiovascular Risk Consortium identified certain modifiable risk factors of cardiovascular disease and all-cause mortality [7]: body mass index, systolic blood pressure (BP), low-density lipoprotein cholesterol levels, tobacco use, and diabetes, which contribute significantly to the burden of cardiovascular disease. However, the precise contribution of each factor varies according to the country and area [8,9].

According to a multinational prospective cohort study of 155,000 participants living in 21 high-income, middle-income, and low-income country regions followed for an average of 9 years and examining associations with potential risk factors for cardiovascular disease (CVD), more than 70% of CVD cases and deaths in the entire cohort were associated with modifiable risk factors [10]. This report examined the prevalence, hazard ratios, and population attributable fractions (PAFs) of CVD and mortality associated with behavioral factors (e.g., tobacco, alcohol, diet, physical activity, and sodium intake), metabolic factors (e.g., lipids, blood pressure, diabetes, and obesity), socioeconomic and psychosocial factors (e.g., education and depressive symptoms), physical fitness, household (solid cooking fuel), and ambient PM_2.5 air pollution. Metabolic factors were the main risk factor for CVD (41% of PAFs), with hypertension being the largest (22% of PAFs). As a cluster, behavioral risk factors contributed the most to death (26% of PAFs), while the largest single risk factor was a low level of education [10]. Thus, the majority of CVD cases and deaths are attributable to a small number of common modifiable risk factors. Factors such as hypertension and education level have a broad global impact, while other factors such as air pollution in the home and unhealthy diets vary by country’s economic level. Health policy should focus on those risk factors that have the greatest effect on avoiding CVD and death globally, with additional emphasis on the most important risk factors in specific country groups.

In 2008, Japan’s Ministry of Health, Labor, and Welfare initiated a specific health checkup (SHC) policy to screen high-risk groups for metabolic syndrome, aiming to prevent metabolic syndrome, promote lifestyle interventions, and reduce mid- to long-term medical expenses. The checkup included 30 parameters based on biochemical tests and a 24-item questionnaire [11]. In addition, the validity of the health checkups and questionnaire items is verified every four years by a research team of the Ministry of Health, Labor, and Welfare [12]. The questions in the specific health checkup are designed to understand the lifestyle habits of residents. By combining the residents’ answers with the health checkup results, it is possible to understand the relationship between lifestyle habits and health checkup results [12]. In Japan, universal health insurance was introduced in 1961. Anonymized data are available from the SHC and medical information submitted to the National Health Insurance [13], accessible for academic research to optimize medical costs and improve the quality of medical services [14,15,16,17]. Here, we analyzed heart disease risk factors by combining data from past SHCs with receipt data collected after 10 years.

To investigate the items that can be used for screening heart disease, we conducted an exploratory analysis of the relationship between the incidence rate of heart disease and the results of health checkups using 10 years of data from a Japanese SHC. We also examined whether Japan’s specific health checkups can be used as a primary prevention strategy for screening for heart disease.

2. Materials and Methods

2.1. Participants

Japanese National Health Insurance is a medical insurance system that covers all residents of Japan. It is a legally compulsory insurance based on the Japanese National Health Insurance Act. It is mainly run by municipalities and forms the core of Japan’s universal healthcare system. This study targeted all residents of Mishima City who are enrolled in National Health Insurance and Medical Insurance for the Elderly. Furthermore, among these subjects, we targeted residents who underwent specific health checkups in 2009. To exclude the impact of COVID-19 on heart disease, the study focused on data from the period before the pandemic. We enrolled 36,407 individuals with insurance from 107,000 residents of Mishima City, Japan, who had received medical care in the city in 2019 under the National Health Insurance program. Additionally, in 2009, 9,164 residents aged over 40 received SHCs that included 29 physical examinations, biochemical tests, and a 22-item questionnaire. From this group, we identified 6,587 participants (2,477 males and 4,110 females; age: 75.6 ± 6.73 years, range 49–85 years) who had received both medical treatment in 2019 and a standard health examination in 2009.

Heart disease was defined as residents receiving treatment for the following ICD-10 diagnoses: angina (I20.0, I20.1, I20.2, I20.3, I20.9), cardiac arrest (I21.0, I21.1, I21.8, I21.9), and cardiac infarct (I50.0, I50.1, I50.9). Individuals with a history of these heart diseases in 2009 were excluded from the study. We investigated the association between new-onset heart disease cases since 2009 and the results of the SHC survey conducted a decade prior.

2.2. Statistical Analysis

Logistic regression analyses were conducted to determine factors associated with heart disease. To control for potential confounders, multiple logistic regression analyses were employed to determine adjusted odds ratios (ORs). All SHC items were included together as explanatory variables. Participants were grouped based on the presence or absence of heart disease based on 2019 National Health Insurance data. Heart disease was defined according to the ICD-10 classifications described above.

For the multivariate logistic regression analysis, the presence or absence of heart disease was the dependent variable, and the independent variables were the results of the SHCs or questionnaire. Separate multivariate logistic regression analyses were performed for biochemical tests and questionnaire data. Potential common confounders such as age, sex, medication intake, and history of disease were included as explanatory variables in both analyses. In addition, a directed acyclic graph was created to visualize the causal relationships between variables. The type of directed acyclic graph is the Markov Blanket type of Bayesian network. Data analysis was carried out using SPSS software (version 27) and Modeler (version 18.3) (IBM Japan Ltd., Tokyo, Japan).

The study followed the International Ethical Guidelines for Epidemiology [18], the Guidelines for the Utilization of the Database for National Health Insurance Claims, SHC Guidance [19], and the Guidelines for Security of Health Information Systems [20]. Residents’ data were anonymized by local authorities to prevent personal identification.

3. Results

Table 1. Incidence rate of heart disease in 2019.

			Sex		Total
			Male	Female
Angina (I210, I211, I212, I213, I219)	Nonexistence	N	2371	3989	6360
		%	95.7%	97.1%	96.6%
	Existence	N	106	121	227
		%	4.3%	2.9%	3.4%
Cardiac arrest (I200, I201, I208, I209)	Nonexistence	N	2068	3499	5567
		%	83.5%	85.1%	84.5%
	Existence	N	409	611	1020
		%	16.5%	14.9%	15.5%
Cardiac infarct (I500, I501, I509)	Nonexistence	N	1978	3437	5415
		%	79.9%	83.6%	82.2%
	Existence	N	499	673	1172
		%	20.1%	16.4%	17.8%
Total Heart Disease	Nonexistence	N	1720	2988	4718
		%	69.4%	72.9%	71.6%
	Existence	N	757	1112	1869
		%	30.6%	27.1%	28.4%
Total		N	2477	4110	6587
		%	100.0%	100.0%	100.0%

presents the number of heart disease cases and the incidence rate of each type of heart disease in 2019.

Table 2. Cross-tabulation results of a questionnaire survey conducted in 2009 and the incidence rate of heart disease in 2019.

		Incidence of Heart Disease
		Nonexistence	Existence	Total	p-Value
Taking blood pressure-lowering medication	Yes	1473	844	2317	<0.001
		63.6%	36.4%	100.0%
	No	3245	1025	4270
		76.0%	24.0%	100.0%
Drugs to lower blood sugar or insulin injections	Yes	314	195	509	<0.001
		61.7%	38.3%	100.0%
	No	4404	1674	6078
		72.5%	27.5%	100.0%
Taking cholesterol-lowering medication	Yes	1077	552	1629	0.001
		66.1%	33.9%	100.0%
	No	3641	1317	4958
		73.4%	26.6%	100.0%
Stroke history	Yes	184	147	331	<0.001
		55.6%	44.4%	100.0%
	No	4534	1722	6256
		72.5%	27.5%	100.0%
History of chronic kidney failure	Yes	4	5	9	0.070
		44.4%	55.6%	100.0%
	No	4714	1864	6578
		71.7%	28.3%	100.0%
Anemia history	Yes	445	172	617	0.774
		72.1%	27.9%	100.0%
	No	4273	1697	5970
		71.6%	28.4%	100.0%
Currently smoking regularly	Yes	538	216	754	0.860
		71.4%	28.6%	100.0%
	No	4180	1653	5833
		71.7%	28.3%	100.0%
WT gain about 10 kg or more since age 20	Yes	1365	657	2022	<0.001
		67.5%	32.5%	100.0%
	No	3353	1212	4565
		73.5%	26.5%	100.0%
Exercise for 30 min at least two days a week	Yes	2514	1011	3525	0.553
		71.3%	28.7%	100.0%
	No	2204	858	3062
		72.0%	28.0%	100.0%
Walk for more than an hour every day	Yes	2412	408	2820	0.443
		85.5%	14.5%	100.0%
	No	2620	469	3089
		84.8%	15.2%	100.0%
Walking faster than peers of the same age and gender	Yes	2688	972	3660	<0.001
		73.4%	26.6%	100.0%
	No	2030	897	2927
		69.4%	30.6%	100.0%
WT loss or gain 3 kg or more in the past year	Yes	817	314	1131	0.616
		72.2%	27.8%	100.0%
	No	3901	1555	5456
		71.5%	28.5%	100.0%
Eating pace	Faster	1126	439	1565	0.913
		71.9%	28.1%	100.0%
	Normal	3247	1289	4536
		71.6%	28.4%	100.0%
	Slower	345	141	486
		71.0%	29.0%	100.0%
Dinner 2 h before bedtime	Yes	528	225	753	0.330
		70.1%	29.9%	100.0%
	No	4190	1644	5834
		71.8%	28.2%	100.0%
Eat a snack after dinner at least three times a week	Yes	4260	1692	5952	0.769
		71.6%	28.4%	100.0%
	No	458	177	635
		72.1%	27.9%	100.0%
Skipping breakfast 3 or more days a week	Yes	292	99	391	0.167
		74.7%	25.3%	100.0%
	No	4426	1770	6196
		71.4%	28.6%	100.0%
Drink alcohol	Non-drinker	2657	1116	3773	0.043
		70.4%	29.6%	100.0%
	Sometimes	1083	395	1478
		73.3%	26.7%	100.0%
	Everyday	978	358	1336
		73.2%	26.8%	100.0%
Feel refreshed upon awakening	Yes	3823	1472	5295	0.036
		72.2%	27.8%	100.0%
	No	895	397	1292
		69.3%	30.7%	100.0%
Lifestyle changes	No plans for improvement	1449	565	1258	0.50
		71.9%	28.1%	100.0%
	Planning to start within 6 months	1125	469	486
		70.6%	29.4%	100.0%
	Planning to start within a months	692	287	659
		70.7%	29.3%	100.0%
	Already started within 6 months	420	146	1705
		74.2%	25.8%	100.0%
	Already started over 6 months ago	1032	402	1801
		72.0%	28.0%	100.0%
Willing to receive health guidance	Yes	2199	857	3056	0.579
		72.0%	28.0%	100.0%
	No	2519	1012	3531
		71.3%	28.7%	100.0%
	Total	4718	1869	6587
		71.6%	28.4%	100.0%

p-value: Chi-square test.

presents the findings from the cross-tabulation of questionnaires administered in 2009 and incidence of heart disease in 2019, which showed significant differences in eight items: “Taking blood pressure lowering medication”, “Drugs to lower blood sugar or insulin injections”, “Taking cholesterol-lowering medication”, “Stroke history”, “WT gaine about 10kg or more since age 20”, “Walking faster than peers of the same age and gender”, “Drink alcohol every day”, and “Feel refreshed upon awakening”.

We cross-tabulated the biochemical tests in 2009 and incidence of heart disease in 2019 (

Table 3. Cross-tabulation results of biochemical tests in 2009 and the incidence rate of heart disease in 2019.

		Incidence of Heart Disease
Item		Nonexistence	Existence	Total	p-Value
Uric protein	Normal	4480	1724	6204	<0.001
		72%	28%	100%
	Follow-up	148	79	227
		65%	35%	100%
	Requires further testing	68	40	108
		63%	37%	100%
	Requires treatment	22	26	48
		46%	54%	100%
Urinary sugar	Normal	4630	1821	6451	0.153
		72%	28%	100%
	Follow-up	15	10	25
		60%	40%	100%
	Requires further testing	19	6	25
		76%	24%	100%
	Requires treatment	54	32	86
		63%	37%	100%
Uric blood	Normal	3990	1594	5584	0.058
		71%	29%	100%
	Follow-up	458	153	611
		75%	25%	100%
	Requires further testing	260	113	373
		70%	30%	100%
	Requires treatment	10	9	19
		53%	47%	100%
Creatinine	Normal	4503	1727	6230	<0.001
		72%	28%	100%
	Follow-up	215	142	357
		60%	40%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
Urea nitrogen	Normal	4571	1806	6377	0.595
		72%	28%	100%
	Follow-up	147	63	210
		70%	30%	100%
	Requires further testing	4718	1869	6587
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
Urinary acid	Normal	4329	1676	6005	0.007
		72%	28%	100%
	Follow-up	389	193	582
		67%	33%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
Leucocyte	Normal	4486	1783	6269	0.590
		72%	28%	100%
	Follow-up	232	86	318
		73%	27%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
Erythrocyte	Normal	3224	1264	4488	0.331
		72%	28%	100%
	Follow-up	1351	535	1886
		72%	28%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	143	70	213
		67%	33%	100%
Hemoglobin	Normal	4281	1692	5973	0.646
		72%	28%	100%
	Follow-up	344	145	489
		70%	30%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	93	32	125
		74%	26%	100%
Hematocrit	Normal	4479	1752	6231	0.950
		72%	28%	100%
	Follow-up	213	104	317
		67%	33%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	26	13	39
		67%	33%	100%
Platelet	Normal	4593	1799	6392	0.018
		72%	28%	100%
	Follow-up	125	70	195
		64%	36%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
AST(GOT)	Normal	4200	1647	5847	0.430
		72%	28%	100%
	Follow-up	482	203	685
		70%	30%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	36	19	55
		65%	35%	100%
ALT(GPT)	Normal	4090	1619	5709	0.954
		72%	28%	100%
	Follow-up	548	220	768
		71%	29%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	80	30	110
		73%	27%	100%
γGTP	Normal	4074	1600	5674	0.563
		72%	28%	100%
	Follow-up	469	190	659
		71%	29%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	175	79	254
		69%	31%	100%
Amylase	Normal	4468	1759	6227	0.345
		72%	28%	100%
	Follow-up	250	110	360
		69%	31%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
ALP	Normal	4528	1782	6310	0.252
		72%	28%	100%
	Follow-up	190	87	277
		69%	31%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
LDL-cholesterol	Normal	1595	673	2268	0.090
		70%	30%	100%
	Follow-up	3123	1196	4319
		72%	28%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
Total protein	Normal	4610	1812	6422	0.075
		72%	28%	100%
	Follow-up	108	57	165
		65%	35%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	0	0	0
		0%	0%	0%
Total-cholesterol	Normal	2112	867	2979	0.335
		71%	29%	100%
	Follow-up	2042	809	2851
		72%	28%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	564	193	757
		75%	25%	100%
HDL-cholesterol	Normal	4526	1768	6294	0.041
		72%	28%	100%
	Follow-up	140	69	209
		67%	33%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	52	32	84
		62%	38%	100%
Neutral fat	Normal	3744	1450	5194	0.060
		72%	28%	100%
	Follow-up	881	366	1247
		71%	29%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	93	53	146
		64%	36%	100%
Blood glucose level	Normal	3383	1287	4670	0.048
		72%	28%	100%
	Follow-up	1041	443	1484
		70%	30%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	294	139	433
		68%	32%	100%
HbA1C	Normal	2586	914	3500	<0.001
		74%	26%	100%
	Follow-up	1785	775	2560
		70%	30%	100%
	Requires further testing	0	0	0
		0%	0%	0%
	Requires treatment	347	180	527
		66%	34%	100%
Total		4718	1869	6587
		72%	28%	100%

p-value: Chi-square test.

); the p-value was calculated by a chi-square test and seven items were identified as significantly different: “Uric protein”, “Creatinine”, “Urinary acid”, “Platelet”, “HDL-cholesterol”, “Blood glucose level”, and “HbA1C”.

Table 4. Multivariate logistic regression analysis of factors based on questionnaires.

Item		Crude Odds Ratio	95% Confidence Interval		p-Value	Multivariate Adjusted Odds Ratio	95% Confidence Interval		p-Value
			Lower	Upper			Lower	Upper
Age		1.074	1.063	1.084	<0.001	1.072	1.059	1.085	<0.001
Gender	(F/M)	0.843	0.755	0.941	0.002	0.765	0.655	0.892	0.001
Taking blood pressure-lowering medication	(+/−)	1.814	1.625	2.025	<0.001	1.452	1.270	1.660	<0.001
Drugs to lower blood sugar or insulin injections	(+/−)	1.634	1.355	1.970	<0.001	1.458	1.178	1.804	0.001
Taking cholesterol-lowering medication	(+/−)	1.417	1.256	1.598	<0.001	0.956	0.823	1.110	0.553
Stroke history	(Yes/No)	2.104	1.682	2.631	<0.001	1.303	1.008	1.684	0.022
History of chronic kidney failure	(Yes/No)	3.161	0.848	11.785	0.086	6.261	1.540	25.463	0.010
Anemia history	(Yes/No)	0.973	0.809	1.171	0.774	1.149	0.921	1.432	0.347
Currently smoking regularly	(Yes/No)	1.015	0.858	1.201	0.860	1.112	0.900	1.372	0.315
WT gain about 10 kg or more since age 20	(Yes/No)	1.332	1.188	1.492	<0.001	1.249	1.087	1.437	<0.001
Exercise for 30 min at least two days a week	(Yes/No)	1.024	0.920	1.140	0.667	0.980	0.849	1.132	0.871
Walk for more than an hour every day	(Yes/No)	1.033	0.928	1.150	0.553	1.015	0.881	1.169	0.840
Walking faster than peers of the same age and gender	(Yes/No)	0.818	0.735	0.911	<0.001	0.841	0.737	0.960	0.003
WT loss or gain 3 kg or more in the past year		0.964	0.836	1.112	0.616	0.980	0.826	1.162	0.828
Eating pace     Slower	Reference Group								0.400
Normal		0.982	0.864	1.116	0.782	1.019	1.056	0.813	0.683
Faster		1.030	0.838	1.265	0.782	0.929	0.952	0.750	0.687
Dinner 2 h before bedtime	(Yes/No)	0.921	0.780	1.087	0.330	0.914	0.748	1.117	0.380
Eat a snack after dinner at least three times a week	(Yes/No)	0.973	0.811	1.168	0.769	1.059	0.850	1.318	0.610
Skipping breakfast 3 or more days a week	(Yes/No)	0.848	0.671	1.072	0.168	1.047	0.786	1.394	0.753
Drink alcohol     Non-drinker	Reference Group								0.070
Sometimes		0.966	1.004	0.849	1.186	0.897	0.761	1.057	0.195
Everyday		0.040	1.152	1.006	1.318	0.805	0.666	0.973	0.025
Feel refreshed upon awakening	(Yes/No)	0.868	0.760	0.991	0.036	0.769	0.656	0.900	0.001
Lifestyle changes									0.360
No plans for improvement	Reference Group
Planning to start within 6 months		0.288	1.122	0.907	1.386	1.143	0.955	1.366	0.144
Planning to start within a months		0.101	1.199	0.965	1.490	1.022	0.831	1.257	0.838
Already started within 6 months		0.138	1.193	0.945	1.506	0.888	0.688	1.148	0.365
Already started over 6 months ago		0.312	1.121	0.899	1.397	1.018	0.846	1.224	0.854
Willing to receive health guidance	(Yes/No)	0.970	0.871	1.080	0.579	0.924	0.811	1.054	0.239
Outpatient medical expenses (2009)		1.000	1.000	1.000	<0.001	1.000	1.000	1.000	0.104
_cons						0.005			<0.001

In total, 24 items were used simultaneously as independent variables. (Hosmer–Lemeshow test: p-value = 0.857, Cox–Snel R-squared values: 0.0878).

summarizes the logistic regression analysis for factors associated with the prevalence of heart disease based on questionnaires administered in 2009.

The crude Odds ratios (ORs) revealed associations for seven items. To account for potential confounders, all explanatory variables potentially related to heart disease were included in the multivariate logistic regression analysis, irrespective of significance in the univariate analysis. The multivariate logistic regression analysis identified eight significant ORs associated with the incidence of heart disease.

Residents with a history of stroke (OR: 1.71, 95% Confidence Interval (CI): 1.05–2.80), history of chronic renal failure (OR: 6.26, 95% CI: 1.54–25.46), and weight gain ≥ 10 kg since age 20 (OR: 1.25, 95% CI: 1.09–1.44) were prone to developing heart disease. Conversely, residents identified as female (OR: 0.77, 95% CI: 0.66–0.89), those who walk faster than people of their age (OR: 0.84, 95% CI: 0.74–0.96), those who drink alcohol every day (OR: 0.81, 95% CI: 0.67–0.97), and those who feel refreshed upon awakening (OR: 0.77, 95% CI: 0.666–0.90) had diminished chances of heart disease.

Table 5. Multivariate logistic regression analysis of biochemical parameters.

Item	Crude Odds Ratio	95% Confidence Interval		p-Value	Multivariate Adjusted Odds Ratio	95% Confidence Interval		p-Value
		Lower	Upper			Lower	Upper
Age	1.074	1.063	1.084	<0.001	1.067	1.053	1.080	<0.001
Gender (W/M)	0.843	0.755	0.941	<0.001	0.874	0.703	1.087	0.225
Height (cm)	0.997	0.990	1.003	0.002	0.996	0.982	1.009	0.545
Weight (kg)	1.010	1.004	1.015	0.342	1.008	0.993	1.022	0.293
Abdominal circumference(cm)	1.020	1.014	1.026	<0.001	1.004	0.992	1.017	0.491
Taking blood pressure-lowering medication (+/−)	1.814	1.625	2.025	<0.001	1.364	1.186	1.569	<0.001
Drugs to lower blood sugar or insulin injections (+/−)	1.634	1.355	1.970	<0.001	1.582	1.234	2.028	<0.001
Taking cholesterol-lowering medication (Yes/No)	1.417	1.256	1.598	<0.001	0.961	0.825	1.119	0.608
Systolic blood pressure (mmHg)	1.014	1.010	1.017	<0.001	1.007	1.001	1.012	0.019
Diastolic blood pressure (mmHg)	1.008	1.002	1.013	<0.001	0.995	0.987	1.004	0.303
Uric protein (+±/−)	1.585	1.280	1.962	<0.001	1.216	0.946	1.564	0.127
Urinary sugar (+±/−)	1.388	0.973	1.982	<0.001	1.135	0.732	1.759	0.571
Uric blood (+±/−)	0.947	0.814	1.100	0.071	0.933	0.778	1.118	0.451
Creatinine (+±/−)	1.723	1.384	2.145	0.474	1.440	1.117	1.856	0.005
Urea nitrogen (+±/−)	1.084	0.803	1.464	<0.001	0.933	0.660	1.319	0.695
Urinary acid (+±/−)	1.282	1.069	1.537	0.597	1.193	0.956	1.488	0.119
Leucocyte (+±/−)	0.933	0.724	1.202	<0.001	0.934	0.682	1.279	0.672
Erythrocyte (+±/−)	1.033	0.921	1.158	0.590	1.108	0.959	1.281	0.164
Hemoglobin (+±/−)	1.025	0.853	1.231	0.580	1.085	0.864	1.363	0.482
Hematocrit (+±/−)	1.252	0.997	1.573	0.794	1.088	0.814	1.455	0.569
Platelet (+±/−)	1.431	1.062	1.926	0.053	1.093	0.766	1.560	0.623
AST(GOT) (+±/−)	1.093	0.925	1.292	0.018	0.867	0.677	1.111	0.260
ALT(GPT) (+±/−)	1.006	0.859	1.177	0.298	0.944	0.742	1.199	0.635
γGTP (+±/−)	1.064	0.912	1.240	0.944	1.007	0.821	1.234	0.949
Amylase (+±/−)	1.117	0.887	1.408	0.432	1.051	0.802	1.377	0.720
ALP (+±/−)	1.164	0.898	1.509	0.347	0.922	0.673	1.262	0.610
LDL-cholesterol (+±/−)	0.908	0.811	1.015	0.252	0.904	0.767	1.065	0.227
Total protein (+±/−)	1.345	0.971	1.862	0.090	1.237	0.849	1.800	0.268
Total-cholesterol (+±/−)	0.936	0.841	1.042	0.074	1.072	0.917	1.254	0.382
HDL-cholesterol (+±/−)	1.347	1.052	1.725	0.229	1.014	0.753	1.364	0.928
Neutral fat (+±/−)	1.111	0.976	1.264	0.018	1.013	0.861	1.191	0.878
Blood glucose level (+±/−)	1.084	1.016	1.157	0.112	0.915	0.831	1.009	0.074
HbA1C (+±/−)	1.270	1.140	1.414	0.015	1.043	0.907	1.200	0.557
_cons				<0.001	0.001			<0.001

In total, 34 items were used simultaneously as independent variables. (Hosmer–Lemeshow test: p-value = 0.996, Cox–Snel R-squared values: 0.081).

summarizes the logistic regression analysis for variables related to the prevalence of heart disease based on biochemical tests conducted in 2009.

The crude ORs indicated significant associations for 13 items. However, the multivariate logistic regression analysis identified five significant ORs for the incidence of heart disease. Residents who were prescribed medications to lower blood pressure(BP) (OR: 1.36, 95% CI: 1.19–1.57), used insulin injections or medications to lower blood glucose (OR: 1.58, 95% CI: 1.23–2.03), had higher systolic BP (OR: 1.01, 95% CI: 1.00–1.01), or had abnormal creatinine levels (OR: 1.44, 95% CI: 1.12–1.86) were at a higher risk of developing heart disease.

Figure 1 and Figure 2 show the directed acyclic graphs of the questionnaire and biochemical test.

4. Discussion

The rise in the prevalence of heart failure in Japan is largely attributed to the aging population. Japan is experiencing an unprecedented rate of aging. According to the 2024 edition of the World Health Statistics published by the WHO, Japan has the highest life expectancy in the world, at 84.5 years.

By gender, Israel ranks first for men with a life expectancy of 82.4 years, while Japan ranks first for women with a life expectancy of 87.2 years. Japanese men are ranked second at 81.7 years [21]. A major challenge unique to our super-aging society is the sharp increase in elderly heart failure patients, often referred to as the heart failure pandemic. This number is expected to increase to 1.3 million nationwide in 2030 [5].

In this study, a history of stroke was found to be associated with heart disease. It has been reported that arrhythmias, such as atrial fibrillation, are significantly linked to an increased risk of stroke [22]. Therefore, patients who had a history of stroke due to undiagnosed arrhythmias in 2009 may have developed heart failure 10 years later. Additionally, a history of chronic renal failure was associated with heart disease in this study. Previous random-effects meta-analyses have indicated that chronic kidney disease and worsening renal function are linked to adverse outcomes in heart failure, which is consistent with our findings [23].

Furthermore, this study found that individuals who answered “weight gain ≥ 10 kg since age 20” had significantly lower ORs for developing heart disease. Recent Mendelian randomization studies have demonstrated that each 1 kg/m² increase in body mass index predisposes to several conditions, including pulmonary embolism, coronary heart disease, peripheral arterial disease, atrial fibrillation, hypertension, deep vein thrombosis, heart failure, and aortic stenosis [24].

Additionally, this study found that individuals who walked faster than their peers of the same sex and age had significantly lower ORs for developing heart disease. Previous research has shown that increasing the number of daily steps can reduce the risk of heart disease and mortality [25,26]. However, more recent investigations have highlighted that walking speed, rather than just the number of steps taken, is significantly linked to the mortality risk after adjusting for step count [27]. These findings are consistent with our results.

This study found that individuals who drink alcohol daily had significantly lower ORs for developing heart disease. Epidemiological research has demonstrated that low to moderate alcohol intake can minimize the risk of heart disease [28,29,30]. However, the impact of alcohol on cardiovascular health is complex and multifaceted. It has been suggested that the relationship between alcohol intake and overall cardiovascular disease risk may not follow a simple J- or U-shaped curve, but rather involve multiple opposing dose–response curves [31]. Additionally, the impact of alcohol on cardiovascular risk varies by type; for instance, beer and spirits have been associated with higher risks compared to wine [31]. Our study focused on the presence, absence, and frequency of alcohol consumption, but future research should include more detailed analyses of the type and quantity of alcohol consumed.

In this study, feeling refreshed upon awakening was significantly associated with heart disease. A dose–response meta-analysis of prospective studies has shown a U-shaped association between sleep duration and the risk of all-cause mortality, cardiovascular diseases, coronary heart disease, and stroke, with the lowest risk associated with approximately 7 h of sleep. Both shorter and longer sleep durations were significantly linked to increased risks of all-cause mortality, cardiovascular diseases, coronary heart disease, and stroke [32,33]. Further long-term randomized controlled trials are imperative to clarify the causal relationship and underlying mechanisms between sleep duration and the development of heart disease.

Among biochemical tests, significantly higher ORs were found for taking medication to lower BP and having high systolic BP. Hypertension predisposes to cardiovascular and renal events, both fatal and nonfatal, including myocardial infarction, stroke, atherosclerosis, aortic aneurysm, hypertensive heart disease, heart failure, peripheral arterial disease, and end-stage renal disease [34]. A recent meta-analysis of individual participant-level data demonstrated that a 5 mmHg reduction in systolic BP on average reduced the risk of major cardiovascular events by approximately 10%. The corresponding proportional risk reductions were 13% for stroke, 13% for heart failure, 8% for ischemic heart disease, and 5% for cardiovascular death [35]. However, it has been reported that aiming to lower BP to a uniform threshold for all individuals may not be ideal [35]. Therefore, the optimal level of BP reduction should be determined.

In this study, biochemical tests revealed a significantly higher OR for the use of insulin or medicines to lower blood glucose. A meta-analysis of 77 prospective studies found that diabetes leads to a two-fold increased risk of heart failure in the general population [36]. Additionally, elevated blood glucose, even within the pre-diabetic range, also elevates the risk of heart failure [36]. Another meta-analysis of individual records has shown that approximately 10% of vascular deaths in the developed world over the past decade were attributable to diabetes in adults, equating to about 325,000 deaths annually in high-income countries alone [37]. This burden is expected to rise if the incidence of diabetes continues to increase. Moreover, in the biochemical tests of this study, individuals with abnormal creatinine levels had a significantly higher OR for developing heart disease 10 years later. Previous meta-analyses have highlighted that elevated serum creatinine and associated changes in glomerular filtration rate are linked to increased mortality in heart failure patients [23]. However, reduced serum creatinine levels do not necessarily translate to improved survival [23]. Future research should focus on identifying at-risk patients, accurately defining or calculating changes in renal function, and emphasizing the maintenance or improvement in renal function over time to prevent heart failure.

As a limitation of the research, this study focuses on one city rather than the data for the whole of Japan. In addition, in this study, the multivariate analysis used all items from the health checkups and questionnaires as an exploratory study, but further verification analysis of the relationship between each variable is necessary.

To prevent heart disease in Japan in the future, it is essential to accurately identify the risks associated with lifestyle-related diseases. By clarifying these risks, it may be possible to leverage the existing SHCs and guidance programs aimed at preventing metabolic syndrome for heart disease prevention as well. Therefore, future efforts should focus on identifying specific health management items related to heart disease risk through larger sample sizes and long-term studies.

5. Conclusions

We found that individuals who reported a history of stroke, history of chronic renal failure, or weight gain ≥ 10 kg since age 20 in the questionnaire-based health checkup are at high risk for developing heart disease. Additionally, those with biochemical test results from 10 years ago indicating the use of antihypertensive drugs, use of insulin injections or hypoglycemic medications, systolic hypertension, and abnormal creatinine levels also exhibited a significantly higher risk of heart disease. Conversely, individuals who identified as female, walked faster than people of the same age, drank alcohol daily, and felt refreshed upon awakening had a reduced prevalence of heart disease. Long-term studies with more participants are required to determine health management criteria that can effectively determine the heart disease risk.