Risk Factors for Stroke Based on the National Health and Nutrition Examination Survey

Abstract

Objectives

To explore the risk factors for stroke by using National Health and Nutrition Examination Survey (NHANES).

Methods

A total of 19384 cases from the NHANES database during 2005 to 2016 were included in this study, of which 661 were stroke patients, with a weighted prevalence of 2.60%. Univariate and multivariate logistic regression analysis was used to analyze possible risk factors for stroke.

Results

The risk of stroke was 1.839 times higher in subjects with diabetes than in those without diabetes, P<0.001. An average increase of 1 ug/dL in blood lead was associated with a 1.082-fold increase in stroke risk, P<0.001. Subjects with mild-, moderate-, and moderately severe or severe depression had a 1.567-fold, 1.836-fold, and 3.279-fold higher risk than those without depression respectively, P<0.001. The risk of stroke in subjects with sleep disorders was 1.622 times higher than those without sleep disorders, P<0.001.

Conclusions

Patients with diabetes, lower household income (<$2000), PHQ-9 depression score (5–27), trouble sleep, older age and higher concentration of blood lead are associated with a higher risk of stroke.

Introduction

Stroke occurs for a variety of reasons, interfering the blood flow in the brain and then damaging brain tissue in the affected area, causing central nervous system injury (1, 2). In the light of World Health Organization criteria, the definition of stroke is a rapidly developing signs of continuous disturbance of cerebral function >24 h (other than interrupted by treatment or death) with no apparent except vascular cause, and the stroke is considered as the second leading cause of death all over the world which can bring patients severe, long-term disability (3, 4, 5).

Every year, 86.5 per 100000 people died from stroke around the world (6). Stroke survivors suffer from the neurological injury after stroke which brings them the dysfunction of strength, cognition and speech (7). More than 75% of stroke survivors suffer from a disability which makes it difficult for them to find jobs, and more than 80% of survivors who acquired motor dysfunction need rehabilitation treatments (5). Besides neurological impairments, various of medical complications can occur with stroke (8). The stroke related complications during the rehabilitation process can reduce the chance for improvement, reported by Janus-Laszuk B et al., and the knowledge of early prevention and treatment is benefit to the outcomes improvement of stroke patients (9).

Considering the high mortality and disability, the prevention and treatment for stroke patients are essential. Current study generally suggests that the occurrence of stroke is related to various risk factors including smoking or drinking, hypertension, diabetes, dyslipidemia, etc. which can be divided into controllable factors and uncontrollable factors. Special emphasis should be placed on risk factors identification for stroke, and the influence of identified risk factors could be diminished or eliminated by lifestyle changes or medication. The present study was conducted to explore the risk factors for stroke by using National Health and Nutrition Examination Survey (NHANES), a large nationally representative database, with the aim of providing references for clinical diagnosis, treatment and prevention of stroke.

Methods

Patients and materials

In the present study, a total of 19384 subjects was retrieved from the NHANES database from 2005 to 2016. The NHANES conducted by the national center for health statistics was designed to assess the health and nutritional status of different population groups in the US (10).

Definition of stroke: (1) Has a doctor or other health professionals ever told you that you had a stroke; (2) NHANES (11); (3) the Patient Health Questionnaire (PHQ-9): Starting from 2005, NHANES has included PHQ-9 to measure signs and symptoms of depression. Total PHQ-9 score ranges from 0 to 27 and are categorized as “none or minimum” (0–4), “mild” (5–9), “moderate” (10–14), “moderately severe” (15–19), and “severe” (20–27) for depression severity (12).

Collection of baseline information

Variables including marital status, gender, age, total cholesterol (mg/dL), blood lead (µg/dL), body mass index (BMI), systolic blood pressure (SBP), diabetes, high-density lipoprotein (HDL) (mg/dL), medical insurance, PHQ-9 entry content (12), household income ($) and sleep status were retrieved in this study.

Statistical analysis

All Statistical tests adopted two-side test, P<0.05 was considered as statistically significant. Statistical analysis was completed by SPSS statistical professional analysis software (version 16.0), and data combination was completed by SAS software (SAS Campus Drive, Cary, North Carolina 27513, USA). Mean±standard deviation was used to describe the measurement data of normal distribution, and independent sample t-test was used for inter-group comparison. Median and interquartile range (M (Q1, Q3)) were used for description of abnormal data, the Mann-Whitney U rank sum test was used for comparison between groups. Univariate and multivariate logistic regression analysis was used to analyze possible risk factors for stroke.

Results

Patients

A total of 19384 cases were included in this study, including 9600 (49.50%) males and 9784 (50.50%) females. There were 661 cases of stroke, and the incidence of stroke was 2.60%, patients with stroke were divided into the stroke group, others were divided into the control group. The stroke group had older age (48.52 vs 65.78, P<0.001) and lower proportion of family income ≥$20,000 (79.00 vs 64.90, P<0.001) than the control group. There was no statistical difference in age composition between the two groups. The blood lead level and PHQ-9 score in the stroke group were higher than those in the control group (1.27 vsl.75, P<0.001; 2.00 vs 3.00, P<0.001) (Table 1).

Table 1.

Basic characteristics of the included study

Variables	Stroke group (n=18723)	Control group (n=661)	Stastatic	P
Material status			214.28	<0.001
Married	9898(52.90)	331(50.10)
Widowed	1374(7.30)	137(20.70)
Divorce/separation	2617(14.00)	122(18.50)
Unmarried	4837(25.80)	71(10.70)
Household income ($)			75.84	<0.001
<20000	3924(21.00)	232(35.10)
≥20000	14802(79.00)	429(64.90)
Diabetes			294.11	<0.001
No	16317(87.10)	425(64.30)
Borderline	383(2.00)	25(3.80)
Yes	2026(10.80)	211(31.90)
Covered by health insurance			73.15	<0.001
No	4395(23.50)	61(9.20)
Yes	14331(76.50)	600(90.80)
Sleep duration (h)			36.69	<0.001
≤5	2873(15.30)	132(20.00)
6–8	14520(77.50)	450(68.10)
≥9	1333(7.10)	79(12.00)
PHQ-9 depression score	2.00(0.00,4.00)	3.00(1.00,7.00)	Z=9.78	<0.001
0–4	14384(76.80)	411(62.20)	107.78	<0.001
5–9	2798(14.90)	136(20.60)
10–14	980(5.20)	57(8.60)
15–27	564(3.00)	57(8.60)
Trouble sleep			111.50	<0.001
No	14315(76.40)	387(58.50)
Yes	4411(23.60)	274(41.50)
Gender			0.04	0.838
Male	9273(49.50)	330(49.90)
Female	9453(50.50)	331(50.10)
SBP	123.09±18.53	131.61±21.68	t=11.54	<0.001
BMI	29.01±6.73	30.01±6.65	t=3.75	<0.001
Age	48.52±17.65	65.78±13.22	t=24.89	<0.001
Blood lead (ug/dL)	1.27(0.81,2.00)	1.75(1.19,2.72)	Z=13.27	<0.001
HDL (mg/dL)	53.05±16.08	50.88±16.29	t=3.41	0.001
Total Cholesterol (mg/dL)	196.09±41.73	188.71±46.87	t=4.45	<0.001

PHQ: Patient Health Questionnaire; SBP: Systolic blood pressure; BMI: Body mass index; HDL: High-density lipoprotein.

Univariate logistic regression analysis of stroke risk factors

As shown in Tale 2, univariate analysis showed that compared with the married, the widowed (OR=2.98, P<0.001) and divorced (OR=1.39, P=0.002) were risk factors for stroke, and unmarried was a protective factor (OR=0.44, P<0.001). The risk of stroke was 0.49 times higher in subjects with the household income ≥$20000 than those with the income <$20000, (P<0.001). Subjects with mild-, moderate-, and moderately severe or severe depression had a 1.70-fold, 2.04-fold, and 3.54-fold higher risk for stroke than those without depression respectively, P<0.001. Compared with sleep duration ≤5h, the sleep duration of 6.8h was a protective factor for stroke, P<0.001. An average increase of 1 ug/dL in blood lead levels was associated with a 1.119-fold increase in stroke risk, P<0.001.

Table 2.

Logistic regression analysis of stroke risk factors

Variables	Single factor		Multi-factor
	OR (95% CI)	P	OR (95% CI)	P
Material status
Married	(Ref)		(Ref)
Widowed	2.98(2.42,3.66)	<0.001	1.045(0.825,1.323)	0.714
Divorce/separation	1.39(1.13,1.72)	0.002	1.151(0.913,1.451)	0.234
Unmarried	0.44(0.34,0.56)	<0.001	1.050(0.790,1.396)	0.736
Household income ($)
<20000	(Ref)		-
≥20000	0.49(0.42,0.58)	<0.001	0.749(0.623,0.900)	0.002
Diabetes
No	(Ref)		(Ref)
Borderline	2.51(1.65,3.79)	<0.001	1.417(0.922,2.177)	0.112
Yes	3.99(3.37,4.75)	<0.001	1.839(1.521,2.223)	<0.001
Covered by health insurance
No	(Ref)		(Ref)
Yes	0.33(0.25,0.43)	<0.001	0.738(0.553,0.984)	0.038
Sleep duration (h)
≤5	(Ref)		(Ref)
6–8	0.68(0.55,0.82)	<0.001	0.949(0.764,1.178)	0.634
≥9	1.29(0.97,1.72)	0.081	1.275(0.937,1.735)	0.122
PHQ-9 depression score
none or minimum (0–4)	(Ref)		(Ref)
mild (5–9)	1.70(1.40,2.07)	<0.001	1.567(1.266,1.941)	<0.001
moderate (10–14)	2.04(1.53,2.71)	<0.001	1.836(1.346,2.504)	<0.001
moderately severe or
severe (15–27)	3.54(2.65,4.73)	<0.001	3.279(2.370,4.536)	<0.001
Trouble sleep
No	(Ref)		(Ref)
Yes	2.29(1.96,2.69)	<0.001	1.622(1.358,1.937)	<0.001
Gender
Male	(Ref)		(Ref)
Female	0.98(0.84,1.15)	0.838	1.015(0.845,1.219)	0.871
SBP	1.02(1.01,1.02)	<0.001	1.001(0.997,1.005)	0.559
BMI	1.02(1.01,1.03)	<0.001	1.006(0.994,1.019)	0.329
Age	1.06(1.05,1.07)	<0.001	1.060(1.052,1.068)	<0.001
Blood lead (ug/dL)	1.119(1.089,1.150)	<0.001	1.082(1.048,1.117)	<0.001
HDL (mg/dL)	0.991(0.986,0.996)	0.001	0.992(0.986,0.997)	0.005
Total Cholesterol (mg/dL)	0.996(0.994,1.002)	0.069	0.998(0.996,1.001)	0.057

PHQ: Patient Health Questionnaire; SBP: Systolic blood pressure; BMI: Body mass index; HDL: High-density lipoprotein.

Multivariate logistic regression analysis of stroke risk factors

Multivariate analysis demonstrated that the risk of stroke was 1.839 times higher in subjects with diabetes than in those without diabetes, P<0.001. An average increase of 1 ug/dL in blood lead was associated with a 1.082-fold increase in stroke risk, P<0.001. Subjects with mild-, moderate-, and moderately severe or severe depression had a 1.567-fold, 1.836-fold, and 3.279-fold higher risk than those without depression respectively, P<0.001. The risk of stroke in subjects with sleep disorders was 1.622 times higher than those without sleep disorders, P<0.001.

Discussion

In this study, we investigated the risk factors for stroke. A total of 19384 subjects were retrieved from the NHANES database, of which 661 cases had stroke. Our outcome demonstrated that the patients with diabetes, lower household income (<$2000), PHQ-9 depression score between 5–27, trouble sleep, older age and higher concentration of blood lead had a higher risk of stroke.

Age is one of the risk factors for stroke, our study confirms the higher risk of stroke in older patients. The incidence of cerebrovascular disease increases with age. Bo-Ram Kim et al. also reported the relationship between age and stroke, study showed more severe stroke and complications in elderly (5). Multivariate analysis demonstrated that the risk of stroke was 1.839 times higher in subjects with diabetes than in those without diabetes. The study conducted by Barrett Connor et al. explored the relationship between diabetes and stroke that the mortality and morbidity of stroke in patients with diabetes were higher than those without diabetes (13). The reason for the higher risk of stroke may due to cerebrovascular atherosclerosis, cardiac embolism, or rheologic abnormalities in patients with diabetes (14).

Current study exhibited the increased risk of stroke in subjects with mild-, moderate-, and moderately severe or severe depression, and the conclusion of previous study that the depression is prospectively associated with a significantly increased risk of developing stroke supported our results (15). Trace elements play an important role in biological structures. However, those elements could be toxic when the concentrations are not in the desired range. Changes in the concentration of essential trace elements and heavy metals in human body may lead to acute hemorrhagic stroke (16). Karadas et al. found the higher serum lead level in patients with acute hemorrhagic stroke (17). In this study, we found the risk of stroke increases with increased blood lead level.

Sleep plays an important role in human life. Studies confirmed that both short and long sleep duration is associated with the risk of stroke (18, 19). A dose response meta-analysis exhibited the increased incidence of stroke by the long sleep duration (20). Our study also indicated the appropriate sleep duration (6–8h) as a proactive factor for stroke. As regard to the sleep disorders, current study showed the increased risk of stroke in patients with disorders. Sleep disorders are generally known as modifiable risk factors for stroke (21). The vascular pathology of stroke attributed to sleep disorder is influenced by various direct or indirect mechanisms (22).

Household income is not often discussed as risk factors for stroke. Subjects with the household income ≥$20000 had a lower risk than those with the income <$20000. Bird et al. pointed out that the household income was strongly related to the main disease intermediary for stroke - high blood pressure and the stroke risk factor - smoking (23). Therefore, the household income is a kind of modifiable risk factor for stroke.

Although the samples retrieved from NHANES database in this study is large enough, the number of stroke patients is still small which may influence the power of this analysis. The results of this study may also be influenced by the survival bias that wherein the subjects with interest or severe risk factors may not able to participate in the study.

Conclusion

Our study explores the risk factors for stroke based on the NHANES database. Results indicated that patients with diabetes, lower household income, PHQ-9 depression score, trouble sleep, older age and higher concentration of blood lead had a higher risk of stroke.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethics approval and consent to participate

The current study was based on the National Health and Nutrition Examination Survey database to explore the risk factors for stroke. The NHANES is a program of studies designed to assess the health and nutritional status of adults and children in the United States. It was the previous data provided by the NHANES and free of approval by the ethics committee.

Conflict of interest

The authors declare that they have no conflict of interest.