Increased Risk of Dementia Among Patients With Pulmonary Tuberculosis

Abstract

Objective:

We investigated and compared the risk of dementia development in a cohort of patients with tuberculosis (TB).

Methods:

The study involved 6473 patient with newly diagnosed TB, and each patient was randomly frequency matched with 4 people without TB based on age, sex, and index year. The risk of dementia development was analyzed using Cox proportional hazards regression.

Results:

Among the patients with TB, the overall risk of developing dementia was 1.21-fold significantly higher than the non-TB cohort. In the stratified analysis of dementia risks, only the patients with TB who were male or 50 to 64 years of age exhibited a significantly higher risk of dementia development compared with those without TB. An analysis of the follow-up duration revealed that patients with TB had a 1.78-fold increased risk within 1 year of follow-up.

Conclusion:

Patients with TB have a significantly higher risk of developing dementia than that of the general population.

Introduction

Dementia is a prevalent neurodegenerative disorder affecting elderly people, characterized by considerable impairment of intellect, memory, and personality without impairment of consciousness beyond what might be expected from natural aging. Because of its major medical, social, and economic impacts, dementia is an increasing burden worldwide. ¹ Previous studies have associated several risk factors with dementia, including socioeconomic status, hypertension, hypercholesterolemia, diabetes, head injury, and depression. ^2,3

Pulmonary tuberculosis (TB) is a critical global public health concern. One third of the world’s population is believed to have latent TB infection, which carries a risk of subsequent progression to an active state at any time. Moreover, TB is second only to HIV/AIDS as the major cause of death from a single infectious agent, primarily caused by the species Mycobacterium tuberculosis. In 2012, the World Health Organization reported that 8.6 million people were affected by TB worldwide, and 1.3 million people died of TB. In Taiwan, the estimated incidence of TB in 2011 was approximately 54.5 per 100 000 person-years, with pulmonary TB accounting for approximately 94% of all cases with TB. ⁴

Although the etiology of dementia is heterogeneous, a growing number of studies have provided evidence indicating the critical role of inflammation in the pathogenesis of dementia. ^{5 –7} In addition, inflammation is crucial for pathogenesis of TB. ⁸ A previous epidemiologic study suggested that inflammation triggered by an infectious agent may play a role in the pathogenesis of dementia. ⁹ However, pulmonary TB is not generally considered a risk factor for development of dementia. Therefore, we conducted a longitudinal nationwide population-based cohort study to test the hypothesis of whether pulmonary TB increases the subsequent risk of dementia.

Methods

Data Source

This retrospective cohort study was conducted using data obtained from the Longitudinal Health Insurance Database (LHID), which is a subset of the National Health Insurance Research Databases (NHIRD) maintained by the Bureau of National Health Insurance, Taiwan. The National Health Insurance (NHI) program was established in March 1995, providing insurance coverage for more 99% the population in Taiwan (http://www.nhi.gov.tw/). Taiwan launched the NHI program in 1995, providing insurance coverage for more than 99% of the population in Taiwan (http://www.nhi.gov.tw/), operated by the single payer, the government. Medical reimbursement specialists and peer review should scrutinize all insurance claims. The identification of diseases in the NHIRD was based on the International Classification of Disease, 9th Revision, Clinical Modification (ICD-9-CM), which were judged and determined by related specialists and physicians according to the standard clinical criteria. The LHID contains the medical records (from 1996 to 2011) of 1 million insurant who were randomly selected from the original 2000 Beneficiary Registry. In accordance with the Personal Information Protection Act, the insurant data were electronically depersonalized before being released for research purposes.

Study Participants

From the LHID, we collected data on 6904 patients whose pulmonary TB (ICD-9-CM 011.XX) was newly diagnosed between 2000 and 2010. The index date was defined as the date of TB diagnosis. Patients younger than 20 years or with a history of dementia before the index date were excluded. Subsequently, the TB cohort in this study comprised 6473 patients with TB. For the comparison cohort, people without TB or dementia before the index date were selected from the LHID and frequency matched by age, sex, and index year as the TB cohort at a ratio of 4:1.

End Point and Comorbidity

All study participants were followed up from the index date until the onset of dementia (ICD-9-CM 290.XX, 294.1X and 331.0X), withdrawal from the NHI program, or the end of 2011. The diagnoses of dementia were based on the ICD-9-CM codes which were judged and determined by related specialists and physicians. The comorbidities (ICD-9-CM) considered in this study were atrial fibrillation (AF; 427.31), hypertension (401.XX-405.XX), hyperlipidemia (272.XX), diabetes (250.XX), heart failure (428.XX), head injury (850.XX-854.XX and 959.01), stroke (430.XX-438.XX), and depression (296.2X, 396.3X, 300.4X, and 311.XX). The comorbidities were considered only if they were diagnosed before the index date.

Statistical Analysis

A chi-square test was performed to determine the difference in demographic characteristics between the TB and the comparison cohorts. The incidence of dementia per 1000 person-years was calculated for the 2 cohorts. Cox proportional hazard regression was used to estimate the hazard ratio (HR) and 95% confidence interval (CI) of dementia and dementia-associated risk factors. A multivariable Cox proportional hazard model was adjusted to account for the effects of age, sex, and comorbidity, differing significantly from the unadjusted model. Age-, sex-, and comorbidity-specific risks were assessed in multivariable model. For the future analysis, we assessed the risk of dementia development stratified by study duration. Cumulative incidence curves for dementia in the 2 cohorts were plotted based on Kaplan-Meier analysis, and the difference between the 2 cohorts was tested by performing a log-rank test. SAS Version 9.3 for Windows (SAS Institute, Cary, North Carolina) was employed for all statistical analyses, and the level of significance was set at .05 in a 2-tailed test.

Data Availability Statement

All data and related metadata were deposited in an appropriate public repository. The data on the study population that were obtained from the NHIRD (http://w3.nhri.org.tw/nhird//date_01.html) are maintained in the NHIRD (http://nhird.nhri.org.tw/). The NHRI is a nonprofit foundation established by the government.

Ethics Statement

The NHIRD encrypts patient personal information to protect privacy and provides researchers with anonymous identification numbers associated with relevant claims information, including sex, date of birth, medical services received, and prescriptions. Patient consent is not required to access the NHIRD. This study was approved by the Institutional Review Board (IRB) of China Medical University (CMU-REC-101-012). The IRB specifically waived the consent requirement.

Results

Among 32 363 study participants selected for this retrospective cohort study, 6473 were assigned the TB cohort, and the remaining 25 890 were assigned to the comparison cohort. The mean age of the TB cohort was 59.7 (standard deviation = 18.2), and most of them were men (69.0%). Compared with the comparison cohorts, the patients with TB were more likely to have comorbidities, except for hyperlipidemia (Table 1). During a mean 5.9-year follow-up period, 288 and 1128 participants had developed dementia among the TB and comparison cohorts, respectively, with a corresponding incidence of 8.36 and 7.21 per 1000 person-years (Table 2). The cumulative incidence of dementia development in the TB cohort was slightly higher than that in the comparisons cohort after a 12-year follow-up period (log-rank test P = .03, Figure 1).

Table 1.

Demographic Characteristics Between the TB and the Comparison Cohorts.

Variable	TB (N = 6473)		Comparison (N = 25 890)		P Value
	n	%	n	%
Age, year					.99
20-34	810	12.5	3240	12.5
34-49	1169	18.1	4676	18.1
50-64	1529	23.6	6116	23.6
65-79	2159	33.3	8624	33.3
80+	809	12.5	3234	12.5
Mean (SD)	59.7 (18.2)		59.0 (18.2)
Gender					.99
Women	2004	31.0	8016	1.0
Men	4469	69.0	17 874	69.0
Comorbidity
AF	121	1.87	367	1.42	.008
Hypertension	2726	42.1	10457	40.3	.01
Hyperlipidemia	1274	19.7	5405	20.9	.03
Diabetes	1366	21.1	3599	13.9	<.0001
Heart failure	454	7.01	1181	4.56	<.0001
Stroke	569	8.79	1495	5.77	<.0001
Depression	294	4.54	978	3.78	.005
Head injury	302	4.67	675	2.61	<.0001

Abbreviations: TB, pulmonary tuberculosis; SD, standard deviation; AF, atrial fibrillation.

Table 2.

Incidence and Hazard Risk of Dementia and Dementia-Associated Risk Factors.

Variable	Event no.	Person-Years	Incidencea	HR (95% CI)
				Crude	Adjusted
TB
No	1128	156 404	7.21	1.00	1.00
Yes	288	34 434	8.36	1.16 (1.02-1.32)b	1.21 (1.06-1.38)c
Age, years
20-49	19	67 768	0.28	1.00	1.00
50-64	109	48 530	2.25	8.11 (4.98-13.2)d	6.76 (4.14-11.0)d
65-79	808	59 704	13.53	49.6 (21.5-78.3)d	36.2 (22.8-57.5)d
80+	480	14 836	32.35	125 (78.7-197)d	87.0 (54.5-139)d
Gender
Women	389	61 705	6.30	1.00	1.00
Men	1027	129 134	7.95	1.26 (1.12-1.42)d	1.00 (0.89-1.13)
Comorbidity
AF
No	1380	188 848	7.31	1.00	1.00
Yes	36	1991	18.08	2.46 (1.77-3.42)d	0.78 (0.56-1.10)
Hypertension
No	398	121 971	3.26	1.00	1.00
Yes	1018	68 868	14.78	4.55 (4.05-5.11)d	1.38 (1.21-1.56)d
Hyperlipidemia
No	998	154 779	6.45	1.00	1.00
Yes	418	36 059	11.59	1.80 (1.60-2.01)d	1.06 (0.94-1.20)
Diabetes
No	1048	166 154	6.31	1.00	1.00
Yes	368	24 685	14.91	2.36 (2.10-2.66)d	1.39 (1.23-1.56)d
Heart failure
No	1274	184 329	6.91	1.00	1.00
Yes	142	6510	21.81	3.15 (2.65-3.75)d	1.09 (0.91-1.30)
Stroke
No	1216	182 728	6.65	1.00	1.00
Yes	200	8110	24.66	3.71 (3.19-4.31)d	1.46 (1.25-1.70)d
Depression
No	1314	184 571	7.12	1.00	1.00
Yes	102	6268	16.27	2.28 (1.86-2.79)d	1.63 (1.32-1.99)d
Head injury
No	1330	186 344	7.14	1.00	1.00
Yes	86	4495	19.13	2.67 (2.15-3.32)d	1.79 (1.43-2.23)d

Abbreviations: TB, pulmonary tuberculosis; HR, hazard ratio; CI, confidence interval; AF, atrial fibrillation.

^aPer 1000 person-years.

^b P < .05.

^c P < .01.

^d P < .001.

Figure 1.

Kaplan-Meier analysis for cumulative incidence of dementia between the tuberculosis (TB) and the non-TB cohort.

Compared with the comparison cohort, the risk of developing dementia among the patients with TB was 1.21-fold higher after we controlled for age, sex, AF, hypertension, hyperlipidemia, diabetes, heart failure, stroke, depression, and head injury (95% CI = 1.06-1.38; Table 2). The risk of dementia development increased with age from 6.76 (50-64 years old) to 87.0 (80+ years old) compared with participants at 20 to 49 years old (95% CI = 4.14-11.0 and 54.5-139, respectively). Except for the patients with TB, those with a head injury exhibited the highest risk of developing dementia (HR = 1.70, 95% CI = 1.43-2.23), followed by depression (HR = 1.63, 95% CI = 1.32-1.99), stroke (HR = 1.46, 95% CI = 1.25-1.70), diabetes (HR = 1.39, 95% CI = 1.23-1.56), and hypertension (HR = 1.38, 95% CI = 1.21-1.56).

Patients with TB had a higher incidence of dementia than that of comparison cohort based on sex, 4 age-groups, and the presence of comorbidities. However, only men with TB (HR = 1.22, 95% CI = 1.05-1.42), aged 50 to 64 years (HR = 1.65, 95% CI = 1.09-2.51), or with comorbidities (HR = 1.21, 95% CI = 1.05-1.40) had a significantly higher risk compared to the comparison cohort with matching criteria (Table 3). In the follow-up stratified analysis, the patients with TB exhibited a 1.78-fold risk compared with the comparison cohort within the first year of follow-up (95% CI = 1.36-2.32; Table 4). Beyond 1 year of follow-up, no significant difference was observed between the 2 cohorts.

Table 3.

Incidence and Hazard Risk of Dementia Stratified by Age, Gender, or Comorbidity.

Variable	TB		Comparison		HR (95% CI)
	Event no.	Incidencef	Event no.	Incidencef
Age, yeara
20-49	6	0.46	13	0.24	1.43 (0.51-4.02)
50-64	33	3.72	76	1.92	1.65 (1.09-2.51)b
65-79	157	15.37	651	13.15	1.14 (0.95-1.35)
80+	92	39.28	388	31.06	1.24 (0.98-1.55)
Genderc
Women	75	6.47	314	6.27	1.17 (0.91-1.51)
Men	213	9.33	814	7.66	1.22 (1.05-1.42)d
Comorbiditye
No	47	2.81	207	2.44	1.27 (0.92-1.74)
Yes	241	13.62	921	12.90	1.21 (1.05-1.40)d

Abbreviations: TB, pulmonary tuberculosis; HR, hazard ratio; CI, confidence interval; AF, atrial fibrillation.

^aAdjusted for gender, hypertension, hyperlipidemia, heart failure, depression, stroke, head injury, diabetes, and AF.

^b P < .05.

^cAdjusted for age, hypertension, hyperlipidemia, heart failure, depression, stroke, head injury, diabetes, and AF.

^d P < .01.

^eAdjusted for age and gender.

^fPer 1000 person-years.

Table 4.

Incidence and Hazard Risk of Dementia Stratified by Follow-Up Duration.^a,b

Follow-Up Duration, year	TB		Comparison		HR (95% CI)
	Event no.	Incidenced	Event no.	Incidenced
1	79	13.04	180	7.08	1.78 (1.36-2.32)c
2-3	69	6.86	322	7.17	0.99 (0.76-1.29)
4-5	60	7.84	256	7.19	1.18 (0.89-1.57)
>5	80	7.50	370	7.33	1.10 (0.87-1.41)

Abbreviations: TB, pulmonary tuberculosis; HR, hazard ratio; CI, confidence interval; AF, atrial fibrillation.

^aManually adjusted for age, gender, hypertension, hyperlipidemia, heart failure, depression, stroke, head injury, diabetes, and AF.

^bCox proportional hazard regression assumption P = .03.

^c P < .0001.

^dPer 1000 person-years.

Discussion

In this large cohort study, we examined the longitudinal association between TB and dementia risk using data retrieved from the NHIRD. We observed that patients with pulmonary TB were 1.21-fold more likely to develop dementia, particularly men aged 50 to 64 years of age, compared to those without TB. Based on our research, this is the first study indicating that TB increases the subsequent risk of dementia.

Tuberculosis is a continuum consisting of a spectrum of diseases resulting from complex inflammatory responses. The infection process of M tuberculosis and the associated immunological events can be divided into the following 3 distinct yet interrelated stages: first, primary or secondary infection is characterized by an innate immune response; second, latency following an immunological equilibrium state; and, third, reactivation caused by a decline in the host’s immunity or a failure to develop immune signals. ¹⁰ Proinflammatory cytokines, including interferon, tumor necrosis factor α (TNF-α), and interleukin 1β (IL-1β), are crucial for regulating the complex interplay between the host immune system and M tuberculosis. Because of the systemic immune response, TB not only presents as pulmonary disease but also with accompanies manifestations such as wasting and anorexia. ¹¹

Dementia is a clinical syndrome characterized by memory loss and impaired cognitive ability with adequate severity to affect daily functioning. Alzheimer’s disease (AD) is the leading form of dementia, accounting for approximately 75% of cases. ¹² The specific mechanism responsible for development of AD remains unclear. The hypothesized mechanisms include amyloid metabolic cascade, apolipoprotein E genotyping, the hyperphosphorylation of tau protein, oxidative stress, abnormal cell cycle reentry, inflammation, and amyloid β metabolism. ¹³ Among these hypothesized mechanisms, inflammation associated with infection or injury could be a practical explanation because it may provide a potential target for intervention and prevention.

Although the mechanisms underlying the relationship between TB and dementia are unclear, evidence from previous research might explain this association. Most people with TB are asymptomatic because M tuberculosis colonizes in the lungs in a latent state, and fewer than 10% of infected people become TB active. ¹⁴ Both TNF-α and IL-1β are critical proinflammatory cytokines involved in the pathogenesis of TB and have been shown to be associated with subsequent risk of AD in older adults. ¹⁵ Increasing evidence has also shown that the inflammatory response signals generated outside the central nervous system, as is typical in cases of pulmonary TB infection, can cross the blood–brain barrier, communicate with the brain, and become involved in multiple crucial pathogenic events of neurodegeneration through several mechanisms. ^7,16 Conversely, the use of nonsteroidal antiinflammatory drugs may reduce the risk of AD development. ¹⁷ Moreover, in a cohort study involving 300 community-dwelling patients with mild to severe AD, both acute and chronic systemic inflammation—which are associated with increases in serum TNF-α—were associated with an increase in cognitive decline over a 6-month period. ¹⁸ A case–control study by Dunn et al showed that elderly patients with dementia have a higher ratio of infection episodes in the 4 years preceding the diagnosis of dementia. ¹⁹

In our study, among the patients with pulmonary TB, most (69%) were men, which is consistent with previous reports in both Asian and Western countries. ^20,21 We observed a comparative lower incidence of dementia in both pulmonary TB and comparison groups in people younger than 49 years, as well as a positive association with dementia development (Table 2), which is consistent with the findings of a previous study. ²² Among the patients with TB, men and people between 50 and 64 years exhibited a significantly higher risk of dementia development compared with the comparison cohort. During the 12-year follow-up, the risk of dementia among patients with pulmonary TB was significantly higher than in the comparison cohort during the first year only. Because no study has investigated the longitudinal relationship between pulmonary TB and dementia, future research is required to corroborate or refute this finding.

In this epidemiologic study using a nationwide insurance database, the accuracy of the clinical coding system may be questioned. However, in Taiwan, TB is under intense scrutiny, and all suspected cases with TB are confirmed by sputum smears and radiologic examination; moreover, reporting of patients with TB is mandatory and enforced by the Centers for Disease Control (Ministry of Health and Welfare, Taiwan). In addition, all insurance claims must be reviewed and scrutinized by a medical reimbursement specialist. Therefore, the validity regarding diagnoses of TB in this study is reliable.

Several limitations were encountered while conducting this study. First, despite our adequate model adjustment to account for potential confounding factors, bias might remain in this retrospective study. The study data were retrieved from the NHIRD, in which all data are deidentified for research purposes; consequently, detailed information on the demographic characteristics and daily lives of the participants are unavailable, such as their education level, occupation, exercise, diet, and cigarette smoking behaviors, all of which are potentially confounding factors. Second, dementia is an underdiagnosed health problem. ²³ Therefore, the incidence of dementia might have been underestimated in both the groups. Third, the different types of dementia, such as AD and vascular dementia, could not be exactly and accurately differentiated in the NHIRD because the information regarding the diagnostic codes of ICD-9-CM is unavailable. Therefore, they were included as all types of dementia in this study. Forth, the effect of anti-TB treatment on dementia development was not evaluated in this study. Regardless, the finding regarding the relationship between pulmonary TB and dementia remains highly reliable and can be generalized because of the nationwide population-based study design, validity of the database, large sample, and long follow-up period.

In conclusion, this study suggests that patients with TB have a higher risk of developing dementia, particularly in men and within 1 year after TB diagnosis. Further studies are necessary to corroborate this finding and to explore the mechanisms underlying this association.