A Systematic Review and Meta-Analysis of the Association between Helicobacter pylori Infection and Dementia

INTRODUCTION

Dementia is a major public health problem worldwide. Alzheimer’s disease (AD) and vascular dementia account for 60–70% and 30% of cases, respectively [1]. Globally, about 35.6 million people were estimated to have dementia in 2010 and this number is expected to double every 20 years [2, 3]. The established risk factors of dementia include sociodemographics (age, female sex, low education) and family history [1, 4]. Diabetes mellitus, hypertension, stroke, and ischemic heart disease, significantly increase the risk dementia [2]. Current evidence also suggests that inflammation is involved in the pathogenesis of dementia [5].

An emergent body of knowledge also suggests the involvements of various infections with AD such as herpes simplex virus type 1, picornavirus, Borna disease virus, Chlamydia pneumoniae, Helicobacter pylori, and spirochete [6, 7]. Among those pathogens, the association between H. pylori infection and dementia [8 –11] is especially important given its chronic course and availability of treatment. H. pylori colonizes the stomach and causes chronic gastritis in all infected subjects. The infection is typically acquired in early life, however most people remain asymptomatic. In about 10–20% of those infected, peptic ulcer disease develops, and in rare occasions gastric cancer [12]. Chronic H. pylori infection induces humoral and cell-mediated immune responses, with predominant T-helper 1 mediated response and T-regulatory responses [13]. H. pylori-associated persistent inflammation does not eliminate the infection; rather it likely contributes the development gastric mucosa damage, and T-regulatory response promotes persistence of the infection [13].

Regarding the association between H. pylori infection, and dementia, the evidence is still limited with conflicting results [8–11 , 15]. Herein, we carried out a systematic review and meta-analysis to examine the association between H. pylori infection and cognitive function in adults.

MATERIALS AND METHODS

The systematic review and meta-analysis were conducted according to the PRISMA guidelines [16]. The literature search was performed using the databases OVID-Medline, PubMed, The Institute of Scientific Information Web of Science, and EMBASE. The terms that were used in the search were: “Helicobacter pylori/ H. pylori or Campylobacter pylori/ C. pylori; and at least one of the following: dementia, cognition disorder, delirium, cognitive function, cognitive decline, mild cognitive impairment, or Alzheimer’s disease”. A sample of detailed search strategy is provided in Supplementary Table 1. Reference lists of original and review articles were also screened. The search was limited to human studies and manuscripts published in English through January 2015.

Eligible studies were observational studies that addressed the association between H. pylori infection and cognitive function in adults. Intervention studies that addressed the role of H. pylori eradication on cognition were also included. Case reports, case series, duplicate reports, letters to the editor, commentaries, and authors’ replies were excluded.

RESULTS

The literature search yielded 420 records, of which 160 were duplicates and were thus excluded. The remaining 260 records were screened based on their titles and abstracts. Of them, 228 were excluded since they addressed different topics. In addition, four were animal studies and one a non-English publication. We reviewed the full text of the remaining 27 studies, and identified 13 that met inclusion criteria for the systematic review and seven for the meta-analysis [8–11 , 19–24] (Fig. 1).

The studies differed in their design, study population, sample size, outcome variables, and methods for detecting H. pylori infection (Table 1). AD as an outcome was evaluated in four studies [9 , 15]. Four studies addressed both AD and non-AD dementia [10 , 25], among which only three presented separate findings for AD and non-AD dementia [22 , 25]. One study evaluated MCI [8] and two evaluated differences between H. pylori positive and negative subjects from the general population, regarding cognitive function as measured by a battery of psychological tests [19, 23]. In two studies, the exposure variable was H. pylori therapy [20, 21]. Two, six and four studies employed cross-sectional [19, 23], case-control [8 , 24], and four were prospective studies [10 , 25].

The studies were conducted in the United States [19, 23], Europe (France [10], Greece [8 , 21], Italy [22], Sweden [24]), Japan [14], and Taiwan [20, 25]. Both males and females were enrolled in all studies. Sample sizes were small in the case-control investigations [9, 22], but some investigations involved large samples [14 , 25]. The sampling frame consisted of healthy adults (>20 years) from the general population who participated in health surveys [19, 23]. In two studies from Taiwan, the sampling frame consisted of nationwide health insurance population-based datasets (AD patients with mean age 76–79 years [20, 25], and cohorts aged ≥40 years with and without H. pylori infection) [25]. The study conducted in France enrolled 603 non-institutionalized individuals aged ≥65 years who participated in the Personas Ages QUID study [10]. The sampling frame for cases in the case-control studies were mental health centers [11], memory clinics [8 , 21], a neurology department [22], and a clinic of geriatric medicine [24]. The sampling frame for controls was health examination centers [11], patients referred to memory clinics who were not diagnosed with dementia [14], healthy volunteers without dementia from a neurological department [22], patients from a gastroenterology department who underwent gastroscopy to investigate mild anemia [8 , 21], patients with prostate hyperplasia or long-term fractures necessitating surgery after epidural anesthesia [15], and individuals screened for gastritis from the general population [24].

The follow-up period in the prospective studies ranged from two [21] to twenty [10] years. The exposure variable in most of the studies was H. pylori infection [8–11 , 22–25]. In one study the exposure was eradication of H. pylori, and in another H. pylori treatment was considered the exposure variable [20]. The methods for detecting H. pylori included measuring IgG antibodies in the serum by Enzyme Linked Immunosorbent Assays (ELISA) [10 , 23] or in urine [14]. One study measured IgG antibodies in the serum and in cerebrospinal fluid (CSF) [15]. Three studies used gastric biopsy employing urease test and histology [8 , 21], together with measuring serum IgG antibodies [8 , 21]. In two studies, H. pylori infection or therapy was determined based on codes in the international classification of diseases 9th edition (ICD-9) in national registries [20, 25]. Methods to account for confounders were conducted in most studies, by matching [9 , 22], multivariable analyses [10 , 25], or both [11].

DISCUSSION

We conducted a systematic review and meta-analysis of the association between H. pylori infection and dementia. To the best of our knowledge, our study is the first meta-analysis of the association between H. pylori infection and dementia. Our study adds original knowledge over a previous systematic review that addressed the association between infections, including H. pylori, and AD, which lacked meta-analysis [7].

We identified 13 observational studies that assessed the relationship between H. pylori infection and dementia and/or cognitive function in adults. The seven studies that were included in the meta-analysis were those in which the exposure variable was H. pylori infection (yes versus no) and the outcome variable was any dementia [8–11 , 25]. Our findings indicate that H. pylori infection was significantly associated with a 1.71-fold (95% CI 1.17–2.49) increased likelihood of dementia. A sub-group analysis according to the type of dementia demonstrated a positive association with both AD and non-AD, with a stronger association found with the latter; pooled OR 1.39 (95% CI 0.76–2.52) for AD and pooled OR 2.04 (95% CI 0.90–4.62) for non-AD dementia, these results were not statistically significant likely due to lack of statistical power. The association between H. pylori infection and non-AD dementia might be of greater magnitude than the association between the infection and AD, possibly given the involvement of H. pylori infection in metabolic risk factors such as diabetes mellitus [26], which is known to affect the risk of vascular dementia. Overall, well-designed large-scale studies showed a positive association between the infection and dementia, with OR ranging between 1.39 and 1.51 [11, 25]; and two small-scale studies showed a stronger association, with OR > 8 [8, 9]. The latter studies [8, 9] determined H. pylori infection status using histological examination of gastric biopsy, contrasting with studies that measured H. pylori specific antibodies and this might explain the variation in the magnitude of the association measures among these studies. Interestingly, the exclusion of these smaller studies [8, 9] slightly attenuated the pooled OR to 1.33 and reduced the heterogeneity among the studies, implying that they might be the source of heterogeneity.

It is not clear yet if association between H. pylori infection and dementia is causal, however if future studies will support causation, then the availability of an effective therapy against H. pylori using a 10–14-day regimen [27] might open new horizons toward the prevention and control of dementia. While the observed association may appear weak, given the high prevalence of H. pylori infection in many populations, a causal association will result in a high population attributable fraction. Noteworthy, only two studies assessed the role of H. pylori therapy on cognitive function and progression of AD, both were observational [20, 21]. A large study (N = 1,538) from Taiwan showed that patients with AD and peptic ulcer disease who received anti-H. pylori therapy had a significantly lower odds to change medications of AD compared to those who did not receive anti-H. pylori therapy [20]. That study employed changing medications of AD as an indication of the progression of the disease [20], however in clinical practice changing medications may not be directly translated to progression of the disease. Kountouras et al. [21] showed, in a small study, less deterioration of cognitive and functional abilities after two years of H. pylori eradication, as measured by the Mini-Mental State Examination, the Cambridge Cognitive Examination for the Elderly, and the Functional Rating Scale for Symptoms of Dementia, among AD patients (n = 61) who had their H. pylori infection eradicated compared to patients with eradication failure and to those who had no H. pylori therapy [21]. These findings support the notion of possible involvement of H. pylori infection, at least in some individuals, in dementia. Moreover, eradication of the infection might be beneficial for some patients with dementia, in terms of delaying the deterioration of their functional and cognitive abilities [21]. To date, randomized controlled trials assessing the role of H. pylori eradication therapy on the risk and progression of dementia are lacking and they are essential to answer this important question.

The association of H. pylori infection with cognition in adults from the general population was assessed in two studies conducted in large samples in the United States [19, 23]. These studies evaluated differences between H. pylori positive and negative subjects in various cognitive function tests. One study [23] based on data of 1,785 participants aged 20–59 years from the National Health and Nutrition Examination Survey III (NHANES III), showed no main effect for H. pylori on the performance of cognitive tests. However, subjects with evidence of H. pylori and latent toxoplasmosis had lower cognitive scores on the memory test (serial-digit learning test) [23]. This study found that higher education attenuated the negative effects of H. pylori infection on cognition [23]. Individuals with lower education who were infected with H. pylori had lower cognitive scores than uninfected ones [23].

Moreover, Beydoun et al. [19] also utilized the NHANES III dataset, but included 5,927 participants aged 20–59 years and 3,468 participants aged 60–90. That study showed poorer performance in the serial digit-learning test among H. pylori sero-positive subjects compared to sero-negatives in the age group 20-59 years. In the age group 60–90 years H. pylori sero-positive individuals showed poorer achievement on the memory test (story recall test) compared to sero-negative ones [19]. It is possible that H. pylori-induced damage becomes more evident with age, or/and that older individuals are more susceptible to any additional stressor that may affect their cognition.

The association between H. pylori infection and dementia is biologically plausible. H. pylori gastritis changes the gastric physiology, causes damage to the gastric mucosa and with time and aging, atrophy of the stomach develops. These changes affect the absorption of essential micronutrients e.g., vitamin B12 and iron, which were shown to be related to both H. pylori infection [28, 29] and to dementia [22 , 31]. Vitamin B12 deficiency leads to increased homocysteine concentration, thus causing vascular and endothelial damage and dementia. Moreover, a positive association was reported between H. pylori infection and diabetes mellitus [32], a major risk factor for cognitive impairment [33]. H. pylori induced persistent inflammatory response might also explain the biological pathways of the association between H. pylori infection and dementia. Experimental studies [34, 35] showed that H. pylori might be involved in the pathophysiological process of AD such as tau hyperphosphorylation [35] and production of amyloid-beta [34]. A study of 53 patients with AD has demonstrated that H. pylori infected patients had significantly l lower serum pepsinogen (PG) I to PGII ratio (as a marker of gastric atrophy), higher levels of homocysteine and inflammatory biomarkers (IL-8 and TNF-α) in the CSF, and lower level of plasma IL-1β [36], in comparison to uninfected AD patients.

In a recent study, we showed that H. pylori infection was significantly associated with 6-point lower mean intelligence quotient (IQ) scores among healthy school-age children [37]. This association was not explained by nutritional status and socio-demographic variables [37]. Thus, it is possible that H. pylori infection- associated cognitive deficits in childhood will extend into adulthood.

Potential limitations of the current study should be acknowledged. As a meta-analysis of observational studies published in English, a publication bias may be possible. However, we carried out a comprehensive search of the literature, and did not find significant evidence of publication bias either by funnel plot or Eager intercept statistic. Additionally, the individual observational studies have their own weaknesses, and, in general, observational studies are prone to confounding factors. We used in the meta-analysis adjusted measures of association that were reported in the original studies; however some studies reported only crude findings. These limitations are carried forward into the meta-analysis, and might threaten the internal validity of the findings. The status of H. pylori infection was determined based tests based on gastric biopsy in some studies, but mostly based on the presence of serum H. pylori IgG antibodies. Worth mentioning that H. pylori IgG antibodies persist in the serum up to one year following H. pylori eradication, thus non-differential misclassification of the exposure variable might be present. Our meta-regression analysis suggested that H. pylori detection method and adjustment for confounders might affect the magnitude of the association between the infection and dementia.

The sample size included in this meta-analysis is larger than that of any individual study, thus enabling more precise estimation of the association measure. We identified important gaps in knowledge, specifically the lack of randomized controlled trials assessing the role of H. pylori therapy on cognitive and functional abilities of adults, and the limited number of studies that addressed the role of H. pylori infection and dementia sub-type, or the role of the infection in cognition in the general population. Addressing these gaps will elucidate the nature of the association between the infection and cognition in adults.

In conclusion, H. pylori infection is associated with increased likelihood of dementia.

The mechanisms that may explain this relationship need to be determined, and the effect of H. pylori therapy on the progression of dementia needs to be evaluated.