Fig. 1.1
Flow of potential reviews through the screening process
5.2.2 Quality of Included Reviews
The overall quality of the five reviews was moderate (Table 1.1). All five reviews specified a priori research questions and inclusion criteria; however, only three [13, 24, 25] reported utilizing at least two independent persons to screen citations or extract data from included citations. Three reviews [23–25] specified comprehensive literature searches, which AMSTAR [17] defines as including at least two electronic database searches plus a supplemental search of at least one other source (e.g., existing reviews, experts, reference lists of retrieved articles), as well as specification of search terms (i.e., keywords, MeSH terms). One review [13] included all of these components except the supplemental search. The other review [22] did not report a supplemental search or any search terms. None of the reviews searched all types of publications (i.e., peer-reviewed journal articles, gray literature) for relevant citations; authors restricted their searches to journal articles. None of the reviews referenced the citations excluded at any level of screening although the authors referenced all included citations. All five reviews listed the characteristics of the included studies. Only two reviews [23, 24] assessed the quality of the included citations, and one of these [23] factored the quality assessment into the conclusions. Both reviews [23, 24] listed cutoff scores on quality assessment instruments as a criterion to exclude poor quality studies from review. Four [13, 23–25] of the five reviews included meta-analyses and the authors of these four reviews employed appropriate methods to combine the data. Three [24, 26, 27] reviews ascertained publication bias. The authors of all five reviews mentioned their funding sources or conflicts of interest.
Table 1.1
Quality of included reviews
Review | A priori design provided | Duplicate study selection/data extraction | Comprehensive literature search | Status of publica-tion as inclusion criterion | List of included/excluded studies provided | Characteristics of included studies provided | Study quality assessed | Study quality used to formulate conclusions | Methods of combining studies appropriate | Publication bias assessed | Conflict of interest acknowledgeda |
|---|---|---|---|---|---|---|---|---|---|---|---|
Outcome: cognitive impairment | |||||||||||
Gifford et al. [13] | Yes | Yes | No | No | No | Yes | No | NA | Yes | No | Yes |
Outcome: Alzheimer’s disease | |||||||||||
Feldstein [22] | Yes | No | No | No | No | Yes | No | NA | NA | No | Yes |
Guan et al. [23] | Yes | No | Yes | No | No | Yes | Yes | Yes | Yes | No | Yes |
Power et al. [24] | Yes | Yes | Yes | No | No | Yes | Yes | No | Yes | Yes | Yes |
Outcome: vascular dementia | |||||||||||
Sharp et al. [25] | Yes | Yes | Yes | No | No | Yes | No | NA | Yes | No | Yes |
None of the reviews graded the strength of evidence, which is recommended to provide readers with a sense of whether additional primary data might change the conclusions of the existing body of evidence [28].
5.2.3 Findings of the Reviews
Cognitive Impairment
One review [13] (Table 1.2—Appendix) of cross-sectional and longitudinal studies examined the relation between blood pressure and cognitive impairment. This review included 12 publications based on 10 unique studies with a total of 3601 participants. The age range was 43–90 years. Although the 12 included publications all contained persons with hypertension (defined as systolic blood pressure >140 mmHg and diastolic blood pressure >90 mmHg), the meta-analyzed outcome was “blood pressure.” The included publications used neuropsychological tests to measure any of seven different domains of cognition: global cognition, episodic memory, language, attention, executive function, information processing speed, and visuoperceptual skills. Measures of effect in the included publications were transformed into Pearson’s correlation coefficients (r) and meta-analyzed without differentiation by study design. The authors provided unadjusted summary r values and also calculated adjusted summary r values from a subset of eight included publications that adjusted results for age, education, diabetes, cholesterol, and cardiac disease. The meta-analysis found weak inverse correlations (adjusted/unadjusted) between hypertension and global cognition (r = −0.07/r = −0.11 [p < 0.001]) and episodic memory (r = −0.18/r = −0.20 [p < 0.001]), and a weak positive correlation (adjusted only) with attention (r = 0.14 [p = 0.002]). The correlation coefficients were not statistically significant for the other cognitive domains. Results of the meta-analyses were not broken down by neuropsychological test.
Table 1.2
Data extraction
Review | Objective | Search strategy | Number of included studies | Total number of participants | Type of cognitive impairment | Outcomes |
|---|---|---|---|---|---|---|
Outcome: cognitive impairment | ||||||
Gifford et al. [13] | To investigate the association between hypertension and cognitive impairment in persons without a history of clinical stroke or dementia | Pubmed (1966–2013), PsycINFO (1887–2013) | 12 publications from 10 unique studies (each publication includes between 1 and 6 cognitive domains) | 3601 (does not include double counting of participants in multiple publications from the same study) Age range: 43.0–91.0 years | Cognitive domains (measured using various neuropsychological tests): | Unadjusted/adjusted (only studies controlling for age, education, diabetes, cholesterol, cardiac disease) |
Search terms provided (cohort or cross-sectional studies only) | Global cognition (5 studies) | Global cognition: r: −0.07 (p < 0.001)/r: −0.11 (p < 0.001) | ||||
English-language only | Episodic memory (10 studies) | Episodic memory: r: −0.18 (p < 0.001)/r: −0.20 (p < 0.001) | ||||
Language (3 studies) | Language: NS/NS | |||||
Attention (5 studies) | Attention: NS/r: 0.14 (p = 0.002) | |||||
Executive function (9 studies) | Executive function: NS/NS | |||||
Information processing speed (5 studies) | Information processing speed: NS/NS | |||||
Visuoperceptual skills (4 studies) | Visuoperceptual abilities: NS/NS | |||||
Outcome: Alzheimer’s disease | ||||||
Feldstein [22] | To investigate the relationship between hypertension and the pathogenesis, development, and progression of AD | PubMed, Medline, Cochrane Reviews of Clinical Trials, AMED Virtual Library, Cumulative Index to Nursing, Allied Health Literature (all databases 1993–2011) | Mid-life hypertension and AD: 5 publications from 4 unique studies | Mid-life hypertension and AD: 22,174, Age range: 40.0–64.0 years | AD diagnosed according to any standardized criteria, supplemented by computed tomography or magnetic resonance imaging | No meta-analysis |
To examine the effect of antihypertensive treatment on the incidence and progression of AD | Search terms not provided (included cohort studies only) | Old age hypertension and AD: 5 | Old age hypertension and AD: 6657, Age range: ≥65.0 years | Mid-life hypertension and AD: 10/12 reported OR/HR > 1.00 (5/10 statistically significant); 2/12 reported OR/HR ≤ 1.00 (both NS) | ||
No mention of language restriction | Old age hypertension and AD: 5/13 reported OR/HR > 1.00 (all NS); 8/13 OR/HR ≤ 1.00 (1/8 statistically significant) | |||||
Guan et al. [23] | To investigate the association between hypertension and the risk of AD | Medline, Embase (both databases 1966–2011) | 9 | 19,178, Age range: 40.0 to ≥75.0 years | AD diagnosed according to any standardized criteria | No association between hypertension and AD: RRs: 1.01 (95 % CI: 0.87–1.18) |
To examine the effect of antihypertensive treatment on the incidence and progression of AD | Search terms provided (cohort studies [n = 10] and RCTs [n = 2]) | |||||
English- or Chinese-language only | ||||||
Power et al. [24] | To investigate the association between hypertension and the risk of AD | Medline, Embase, CINAHL, Web of Science (all databases inception-2010) | 19 (18 publications; 1 publication reported on 2 studies) | 42,968, Age range: 50.0–100.0 years | AD diagnosed according to any standardized criteria | History of hypertension and hypertension at baseline (12 studies): RRs: 0.97 (95 % CI: 0.80–1.16) |
Search terms provided (cohort or nested case–control studies only) | Mid-life hypertension (4 studies, age < 65 years): RRs: blood pressure classification schemes different → no summary estimate and results NS (3 studies); 1 study reported RR: 2.6 (95 % CI: 1.1–6.6) for systolic BP ≥160 mmHg vs <140 mmHg | |||||
No language restriction | Late-life hypertension (12 studies, age ≥ 65 y): RRs: 0.94 (95 % CI: 0.90–1.00) | |||||
Outcome: vascular dementia | ||||||
Sharp et al. [25] | To examine whether hypertension is risk factor for VaD | Medline (1950–2007), Embase (1980–2007), PsycINFO (1967–2007)
Related posts: Hypertension and Brain Inflammation: Role of RAS-Induced Glial Activation
 Arterial Stiffness and the Brain
 Endothelium, the Blood–Brain Barrier, and Hypertension
 Cognitive Dysfunction and Dementia in Animal Models of Hypertension
 Hypertension and Stroke
 Neuroimaging as a Research Tool in Human Essential Hypertension
 Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
Get Clinical Tree app for offline access
| ||||