Y. Zhu1,2,3,*, C. Li4,*, D. Gao1,2,3, X. Huang5, Y. Zhang1,2,3, M. Ji1,2,3, F. Zheng5, W. Xie1,2,3
1. Peking University Clinical Research Institute, Peking University First Hospital, Beijing, China; 2. PUCRI Heart and Vascular Health Research Center, Beijing, China; 3. Key Laboratory of Molecular Cardiovascular Sciences (Peking University), Ministry of Education, Beijing, China; 4. National Institute of Health Data Science at Peking University, Beijing, China; 5. Department of Clinical Nursing, School of Nursing, Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, China; * Yidan Zhu and Chenglong Li contributed equally to this study.
Corresponding Author: Wuxiang Xie, Peking University Clinical Research Institute, Peking University First Hospital, No. 38 Xueyuan Road, Haidian District, 100191, Beijing, China. Telephone: +86-10-82805564-622; Fax: +86-10-62015547; E-mail: xiewuxiang@hsc.pku.edu.cn; Fanfan Zheng, Department of Clinical Nursing, School of Nursing, Chinese Academy of Medical Sciences & Peking Union Medical College, 33 Ba Da Chu Rd, Shijingshan District, 100144, Beijing, China. E-mail: zhengfanfan@nursing.pumc.edu.cn
J Prev Alz Dis 2024;
Published online May 21, 2024, http://dx.doi.org/10.14283/jpad.2024.91
Abstract
BACKGROUND: Hypertension may harm cognitive performance, but the potential correlates of longitudinal patterns of blood pressure (BP), especially diastolic BP (DBP), to cognition have been unclear.
OBJECTIVES: To examine long-term BP trajectories in relation to subsequent cognitive decline, incident dementia and all-cause mortality in the general population.
DESIGN: Population-based cohort study.
SETTING: Communities in England.
PARTICIPANTS: The study included 7566 participants from the English Longitudinal Study of Ageing (ELSA).
MEASUREMENTS: BP were measured in 1998, 2004, 2008. Group-based trajectory modeling was used to identify long-term patterns of systolic BP (SBP) and DBP. Outcomes including cognitive function, incident dementia, and all-cause mortality were followed up to 10 years.
RESULTS: Five distinct trajectories were identified for SBP and DBP, respectively. The normal-stable trajectory was used as the reference. For cognitive decline, both SBP and DBP trajectories were independently associated with subsequent cognitive decline, with the fastest decline appeared in the high-stable SBP group of 180 mmHg and the low-stable DBP group of 60 mmHg (both P<0.005). For incident dementia, the multivariable adjusted hazard ratio (HR) was also greatest in high-stable group (4.79, 95% confidence interval: 2.84 to 8.07) across all SBP trajectories. Conversely, low (HR: 1.58) and moderate-low stable (HR: 1.56) DBP trajectories increased dementia risk (both P<0.005). Similar patterns were found in BP trajectories in relation to all-cause mortality.
CONCLUSION: Our study evaluates the potential health impact from different BP trajectories and suggests that controlling long-term SBP and maintaining adequate DBP may be relevant for the current practice to promote cognitive health and extend lifespan.
Key words: Blood pressure, long-term trajectory, cognitive health, dementia, all-cause mortality.
Introduction
Dementia is one of the major causes of death and disability worldwide (1). As the global population ages, the number of people with dementia is expected to triple over the next 30 years (2), which illustrates the urgent need to reduce dementia’s onset and impact. Hypertension is highly prevalent in middle-aged and older people, previous work has shown that hypertension is associated with cognitive impartment and is one of the potentially modifiable risks for dementia (3-5). However, many studies have focused on the presence or absence of hypertension, or blood pressure (BP) measures at a single point. As BP fluctuate across the lifespan, which may result from normal daily variations, activities, medication use, ages or medical issues, et.al. Hence, these evidences may therefore be insufficient to fully explain the effect of dynamic BP fluctuation on dementia.
Several studies have used varied methodologies to explore the correlation between BP patterns and dementia risk. Some retrospective analyses have examined BP changes preceding dementia diagnosis, yielding inconsistent results (6-8). Recent research has examined cumulative BP exposure over time, revealing an independent association between elevated cumulative systolic BP (SBP) and dementia risk, with inconclusive findings for cumulative diastolic BP (DBP) (9, 10). Other investigations have explored a link between BP variability and dementia risk. These findings offer crucial insights into individual-level BP patterns and their potential impact on dementia risk. However, existing methods have limitations in comprehensively evaluating longitudinal BP trajectories, as individuals with identical cumulative BP values may exhibit diverse trajectories, leading to varying clinical implications. Limited studies so far have assessed the long-term dynamic BP trajectories based on longitudinal repeated measurements and their associations with subsequent cognitive outcomes. Furthermore, there is inconsistent evidence regarding the independent relevance of DBP and its long-term trajectories to incident dementia and mortality (11-13).
Therefore, the present study possesses two objectives. First, the study aims to identify potential dynamic BP trajectories based on 10-year BP longitudinal measurements, using a group-based trajectory modeling approach. The approach has been well-embraced for examining trajectories of various health indices, fully accounting for population-level heterogeneity in longitudinal trajectories (14). Second, based on a prospective cohort design, the study aims to examine associations between identified BP trajectories and subsequent outcomes, including cognitive decline, risk of dementia and all-cause mortality among dementia-free general middle and aged adults.
Methods
Study design and population
In this longitudinal cohort study, participants were drawn from the English Longitudinal Study of Ageing (ELSA) study (15), a nationally representative community-based cohort study of randomly sampled adults aged 50 years and over in England. The ELSA commenced in 2002-2003 (wave 1), follow-up assessments were performed biennially until 2018-2019 (wave 9). BP was measured every two waves from wave 2. ELSA participants were originally drawn from the Health Survey for England with their BP measured in 1998, which serve as Wave 0 of ELSA. The ELSA study was approved by the London Multicenter Research Ethics Committee (MREC/01/2/91). Written informed consent was obtained from all participants.
For the present study, BP measurements on and before wave 4 were used to assess the trajectories of BP, and the subsequent waves were used to assess the rate of cognitive decline, incidence of dementia and all-cause mortality. The full study design is displayed in Supplemental Figure 1. All participants with 1) cognitive function assessment at baseline (wave 1) and no diagnoses of dementia, determined either by self-reported physician diagnosis or alternative methods based on cognition and activities of daily living, during waves 1 to 4; 2) two or more assessments of BP before wave 4; 3) at least one follow-up after wave 4 was included, resulting in an analytical sample of 7566 men and women.
Assessment of blood pressure
BP was measured using the Omron model HEM-907 Monitor at wave 0 (1998), wave 2 (2004) and wave 4 (2008). Three measurements, one minute interval between each, were taken with the patient been seated and quiet. In each patient the average of the three BP values was taken as the value for the visit. Both systolic and diastolic BP were recorded.
Assessment of cognitive function
Three cognitive function domains, memory, verbal fluency, and temporal orientation were assessed in the cognitive test at each wave. All three tests have been proved with good validity and reliability (16-18). Memory was assessed by both immediate and delayed word recall of 10 unrelated words. Each correct recall was scored 1 point, giving a sum score of 0 to 20. Temporal orientation was evaluated using 4 date naming questions, with 1 point for each correct answer, out of 4 points. Verbal fluency was examined with participants required to verbally name as many animals as possible in 60 seconds, scoring 1 point for each correct answer. For all tests, a higher score indicated better cognitive performance.
Global cognitive function was then evaluated by firstly, calculating Z-scores for each three domains by standardizing the score for each wave to the baseline score, and secondly, calculating the global cognitive Z-score by averaging the Z-scores across the three domains and re-standardizing the scores using the same approach. After the above standardization, a global cognitive Z-score of -1 at any wave represented 1 SD below the baseline mean score. This procedure to generate global cognitive Z scores has been widely applied in previous studies (19, 20).
Dementia and all-cause mortality ascertainment
Dementia was defined by either self-reported physician diagnosis or an alternative approach based on the concurrent cognitive and functional impairment at any single follow-up visit. This method offers a more accurate assessment of dementia onset and has been well validated previously (9, 21). Cognitive impairment was defined as a standardized score of 1.5 standard deviations below the population average in at least one cognitive function domain stratified by educational level. Functional impairment was defined as self-reported difficulty with at least one activity of daily living, including bathing, eating, dressing, getting in and out of bed, and walking across a room. Additionally, we implemented a more stringent definition of the alternative approach, which require that cognitive and functional impairments persist for at least two consecutive waves to exclude potential transient issues. This definition was applied in the sensitivity analysis. Death from any cause was ascertained by the National Health Service Central Register held by the Office of National Statistics. Participants were followed starting from wave 4 till 2018.
Assessment of covariates
Potential confounders, such as demographic characteristics (e.g., age, sex, ethnicity, educational level, cohabitation status) and clinical characteristics, including body mass index, diabetes, hyperlipidemia, heart disease, stroke, chronic lung disease and cancer were treated as fixed covariates measured at wave 1, as they were unlikely to undergo significant changes over time. Time varying covariates, ie, current smoking, alcohol consumption, physical activity, depressive symptoms and antihypertension medication use, were included to account for their possible changes over time. Time varying covariates were calculated by summing the total number of waves (from 1 to 4) reporting the presence of the condition. For instance, a value of 3 for current smoking indicates that smoking was reported in 3 out of 4 waves, while a value of 0 signifies non-smoking reported in all 4 waves, consistent with previous studies (9, 19). Detailed description on covariate definition sees the Supplemental Methods.
Statistical analysis
We used group-based trajectory modeling (GBTM) (22) to identify potential long-term trajectories of systolic and diastolic BP independently based on 3 measurements over a 10-year period from wave 0 to wave 4. These models were fit using SAS Proc Traj. The GBTM estimate the probabilities for multiple trajectories simultaneously as opposed to simply fitting the overall population mean. Then the grouping was achieved by estimating the posterior probabilities of each participant belonging to each potential trajectory group and those with the highest probability was determined as the final group membership. Compared with other trajectory modeling methods, GBTM can adapt to complex and changeable blood pressure trajectories (23). Details on the trajectory modeling process see the Supplemental Methods.
To estimate the association of either SBP or DBP trajectory group on cognitive decline, we used linear mixed models with global Z scores included as dependent variable, and trajectory group membership, time, the trajectory group × time interaction term and covariates as independent variables. The intercept and slope of time were fitted as random effects at the participant level. The β coefficient for the group × time interaction term and its 95% confidence interval (CI) were reported. The linear mixed model (24) could appropriately handle random missing cognitive function data and has been widely used.
We used multivariable Cox proportional hazards regression models to derive hazard ratio (HR) and 95% CI for the association between BP trajectories and incident dementia or all-cause mortality. The analyses were adjusted for covariates, with cancer and chronic lung disease included exclusively as covariates for the mortality outcome, given their specific relevance to mortality risk. Individuals contributed person time from the wave 4 until the date of event or censoring. Proportional hazards assumption was assessed by weighted Schoenfeld residuals, whereas age violated this assumption and therefore its interaction term with time was included in the model.
Subgroup analyses were performed based on predefined subgroups, including age, sex and long-term antihypertensive medication use, to explore potential effect modifications. Additionally, several sensitivity analyses were conducted. First, we examined the associations between BP trajectories and subdomain cognitive decline, specifically memory, verbal fluency, and orientation. Second, considering a potential ceiling effect in the orientation domain, we recalculated a new global Z score based solely on memory and verbal fluency, and repeated the analyses. Third, we conducted competing risk analysis using Fine and Gray methods to assess the association with the risk of dementia. Fourth, we applied a stricter definition of supplementary dementia, requiring persistent cognitive and functional impairments for at least 2 consecutive waves, and repeated the Cox analysis based on this criterion.
Data were analyzed using SAS 9.4 (SAS Institute, Cary, North Carolina, United States of America) and statistical testing was conducted at a 2-tailed α level of 0.05.
Results
Baseline characteristics
The cohort included 7566 participants. During a median follow-up of 8.0 [4.0-8.0] years, 771 participants (10.2%) developed dementia and 495 participants (6.5%) died. Flow chart showing process of inclusion of the study population was shown in Supplemental Figure 2. Table 1 and 2 provides participant characteristics by SBP and DBP trajectory groups, respectively. In general, five trajectories were identified for both SBP and DBP. Each trajectory group remained stable with the exception of one group each for SBP and DBP, which experienced rapid decreases towards normal values over time (Figure 1). Predicted group membership for each SBP trajectory was as follows, 26.5% (2008) in the normal-stable group, 55.2% (4180) in the normal-high stable group, 13.7% (1038) in the moderate-stable group, 3.4% (255) in the high-falling group and 1.1% (85) in the high stable group. The five DBP trajectories were labeled as normal-stable, low-stable, moderate-low stable, moderate-high stable and moderate-decreasing, the predicted group membership for each was 40.1% (3032), 6.8% (514), 46.8% (3545), 4.9% (368) and 1.4% respectively. Individuals from the trajectory group that either maintained at high SBP or low DBP were more likely to be older, men and to have higher BMI, worse health status and worse cognitive function.
Table 1. Baseline characteristics of study population by trajectories of systolic blood pressures
a. Values are mean (standard deviation) or median (quartile 1–quartile 3) for continuous variables and number (%) for categorical variables. b. P value reported for differences between trajectories using t test, chi-square test, or Wilcoxon rank test.
Table 2. Baseline characteristics of study population by trajectories of diastolic blood pressures
a. Values are mean (standard deviation) or median (quartile 1–quartile 3) for continuous variables and number (%) for categorical variables. b. P value reported for differences between trajectories using t test, chi-square test, or Wilcoxon rank test.
Figure 1. Trajectories of systolic and diastolic blood pressure
The graphs show the SBP trajectory (top panel) and the DBP trajectory (bottom panel) measured over 10 years. There are five trajectories for both SBP and DBP. Each color represents a BP trajectory.
Associations of trajectories of BP with cognitive decline
Table 3 shows the association between BP trajectories and rates of cognitive decline. Using normal-stable SBP trajectory as a reference, long-term maintenance of higher SBP was independently associated with subsequent faster cognitive decline (all p value<0.05) after adjustment for covariates. The fastest decline appeared in the high-stable trajectory whose SBP maintained at 180 mmHg (β=-0.076, 95% CI: -0.124 to -0.029 SD per year). For DBP, trajectories with lower DBP were associated with faster cognitive decline compared with normal-stable DBP trajectory. Of note, the low-stable group, with DBP maintained at 60 mmHg, exhibited the fastest rate of cognitive decline (β=-0.020, 95%CI: -0.038 to -0.003 SD per year).
Table 3. Cognitive impact of trajectories of blood pressures
a. Adjusted covariates included age, sex, ethnicity, education, cohabitation status, body mass index, diabetes, stroke, cardiovascular diseases, hyperlipidemia as well as total number of waves reporting current smoking, alcohol consumption, physical activity, depressive symptoms, and antihypertensive medication usage.
Associations of trajectories of BP with risk of dementia and all-cause mortality
The findings for SBP trajectories on incident dementia were very similar to that of cognitive decline (Figure 2). Compared with the normal-stable SBP trajectory, the high-stable group had the largest adjusted HR of 4.79 (95% CI: 2.84 to 8.07) for dementia. In addition, the adjusted HRs of the normal-high stable, moderate-stable and high-falling group were 1.56 (1.28 to 1.90), 2.13 (1.66 to 2.73) and 2.04 (1.36 to 3.05), respectively.
Figure 2. Associations of blood pressure trajectories with incident dementia and all-cause mortality
HRs (95% CIs) for incident dementia and all-cause mortality are reported separately for each trajectory group for SBP (top panel) and DBP (bottom panel), by using low-stable trajectory as a reference. Adjusted covariates for dementia outcome included sex, age × time, ethnicity, education, cohabitation status, body mass index, diabetes, stroke, cardiovascular diseases, hyperlipidemia as well as total number of waves reporting current smoking, alcohol consumption, physical activity, depressive symptoms, and antihypertensive medication usage. The mortality outcome was additionally adjusted for cancer and chronic lung disease.
For trajectories of DBP, compared to normal stable group, greater risk of developing dementia was found in individuals maintained at lower DBP levels, but not in those at high levels. Specifically, the adjusted HR was 1.58 (1.16 to 2.14) for low-stable group and 1.56 (1.33 to 1.83) for moderate-low stable group. The association were attenuated and nonsignificant for moderate-high stable (HR:1.01, 95%CI: 0.70 to 1.45) and moderate-decreasing group (HR:1.35, 95%CI: 0.77 to 2.38). Findings were similar for all-cause mortality, which was, SBP trajectory maintained at high levels or DBP trajectory maintained at low level independently predicted higher risk of all-cause mortality.
Sensitivity Analyses
The associations between BP trajectories and subdomain cognitive decline are shown in Table S2-S4. We found broadly similar associations within each subdomain for each BP trajectory pattern, particularly in memory and orientation, with relatively weaker associations in verbal fluency. Additionally, the association of BP trajectories with the recalculated global Z score, excluding the orientation domain, remained consistent, although the estimates appeared weaker (see Table S5). Furthermore, regarding incident dementia, the results were similar in analyses accounting for the competing risk of death, while the association was slightly attenuated (see Table S6). Lastly, the reanalyzed results based on a stricter definition of dementia, requiring cognitive and functional impairment to persist for at least two waves, remained consistent, and the association estimates appeared stronger for low DBP trajectories (see Table S7).
Findings were generally consistent in subgroup analyses (see Figures S3-8) but there was heterogeneity in some subgroups. Male who maintained at lower DBP levels tended to have faster cognitive decline and higher risk of mortality, while women from any SBP trajectory had greater risk of mortality. Besides, older individuals tended to have faster cognitive decline, but their dementia risk did not appear to differ significantly.
Discussion
In this longitudinal cohort of older individuals, we described long-term BP trajectories in the general population using a trajectory approach. Our analysis revealed that that SBP and DBP trajectories each independently predict subsequent cognitive decline, dementia incidence, and all-cause mortality. The findings showed that not only participants whose SBP maintained a stably higher levels at 180 mmHg, but also those DBP remained below 70 mmHg, had the fastest cognitive decline and the highest risk of dementia and all-cause mortality. The nominally lowest risk was found in participants who maintained BP levels at normal levels. These associations were independent of levels and changes in known risk factors, such as age, education level, physical activity, body mass index, medical history and anti-hypertensive drug use.
Our study complements previous findings by detecting longitudinal BP trajectories using repeated BP measurements over a decade and examining their link to dementia risk. While many studies have relied on single or average BP values, others have explored long-term BP patterns through methods like cumulative BP exposure, BP variability, or average changes over time. Our prior studies on the cognitive impact of BP fluctuations found that both long-term cumulative BP9 and short-term time in target range BP25 independently predict subsequent cognitive outcomes. Researches based on data from the Rotterdam Study and the Framingham Heart Study have also examined these topics, with somewhat inconsistent conclusions (10, 26). On this basis, our current study adds to the existing discussion by further examining the predictive value of BP trajectories on cognitive outcomes. Trajectory methods offer the advantage of being able to capture long-term BP fluctuations and classify individuals into distinct classes based on personal response patterns over an extended period of time, which provides a more accurate understanding of long-term BP patterns. For instance, individuals with similar cumulative BP levels may belong to different trajectory groups, exhibiting varying patterns such as decrease, increase, or stability. Thus, longitudinal trajectories can address specific issues that other approaches may overlook and provide more targeted insights for potential interventions.
We observed accelerated cognitive decline and increased risk of dementia among trajectory members who stabilized at high SBP levels. Numerous previous studies have confirmed that hypertension, especially high BP during midlife, is a modifiable risk factor for dementia (27, 28). Our observation was in line with previous studies using a single value to measure BP. Currently, fewer studies have examined BP trajectories and subsequent cognitive outcomes. A study observed a significant association between SBP trajectories and dementia in middle-aged women not receiving antihypertensive treatment, but not with those treated with antihypertensives (6).Together with existing evidence, we suggest the value of focusing on long-term BP patterns while maintaining monitoring in maintaining cognitive health.
Our findings add new information to existing knowledge on long-term patterns of DBP and find that risk in dementia and death is elevated when DBP remains below 70 mmHg for a long period of time, and higher when DBP is maintained at 60 mmHg. Unlike the strong evidence on the effect of SBP on adverse events, previous reports on DBP have varied. Debate continues on the topic of diastolic J-shape phenomenon on cardiovascular events (29, 30), and findings linking DBP levels to dementia and death also remains controversial (11, 12, 31). A study using a similar analytic method found no clear association between DBP trajectory and dementia, but all trajectories they detected at a DBP value of greater than 80 mmHg (8). A post hoc analysis of the SPRINT MIND trial (32) demonstrated that in patients with low baseline DBP (ie, <70mmHg), intensive versus standard BP treatment did not appear to have a detrimental effect on dementia, this conclusion was derived from a P value of 0.06 for heterogeneity in dementia efficacy across DBP quartiles. Nevertheless, individuals with higher DBP appeared to have greater benefit. Interestingly, another two cohort studies (13, 33) based on data from the SPRINT trial found that lowering DBP to 60 mmHg or lower baseline DBP increased risk of a composite of all-cause death and cardiovascular adverse events, even in patients with SBP less than 130mmHg, and suggested an optimum target of DBP between 70 to 80 mmHg. A meta-analysis comparing 138 studies demonstrated a significant association between late-life low DBP and dementia34. Our recent research on cumulative BP also supported an inversely association of DBP levels with dementia and all-cause mortality (9). Taken together, the results suggest a potential preventive strategy of maintaining adequate DBP in the progress of dementia and all-cause mortality.
The association between BP and cerebrovascular dysfunction has been shown to be age-related (5). We observed a stronger association between higher SBP and lower DBP trajectories with cognitive decline in individuals aged 65 and above. Older individuals often have reduced vascular elasticity, which affects cerebral blood flow and thus cognitive functioning (35). Additionally, age-related physiological and metabolic changes, such as neurohumoral activation and endothelial dysfunction, may increase their sensitivity to BP changes and consequently affect cognitive function. Furthermore, we found that men with lower DBP levels experience accelerated cognitive decline compared to women. This gender difference may be linked to specific pathways and the higher traditional vascular risk burden in men, which are also risk factors for late-life dementia (36). These findings provide insights for the development of personalized cognitive care strategies, indicating that personalized blood pressure control may be may be a potential approach to enhancing cognitive health.
Our findings provide important implications for clinical practice and public health actions, further highlighting the significance of monitoring BP longitudinally for the primary prevention of cognitive decline and dementia, along with novel insights into long-term DBP management. The current study extends existing findings to provide a unique insight into the nature of dynamic changes in BP, and quantifying the potential health impact from different SBP and DBP trajectories over an extended period. The results indicate a potential increased risk of dementia and all-cause mortality in older individuals exhibiting specific BP patterns, i.e., sustained high SBP levels, particularly above 180 mmHg. For years, there have been debates over the necessity of lowering DBP, especially in aged population. The excessively low DBP has been associated with elevated risks of cardiovascular events and mortality (37). Our study further demonstrates the adverse neurocognitive consequences of maintaining low DBP, below 70 mmHg, for a decade. While causality cannot be implied from observational study, our findings suggest that long-term BP patterns may serve as potential predictors to cognitive impairment in older adults. These findings contributions necessitate through considerations to develop the optimal BP-lowering regime for older individuals, particularly those with excessively high SBP and low DBP. Further randomized controlled trials are therefore warranted, to validate our results by examining the neurocognitive benefits of lowering long-term SBP and maintaining adequate DBP.
The potential biological mechanisms underlying the impact of blood pressure on cognitive health are not well understood, especially for the observed detrimental neurocognitive consequence of low DBP. Existing explorations mainly focus on cerebrovascular dysfunction, suggesting that changes in the structure and function of cerebrovascular endothelial cells induced by hypertension will lead to increased blood-brain barrier permeability and microhemorrhages, resulting in hypoperfusion that may contribute to cognitive impairment (38). In our study, individuals with prolonged low DBP had correspondingly lower baseline SBP, indicating that vascular fragility is unlikely to be a potential explanation for the association of low DBP with cognitive outcomes. Executive function, a key cognitive domain affected by vascular cognitive impairment (39), is supported by our results showing a correlation with higher SBP trajectories and decreased verbal fluency. However, no significant association was observed with DBP. This implies that sustained high SBP, rather than low DBP, may have a greater impact on executive function through cerebral vascular dynamics, indicating potential differences in cognitive function impact based on different BP indicators. Additionally, we hypothesized left ventricular hypertrophy (LVH) as a possible explanation for the impact of low DBP, as long-term elevated SBP is linked to LVH, leading to diastolic dysfunction characterized by extremely low DBP (40). This is supported by previous research showing an association between increased left ventricular mass and accelerated cognitive decline, independent of arterial stiffness (41). Nevertheless, further research is necessary to elucidate potential differences in the mechanisms linking different BP patterns to cognitive health.
There are several strengths of the current study. First, this is a longitudinal study with national representativeness, a relatively large sample size, and long-time follow-up of twenty years. Second, estimates of BP trajectories over the 10 years prior to outcome assessment provide insights into long-term BP patterns and their potential associations with cognitive outcomes and mortality. Third, the use of reliable method in cognitive outcomes assessment and a covariate-adjustment method that takes into account time-varying variables renders our findings robust.
This study has certain limitations that need to be considered. First, given that this is an observational study, we cannot rule out concerns about reverse causality. Second, although we adjusted for both potentially important fixed and time-varying covariates, residual confounding is still possible. Third, the small number of individuals in few trajectories may reduce statistical power and introduce random errors, although the point estimate suggests a stronger association. Fourth, there may be some misclassification of dementia. Although we considered that relying solely on self-reported dementia may underestimate dementia cases (42) and therefore supplement dementia diagnoses based on cognitive and functional status, the presence of undiagnosed dementia cannot be completely ruled out. Fifth, the global cognitive function assessment may be biased by a ceiling effect in the orientation evaluation. Nevertheless, recalculating the global cognitive score without the orientation domain yielded consistent results, suggesting minimal bias impact. Sixth, the participants excluded from this analysis were more likely to be older, less educated and have more female than the included participants, which might limit the generalizability of our findings to original study population of ELSA. Additionally, attrition bias due to differences in participant dropouts or losses during follow-up between BP trajectories may potentially affect conclusions.
Conclusion
In this large population-based cohort of older individuals, the long-term trajectories of BP and their potential health impact provide additional information on dementia risk stratification. Notably, not only prolonged periods of high SBP, but also persistently low DBP, are linked to accelerated cognitive decline, increased risk of dementia and all-cause-mortality. These findings suggest that controlling long-term SBP and maintaining adequate DBP may offer opportunities to promote cognitive health and extend lifespan.
Acknowledgments: We appreciate all the participants of the English Longitudinal Study of Ageing and acknowledge the efforts made by the original data creators, depositors, copyright holders, the funders of the data collections and their contributions for access of data from the study.
Funding: The present study was supported by the National Natural Science Foundation of China (project number 72304016 and 82373665). The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Author contributions: WX, YZ, CL conceived and designed the study. YZ and FZ obtained funding. YZ and CL analyzed the data. YZ drafted the manuscript. DG, XH and MJ performed the raw data checking. All authors contributed to the critical interpretation of data, and have reviewed and approved the submission of the manuscript.
Competing interests: None reported.
Ethical standards: The ELSA study was approved by the London Multicenter Research Ethics Committee (MREC/01/2/91). Written informed consent was obtained from all participants. All methods were performed in accordance with the relevant guidelines and regulations, including the principles of the Declaration of Helsinki.
References
1. Livingston G, Sommerlad A, Orgeta V, et al. Dementia prevention, intervention, and care. Lancet. Dec 16 2017;390(10113):2673-2734. https://doi.org/10.1016/s0140-6736(17)31363-6
2. Patterson C. World alzheimer report 2018. The State of Art of Dementia Research: New Frontiers; Alzheimer’s Disease International: London, UK, 2018; pp. 34–40
3. Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. Aug 8 2020;396(10248):413-446. https://doi.org/10.1016/s0140-6736(20)30367-6
4. Roberts RO, Cha RH, Mielke MM, et al. Risk and protective factors for cognitive impairment in persons aged 85 years and older. Neurology. May 5 2015;84(18):1854-61. https://doi.org/10.1212/wnl.0000000000001537
5. Qiu C, Winblad B, Fratiglioni L. The age-dependent relation of blood pressure to cognitive function and dementia. Lancet Neurol. Aug 2005;4(8):487-99. https://doi.org/10.1016/s1474-4422(05)70141-1
6. Joas E, Bäckman K, Gustafson D, et al. Blood pressure trajectories from midlife to late life in relation to dementia in women followed for 37 years. Hypertension. Apr 2012;59(4):796-801. https://doi.org/10.1161/hypertensionaha.111.182204
7. Lee ATC, Fung AWT, Richards M, et al. Late-life longitudinal blood pressure trajectories as predictor of dementia. Sci Rep. Jan 31 2022;12(1):1630. https://doi.org/10.1038/s41598-022-05680-3
8. Cheng G, He S, He Q, et al. Trajectory patterns of blood pressure change up to six years and the risk of dementia: a nationwide cohort study. Aging (Albany NY). Jul 1 2021;13(13):17380-17406. https://doi.org/10.18632/aging.203228
9. Li C, Zhu Y, Ma Y, Hua R, Zhong B, Xie W. Association of Cumulative Blood Pressure With Cognitive Decline, Dementia, and Mortality. J Am Coll Cardiol. Apr 12 2022;79(14):1321-1335. https://doi.org/10.1016/j.jacc.2022.01.045
10. Kim H, Ang TFA, Thomas RJ, Lyons MJ, Au R. Long-term blood pressure patterns in midlife and dementia in later life: Findings from the Framingham Heart Study. Alzheimers Dement. Oct 2023;19(10):4357-4366. https://doi.org/10.1002/alz.13356
11. Abell JG, Kivimäki M, Dugravot A, et al. Association between systolic blood pressure and dementia in the Whitehall II cohort study: role of age, duration, and threshold used to define hypertension. Eur Heart J. Sep 1 2018;39(33):3119-3125. https://doi.org/10.1093/eurheartj/ehy288
12. Verghese J, Lipton RB, Hall CB, Kuslansky G, Katz MJ. Low blood pressure and the risk of dementia in very old individuals. Neurology. Dec 23 2003;61(12):1667-72. https://doi.org/10.1212/01.wnl.0000098934.18300.be
13. Li J, Somers VK, Gao X, et al. Evaluation of Optimal Diastolic Blood Pressure Range Among Adults With Treated Systolic Blood Pressure Less Than 130 mm Hg. JAMA Netw Open. Feb 1 2021;4(2):e2037554. https://doi.org/10.1001/jamanetworkopen.2020.37554
14. Zhu Y, Li C, Xie W, Zhong B, Wu Y, Blumenthal JA. Trajectories of depressive symptoms and subsequent cognitive decline in older adults: a pooled analysis of two longitudinal cohorts. Age Ageing. Jan 6 2022;51(1). https://doi.org/10.1093/ageing/afab191
15. Steptoe A, Breeze E, Banks J, Nazroo J. Cohort profile: the English longitudinal study of ageing. Int J Epidemiol. Dec 2013;42(6):1640-8. https://doi.org/10.1093/ije/dys168
16. Bloomberg M, Dugravot A, Dumurgier J, et al. Sex differences and the role of education in cognitive ageing: analysis of two UK-based prospective cohort studies. Lancet Public Health. Feb 2021;6(2):e106-e115. https://doi.org/10.1016/s2468-2667(20)30258-9
17. Zheng F, Yan L, Yang Z, Zhong B, Xie W. HbA(1c), diabetes and cognitive decline: the English Longitudinal Study of Ageing. Diabetologia. Apr 2018;61(4):839-848. https://doi.org/10.1007/s00125-017-4541-7
18. Baars MA, van Boxtel MP, Dijkstra JB, et al. Predictive value of mild cognitive impairment for dementia. The influence of case definition and age. Dement Geriatr Cogn Disord. 2009;27(2):173-81. https://doi.org/10.1159/000200465
19. Zhu Y, Li C, Wu T, et al. Associations of cumulative depressive symptoms with subsequent cognitive decline and adverse health events: Two prospective cohort studies. J Affect Disord. Jan 1 2023;320:91-97. https://doi.org/10.1016/j. jad.2022.09.128
20. Xie W, Zheng F, Yan L, Zhong B. Cognitive Decline Before and After Incident Coronary Events. J Am Coll Cardiol. Jun 25 2019;73(24):3041-3050. https://doi.org/10.1016/j.jacc.2019.04.019
21. Ahmadi-Abhari S, Guzman-Castillo M, Bandosz P, et al. Temporal trend in dementia incidence since 2002 and projections for prevalence in England and Wales to 2040: modelling study. Bmj. Jul 5 2017;358:j2856. https://doi.org/10.1136/bmj. j2856
22. Nagin DS. Analyzing developmental trajectories: a semiparametric, groupbased approach. Psychological Methods. 1999;4:139-57.
23. Franklin JM, Shrank WH, Pakes J, et al. Group-based trajectory models: a new approach to classifying and predicting long-term medication adherence. Med Care. Sep 2013;51(9):789-96. https://doi.org/10.1097/MLR.0b013e3182984c1f
24. Pigott TD. A review of methods for missing data. Educational research and evaluation. 2001;7(4):353-383.
25. Huang X, Deng S, Xie W, Zheng F. Time in target range of systolic blood pressure and cognitive outcomes in patients with hypertension. J Am Geriatr Soc. Nov 2 2023. https://doi.org/10.1111/jgs.18641
26. Ma Y, Wolters FJ, Chibnik LB, et al. Variation in blood pressure and longterm risk of dementia: A population-based cohort study. PLoS Med. Nov 2019;16(11):e1002933. https://doi.org/10.1371/journal.pmed.1002933
27. Gottesman RF, Schneider AL, Albert M, et al. Midlife hypertension and 20- year cognitive change: the atherosclerosis risk in communities neurocognitive study. JAMA Neurol. Oct 2014;71(10):1218-27. https://doi.org/10.1001/jamaneurol.2014.1646
28. Walker KA, Sharrett AR, Wu A, et al. Association of Midlife to Late-Life Blood Pressure Patterns With Incident Dementia. Jama. Aug 13 2019;322(6):535-545. https://doi.org/10.1001/jama.2019.10575
29. Rapsomaniki E, Timmis A, George J, et al. Blood pressure and incidence of twelve cardiovascular diseases: lifetime risks, healthy life-years lost, and age-specific associations in 1·25 million people. Lancet. May 31 2014;383(9932):1899-911. https://doi.org/10.1016/s0140-6736(14)60685-1
30. McEvoy JW, Chen Y, Rawlings A, et al. Diastolic Blood Pressure, Subclinical Myocardial Damage, and Cardiac Events: Implications for Blood Pressure Control. J Am Coll Cardiol. Oct 18 2016;68(16):1713-1722. https://doi.org/10.1016/j. jacc.2016.07.754
31. Iseki K, Miyasato F, Tokuyama K, et al. Low diastolic blood pressure, hypoalbuminemia, and risk of death in a cohort of chronic hemodialysis patients. Kidney Int. Apr 1997;51(4):1212-7. https://doi.org/10.1038/ki.1997.165
32. Jiang C, Li S, Wang Y, et al. Diastolic Blood Pressure and Intensive Blood Pressure Control on Cognitive Outcomes: Insights From the SPRINT MIND Trial. Hypertension. Mar 2023;80(3):580-589. https://doi.org/10.1161/hypertensionaha.122.20112
33. Beddhu S, Chertow GM, Cheung AK, et al. Influence of Baseline Diastolic Blood Pressure on Effects of Intensive Compared With Standard Blood Pressure Control. Circulation. Jan 9 2018;137(2):134-143. https://doi.org/10.1161/circulationaha.117.030848
34. Ou YN, Tan CC, Shen XN, et al. Blood Pressure and Risks of Cognitive Impairment and Dementia: A Systematic Review and Meta-Analysis of 209 Prospective Studies. Hypertension. Jul 2020;76(1):217-225. https://doi.org/10.1161/hypertensionaha.120.14993
35. Rouch L, Cestac P, Sallerin B, et al. Visit-to-Visit Blood Pressure Variability Is Associated With Cognitive Decline and Incident Dementia: The S.AGES Cohort. Hypertension. Oct 2020;76(4):1280-1288. https://doi.org/10.1161/hypertensionaha.119.14553
36. Ernst ME, Ryan J, Chowdhury EK, et al. Long-Term Blood Pressure Variability and Risk of Cognitive Decline and Dementia Among Older Adults. J Am Heart Assoc. Jul 6 2021;10(13):e019613. https://doi.org/10.1161/jaha.120.019613
37. Messerli FH, Panjrath GS. The J-curve between blood pressure and coronary artery disease or essential hypertension: exactly how essential? J Am Coll Cardiol. Nov 10 2009;54(20):1827-34. https://doi.org/10.1016/j.jacc.2009.05.073
38. Ungvari Z, Toth P, Tarantini S, et al. Hypertension-induced cognitive impairment: from pathophysiology to public health. Nat Rev Nephrol. Oct 2021;17(10):639-654. https://doi.org/10.1038/s41581-021-00430-6
39. van der Flier WM, Skoog I, Schneider JA, et al. Vascular cognitive impairment.Nature Reviews Disease Primers. 2018/02/15 2018;4(1):18003. https://doi.org/10.1038/nrdp.2018.3
40. Lorell BH, Apstein CS, Weinberg EO, Cunningham MJ. Diastolic function in left ventricular hypertrophy: clinical and experimental relationships. Eur Heart J. Nov 1990;11 Suppl G:54-64. https://doi.org/10.1093/eurheartj/11.suppl_g.54
41. Scuteri A, Coluccia R, Castello L, Nevola E, Brancati AM, Volpe M. Left ventricular mass increase is associated with cognitive decline and dementia in the elderly independently of blood pressure. Eur Heart J. Jun 2009;30(12):1525-9. https://doi.org/10.1093/eurheartj/ehp133
42. McGrath R, Robinson-Lane SG, Clark BC, Suhr JA, Giordani BJ, Vincent BM. Self-Reported Dementia-Related Diagnosis Underestimates the Prevalence of Older Americans Living with Possible Dementia. J Alzheimers Dis. 2021;82(1):373-380. https://doi.org/10.3233/jad-201212
© Serdi 2024