are noise and air pollution related to the incidence of dementia? a cohort study in london, england

Objective to investigate whether the incidence of dementia is related to air and noise pollution levels in London residents.
Retrospective cohort studies were designed using primary care data.
Set up 75 Great London practices.
Participants were 50-79-year-old adults registered in the general clinic in January 1, 2005, and there was no record of living in dementia or nursing homes.
Primary and secondary results measured the first recorded diagnosis of dementia, as well as the Alzheimer\'s disease and vascular dementia subgroup, in 3-13.
Annual average concentration of carbon dioxide in 2004 (NO2)
, Particulate matter with median aerodynamic diameter ≤ 2. 5u2009µm (PM2. 5)and ozone (O3)
The estimated resolution from the decentralized model is 20 × 20 µm.
Traffic intensity, main road distance and night distance
Level of time noise (Lnight)
It is estimated to be the postal code level.
All exposure measures are linked anonymously to clinical data through residential ZIP codes.
The hr of the Cox model was adjusted according to age, gender, race, smoking and body mass index and further adjusted for regional deprivation and co-morbidity.
Results 2181 subjects (1. 7%)
Received an event diagnosis of dementia (
39% referring to Alzheimer\'s disease, 29% vascular dementia).
There is a positive exposure response relationship between dementia and all air pollution measures other than ozone, which is not easily explained by further adjustments.
Adults living in areas with the fifth highest concentration of NO2 (>41. 5u2009µg/m3)
Compared with the lowest fifth place (0. 95)
, And repeated analysis using alternate years, resulting in the same result (data not shown).
The simulated annual concentration of air pollutants was estimated using the KCLurban diffusion modeling system with a resolution of 20 × 20 µm.
17 it includes hourly meteorological measurements, NO-based on experience-NO2-
Ozone and derived PM (
Particulate matter)
, Using source emission information from London\'s atmospheric emissions list.
18 in this analysis, we focused on the annual concentrations of no 2, pm2 in 2004. 5 (
Fine particulate matter with median aerodynamic diameter ≤ 2. 5u2009µm)and O3 (ozone).
In addition, we give an estimate of primary pm2. 5.
5 Road traffic sources that can be attributed to the sum of the emission source contributions estimated below: tires, brakes, exhaust, surface wear and re-suspension.
Based on the interpolation of the most recent 20 × 20 µm points of the discrete model, 17 air pollution concentration estimates were derived for each of the 190115 London postal code address centers.
This resulted in patients living in the same zip code being assigned to the same level of exposure, typically about 15 households nationwide on average.
In addition, some zip codes covering smaller geographic areas may also be assigned to the same 20 × 20 CUCM points.
The \"trusted third party\" linked the exposure concentration to the CPRD and they subsequently removed the zip code to ensure that we were not able to access any geographic identifiers directly.
Traffic proximity measures related to the density of \"heavy\" vehicles have been developed, which are defined as light trucks, heavy trucks (
Truck/truck)
Bus and coach.
We include a distance measurement (in metres)
From the postal code center to the nearest roads, these roads are divided into four points at the top of the strength of heavy vehicles.
The traffic intensity is estimated to be the total mileage of the vehicle (
Heavy vehicles only)
Each year, for all major roads within the radius of the center of mass of the postal code address 50 m and 50 m.
We use arbitrary cutting.
Under the drive of defining \"high capacity\" in the analysis, turn off> 100000.
Estimation of road traffic noise levels using traffic noise exposure (TRANEX)model.
This uses information on road traffic flow and speed, road geography, land cover and building height to estimate the average sound pressure level (LAeq)in decibels (dB)
In different time periods.
The assessment of TRANEX by the other two UK cities showed a high correlation (r≥0. 85)
Between modeling and measurement 1-hour LAeq.
In our analysis, we focused on the average annual Lnight recorded overnight between 23: 00 and 07: 00, as this period represents the place of residence of most of our subjects.
Alternative analysis using daytime noise (LAeq16)
The same result was produced (data not shown)
Due to the high correlation with night noise (r=0. 999).
20 in order to connect the noise model with the CPRD, use the geometric center of the address location directly in each ZIP code.
The cohort was defined in 75 practices, with a total of 55 5-385 patients actively registered on 1 January 5, accounting for about 7% of the population of Greater London at that time.
From these, we chose 1397718 adults between the ages of 50-79 who have been registered for more than 1 year in a row.
From this group, 131 ÷ 869 (94%)
Successfully linked to our pollution exposure. Non-
The contact was primarily due to the fact that some practices were close to the boundaries of the study area, so the personal zip code of many of their patients was not eligible.
The patient was tracked to the earliest (1)
The date of the first diagnosis of dementia ,(2)
Date of death or cancellation of practice ,(3)
Date on which the practice stops contributing to CPRDand (4)31December2013.
We used the dementia reading code within the framework of quality and results to search for primary health care records for the first dementia diagnosis date (QOF).
21 although most read code is right
Specifically, we identified a subgroup that classified dementia as Alzheimer\'s disease or vascular dementia.
Using the death record, we also used the revised code for the tenth edition of the International Classification of Diseases to identify patients with dementia who were listed as the main cause of death (
If it is not coded on the general practitioner (GP)record)
And further classify them based on specific references to Alzheimer\'s disease or vascular dementia anywhere in the death record.
We also extracted from GP records about race, smoking, body mass index (BMI)
And alcohol consumption, using the last measurement before baseline, if only the measurement before baseline, or during the study period.
A missing category is retained where there is no information.
In addition, based on the list of conditions previously shown for the independent prediction of mortality, we extracted the co-disease information recorded at baseline. 22Patients (n=391)
Was excluded from the existing diagnosis of 1 January dementia.
In addition, we try to exclude patients (n=423)
When there is evidence on their medical records that as of 1 January 5, they live in a residential or nursing home.
We do this by reading the combination of codes, or by recording four or more patients aged 65-99 living at the same address.
We also rule out a small number of patients (n=77)
Places where IMD cannot be assigned.
This has enabled our analysis to reach 978 of the eligible patients at 30 Mbps.
Statistical analysis we used the Cox proportional risk model to investigate the association between all exposure measurements estimated in the year prior to baseline (2004)
And the subsequent incidence of dementia.
Explain clustering through practice (
Homogeneity between patients from the same practice)
We use a common weakness (
At the practical level)
, Which is equivalent to the survival data of random effects.
We are right (1)
Age, gender, ethnicity, smoking, BMI (
Alcohol consumption recorded is not independently predicted, but removed from the model), (2)IMD decile (
According to the UK National Rankings)and(3)comorbidities (
Coronary heart disease (IHD)
Heart failure, diabetes, stroke)
This is an independent predictor of dementia. For model (2)
We discussed the effect of adjusting other pollutants (
Air pollution caused by noise and vice versa).
In sensitivity analysis, we explore
Between practice and
Practice the effect by fitting different models (a non-
According to the fragile model of practice layering, as practice of fixed effect fitting, fitting practice average exposure level in addition to individual level).
For all contaminants, we fitted a continuous measurement based on the IQR to achieve a comparison between the estimated value and the five-digit number, thus describing the shape of the association.
All analyses were performed in Stata V. 13.
Patient and Public participation without patient involvement in the overall design of research issues, outcome measurements, and research.
Due to the anonymity of the patient, we were unable to disseminate the findings directly to the study participants.
Data Link for Clinical Practice Research (CPRD)
It is a large, validated primary care database that has been collecting anonymous patient data from participating UK general practice since 1987.
It has proven to be a broad representation of the UK\'s population, with about 7% actively registered in the UK in 2011.
About 3-
Several quarters of the UK\'s contribution to CPRD practice have agreed to link their data to external sources.
This is facilitated by CPRD\'s \"trusted third party\" to ensure that researchers do not have access to geographic identifiers such as residential postal codes.
Key variables already linked to practice data include the multiple deprivation index (IMD)
Small composite material-area (
About 1500 people)
Measures used in the UK to allocate resources, 15 and the Office of National Statistical death registration data, have enabled the root cause of death to be determined.
In this study, we chose 75. linked)
Practice in the research area located within the geographical scope of the M25 expressway around Greater London (
60 outside London and 15 within London)
Keep recording data between 2004and2010.
16 estimates of air and noise pollution exposure the following exposure measurements estimated annually for the period 2004 to 2010 are related to CPRD :(1)
Concentration of air pollution ,(2)
Measurement of traffic intensity or distance, and (3)
Noise level of road traffic.
First of all, we chose to focus on the annual concentration estimated in 2004 as it was the earliest year.
Simulation estimates for other years (2005–2010)
Highly relevant (r>0. 95)
, And repeated analysis using alternate years, resulting in the same result (data not shown).
The simulated annual concentration of air pollutants was estimated using the KCLurban diffusion modeling system with a resolution of 20 × 20 µm.
17 it includes hourly meteorological measurements, NO-based on experience-NO2-
Ozone and derived PM (
Particulate matter)
, Using source emission information from London\'s atmospheric emissions list.
18 in this analysis, we focused on the annual concentrations of no 2, pm2 in 2004. 5 (
Fine particulate matter with median aerodynamic diameter ≤ 2. 5u2009µm)and O3 (ozone).
In addition, we give an estimate of primary pm2. 5.
5 Road traffic sources that can be attributed to the sum of the emission source contributions estimated below: tires, brakes, exhaust, surface wear and re-suspension.
Based on the interpolation of the most recent 20 × 20 µm points of the discrete model, 17 air pollution concentration estimates were derived for each of the 190115 London postal code address centers.
This resulted in patients living in the same zip code being assigned to the same level of exposure, typically about 15 households nationwide on average.
In addition, some zip codes covering smaller geographic areas may also be assigned to the same 20 × 20 CUCM points.
The \"trusted third party\" linked the exposure concentration to the CPRD and they subsequently removed the zip code to ensure that we were not able to access any geographic identifiers directly.
Traffic proximity measures related to the density of \"heavy\" vehicles have been developed, which are defined as light trucks, heavy trucks (
Truck/truck)
Bus and coach.
We include a distance measurement (in metres)
From the postal code center to the nearest roads, these roads are divided into four points at the top of the strength of heavy vehicles.
The traffic intensity is estimated to be the total mileage of the vehicle (
Heavy vehicles only)
Each year, for all major roads within the radius of the center of mass of the postal code address 50 m and 50 m.
We use arbitrary cutting.
Under the drive of defining \"high capacity\" in the analysis, turn off> 100000.
Estimation of road traffic noise levels using traffic noise exposure (TRANEX)model.
This uses information on road traffic flow and speed, road geography, land cover and building height to estimate the average sound pressure level (LAeq)in decibels (dB)
In different time periods.
The assessment of TRANEX by the other two UK cities showed a high correlation (r≥0. 85)
Between modeling and measurement 1-hour LAeq.
In our analysis, we focused on the average annual Lnight recorded overnight between 23: 00 and 07: 00, as this period represents the place of residence of most of our subjects.
Alternative analysis using daytime noise (LAeq16)
The same result was produced (data not shown)
Due to the high correlation with night noise (r=0. 999).
20 in order to connect the noise model with the CPRD, use the geometric center of the address location directly in each ZIP code.
The cohort was defined in 75 practices, with a total of 55 5-385 patients actively registered on 1 January 5, accounting for about 7% of the population of Greater London at that time.
From these, we chose 1397718 adults between the ages of 50-79 who have been registered for more than 1 year in a row.
From this group, 131 ÷ 869 (94%)
Successfully linked to our pollution exposure. Non-
The contact was primarily due to the fact that some practices were close to the boundaries of the study area, so the personal zip code of many of their patients was not eligible.
The patient was tracked to the earliest (1)
The date of the first diagnosis of dementia ,(2)
Date of death or cancellation of practice ,(3)
Date on which the practice stops contributing to CPRDand (4)31December2013.
We used the dementia reading code within the framework of quality and results to search for primary health care records for the first dementia diagnosis date (QOF).
21 although most read code is right
Specifically, we identified a subgroup that classified dementia as Alzheimer\'s disease or vascular dementia.
Using the death record, we also used the revised code for the tenth edition of the International Classification of Diseases to identify patients with dementia who were listed as the main cause of death (
If it is not coded on the general practitioner (GP)record)
And further classify them based on specific references to Alzheimer\'s disease or vascular dementia anywhere in the death record.
We also extracted from GP records about race, smoking, body mass index (BMI)
And alcohol consumption, using the last measurement before baseline, if only the measurement before baseline, or during the study period.
A missing category is retained where there is no information.
In addition, based on the list of conditions previously shown for the independent prediction of mortality, we extracted the co-disease information recorded at baseline. 22Patients (n=391)
Was excluded from the existing diagnosis of 1 January dementia.
In addition, we try to exclude patients (n=423)
When there is evidence on their medical records that as of 1 January 5, they live in a residential or nursing home.
We do this by reading the combination of codes, or by recording four or more patients aged 65-99 living at the same address.
We also rule out a small number of patients (n=77)
Places where IMD cannot be assigned.
This has enabled our analysis to reach 978 of the eligible patients at 30 Mbps.
Statistical analysis we used the Cox proportional risk model to investigate the association between all exposure measurements estimated in the year prior to baseline (2004)
And the subsequent incidence of dementia.
Explain clustering through practice (
Homogeneity between patients from the same practice)
We use a common weakness (
At the practical level)
, Which is equivalent to the survival data of random effects.
We are right (1)
Age, gender, ethnicity, smoking, BMI (
Alcohol consumption recorded is not independently predicted, but removed from the model), (2)IMD decile (
According to the UK National Rankings)and(3)comorbidities (
Coronary heart disease (IHD)
Heart failure, diabetes, stroke)
This is an independent predictor of dementia. For model (2)
We discussed the effect of adjusting other pollutants (
Air pollution caused by noise and vice versa).
In sensitivity analysis, we explore
Between practice and
Practice the effect by fitting different models (a non-
According to the fragile model of practice layering, as practice of fixed effect fitting, fitting practice average exposure level in addition to individual level).
For all contaminants, we fitted a continuous measurement based on the IQR to achieve a comparison between the estimated value and the five-digit number, thus describing the shape of the association.
All analyses were performed in Stata V. 13.
Patient and Public participation without patient involvement in the overall design of research issues, outcome measurements, and research.
Due to the anonymity of the patient, we were unable to disseminate the findings directly to the study participants.
Results in the research processup period (mean=6. 9 years)
A total of 2181 patients (1.
7%, the disease rate is 2.
£ 4 per 1000 per year)
First diagnosis of dementia was received during follow-upup (table 1).
Of the confirmed patients, 848 (38. 9%)
I have mentioned Alzheimer\'s disease, 634 (29. 1%)
Vascular dementia was mentioned, compared to 747. 34. 3%)received a non-
Read the code specifically.
A total of 48 patients (2. 2%)
He was diagnosed with Alzheimer\'s disease and vascular dementia.
Although the incidence of dementia in smokers was the lowest, later adjustments to age and other covariates explained this association;
However, the risk of dementia in obese subjects is the lowest (data not shown).
View this table: View inline View pop-up table 1 incidence of dementia during tracking
Table 2 shows the features of qualified subjects who simulated basic statistical estimates of air and noise pollutants in 2004.
There is a strong positive correlation between all air pollutants (r>0. 9)
There was a significant negative correlation except O 3 (r=−0. 9u2009or greater)
At the same time, it has n2 and PM2. 5. Night noise (Lnight)
There was a positive correlation with NO2 and pm2.
5, but the scale of the association is small (r=0. 3–0. 4)
All air pollutants
The noise of different patterns is also found to be related within the group (ICCs)
It is calculated through practice.
Although most noise changes are observed in practice (ICC=0. 05)
, The situation of air pollutants is just the opposite, the biggest change between practice (ICC>0. 7).
When the average concentration is calculated by traffic distance and intensity (
Online supplementary form S1)
All contaminants decreased with the increase in distance from the main roads, except for ozone, which increased with the increase in distance.
Patients living in the ZIP code closest to the main road (0–50u2009m)
The noise level at night is much higher than the farthest place (60. 4dB vs 50. 9u2009dB)
The difference in the model of air pollution concentration, especially pm2. 5.
5, a lot less.
Supplementary information [bmjopen-2018-022404-SP1. pdf]
View this table: View the inline View pop-up table 2 summary statistics of the annual air concentration in 2004 (NO2, PM2. 5, PM2. 5traffic, O3)and noise (Lnight)
Pollution stabilization 3 summarizes a series of adjusted diagnostic hours for dementia events that are associated with comparable four-point changes in exposure to different contaminants.
In n2, the strongest positive association occurred, of which a was 7.
The change of 5g/m3 produced 1 hour. 16 (95% CI 1. 05 to 1. 28)
Adjustment for IMD and other mixed factors (HR2 in table 3).
Further adjustment of related complications (HR3=1. 16)
This association was not explained.
The correlation with other measures is small (PM2. 5HR=1. 07, PM2. 5traffic HR=1. 08, LnightHR=1.
02, from roadHR = 1. 02)or negative (O3HR=0. 84).
We explore different ways to estimate the interior.
Between practice and
Practice estimation (
Online supplementary form S2).
For n2 and PM2.
Both of these estimated hours were greater than 1 hour, but the range was wide.
Noise at night (Lnight)
There is stronger evidence that
Association of practice (HR=1. 42, 95%u2009CI 1. 03 to 1. 96).
In order to investigate the shape of the association, figure 1 draws the adjusted HR (HR3 in table 3)
Classified by Level 5 of air and noise pollution, road distance and traffic intensity 50 µm.
The corresponding hours are given completely in the online Supplement Table s3.
For n2 and PM2.
The increased risk of dementia increased significantly in the top five exposed.
Patients who were exposed to no 2 in more than 41 years.
The risk of 5g/m3 increased significantly (HR=1. 40, 95%u2009CI 1. 12 to 1. 74)
Compared with the last five (10 years)
Similar findings have been made; (2)
We do have other simulated years for the next time. up (2005–2010)
These are highly relevant over time (r>0. 95)
Therefore, there is no obvious difference in the alternative analysis using them.
However, we cannot use historical factors as an explanation for our findings.
For example, recent exposure levels may be agents of other historical environmental factors associated with pollution, such as lead in gasoline, 32. Cumulative exposure is associated with a decline in cognitive ability in later years.
A new aspect of our analysis is the ability to simultaneously study the simulated effects of air and noise pollution on dementia and overcome the accepted limitations of other studies.
11 34 previous validation of the pollution models used in this study showed that roadside correlations between them were low, suggesting that independent effects of traffic pollution and road noise could be investigated.
However, a potential limitation is that our estimates of air pollution exposure may be homogeneous to some extent, if more rural geographic areas are not included, we expect to see this change across the country.
Regional contribution in London (background)PM2.
5 and 03 tend to dominate to the overall level.
However, we are able to use a decentralized model with very good resolution (20×20u2009m)
Significant changes in air pollution exposure (such as no 2) between major road and suburban background locations are estimated.
17 nevertheless, the reality is that the subtle roadside changes predicted by the model are small compared to the larger differences estimated between regions represented by GP practice (ICCs>0.
7 all air pollutants)
This shows that most simulated changes in air pollution are (practice)areas.
16 although this test has limited statistical capacity for any internal test
In this study, we found no evidence of an overall association between no 2 and pm2. 5.
Completely explained between 5
Practical differences in simulated exposure.
Finally, another limitation is incomplete information about key confusions and dependence on a domain --based measure (IMD)
Social and economic status. While mid-
Obesity in life is a risk factor for Alzheimer\'s disease, and the BMI measures we extracted near baseline showed a decreased risk of obesity in later years, a finding that has now been observed elsewhere.
35 However, further adjustments to IHD, stroke, diabetes and heart failure do not explain our findings, which will be associated with risk factors that were not measured at an early stage, including individual socio-economic conditions.
Linked to long-standing evidence of Epidemiology
The long-term concentration of the adverse effects of air pollution on health is mainly concentrated on cardiovascular diseases.
36. Describe the global burden of disease research on the effects of air pollution on a worldwide scale, taking into account a wide range of results (
IHD, stroke, lung cancer, chronic pulmonary disease)
But the results of neural degeneration were not considered.
Studies linking air pollution exposure to neurocognitive function gradually increased from observation findings in 2002, from dogs in Mexico City, 37, to larger studies evaluating cognitive decline, 38 39 and a large number of people specifically investigated the association with the diagnosis of dementia.
10 11 40-4 3a 2015 comments on the impact of long-
Duration of exposure to outdoor air pollution (15 studies)and noise (8 studies)
Regarding the cognitive and psychological functions of adults, the two exposures were shown to be related to one or several global cognitive function measures, respectively, but there was no study to consider both exposures at the same time, they stressed the need for further research.
34 The same author then performed data from the Heinz Nixdorf Recall cohort study 13 for 4086 adults using an additional calculated global cognitive score.
They concluded that air pollution and road traffic noise may have a synergistic effect on the cognitive function of adults.
Our research can consider these two measures (
Air pollution and noise at night)
While both show an independent association with dementia, any association with noise is reduced in a combined model and has marginal statistical significance.
The largest cohort study to date to investigate dementia and long-term dementia
Long-term exposure to air pollution comes from North America.
11 11 43 Chen et al 11 11 use a large Canadian population-
A basic cohort of more than 2 million adults aged 55-85 years, to determine approximately a case of dementia, has been established.
In their first analysis, 10 authors found that the risk of dementia increased as they approached a main road (adjusted HR=1. 07, 95%u2009CI 1. 06 to 1.
08, for those who live on the main traffic Road 300 m).
Urban residents, especially those living in big cities, are more closely linked.
In the subsequent analysis, they used the land
Using regression models to estimate the association between event dementia and air pollution, both were found to have a significant positive correlation.
5 and No2. the negative correlation with O3.
IQR increase of Pm2. 5 (4. 8μg/m3)
Related to 1 hour HR. 04 (95%CI 1. 03 to 1. 05)
And for NO2 (14.
2 parts per billion)
HR is estimated to be 1. 10 (95%CI 1. 08 to 1. 12).
In contrast, we estimate that the distribution of the same pollutants in Greater London is much less, and the IQRs is about 1-The size is quarterly.
Therefore, the comparison hours of similar unit changes in our study are much larger (
For example, 1g/m3 change for PM2.
Human resources will be 1.
07 to 1.
From research in Canada, this will be 1. 02 vs 1. 00).
Our estimate of pm2. 5
5 more in line with the situation found in a large study of health insurance participants in the United States
Hospitalized for dementia (HR=1. 08, 95% CI 1. 05 to 1.
11, 1g/m3 change for PM2. 5).
Chen et al speculate that a stronger association with the observed NO2 may be partly due to its better capture of fine-
Change in scale of traffic
Air pollution, while pm2. 5.
5 and ozone have a large regional component.
However, the resolution of their air pollution model is low (
1 × 1 kilometer resolution)
In our research20×20u2009m)
And may not be able to capture major emissions from road traffic.
When our model can estimate traffic
Specific ingredients of Pm2.
The effect estimate of 5,17 no2 is still high.
Some smaller studies separate Alzheimer\'s disease from dementia.
In Europe, 15-
Annual longitudinal study in northern Swedish cities found evidence that land use was positively correlated with vascular dementia, Alzheimer\'s disease, and nitrogen oxides
A regression model with a spatial resolution of 50 × 50 CUCM is used.
A comparison between participants with the highest exposure levels of residents and the lowest participants at baseline yielded similar estimates for Alzheimer\'s disease (HR=1. 38)
And vascular dementia (HR=1. 47).
There\'s a recent cohort study from Taiwan: Jungetal40 shows
Exposure to ozone and PM2.
5 was shown to increase the risk of Alzheimer\'s disease, while Chang et al found an association between dementia and no 2 and carbon monoxide.
A smaller case-control study by Wu et al linked PM10 and ozone to increased risk of Alzheimer\'s disease, and (vascular)dementia.
In our study, lower ozone was negatively correlated with the risk of dementia, mainly due to strong negative correlation with other simulated air pollutants.
45 The meaning of linking exposure to air pollution (such as no 2) to the development of dementia, especially Alzheimer\'s disease, raises many questions.
The causes of these neurodegeneration diseases are largely unclear and may be multi-factor.
While there are several ways in which toxic substances in air pollution may reach the brain, it is still speculative how and when they affect nerve degeneration. 8 30 46 traffic-
Associated Air pollution is associated with cognitive dysplasia in young children, and persistent heavy exposure in 47-year-old children may produce neuroinflammation and change the brain\'s innate immune response in early adulthood.
48 in the later years, the risk of accelerated decline in cognitive ability may involve gene-environmental interactions, such as interactions with Apo E, and 49 evidence is derived from the findings of mouse neurotoxic experiments.
50 our observations of the association between air pollution in older people living in Greater London and the diagnosis of new dementia contrast with earlier analyses of these data that fail to show between air pollution and heart and lung outcomes.
This suggests that there may be a geographical pattern specific to dementia and potential Alzheimer\'s disease, which needs to be further explored nationwide.
In the Ontario cohort, Chen et al estimated 6.
1% of them had dementia caused by air pollution.
11 in our study, all patients were theoretically transferred to the bottom of 20% of NO2 exposure, resulting in a attributable score of 7% (data not shown).
Although this will be smaller than the previous population (PAF)
The UK estimates of dementia for a range of independent risk factors, such as high blood pressure or obesity, 5 even small platelet activation factors for dementia can have an impact, environmental risk factors such as air pollution are more likely to be changed at the population level.
30 even if the effect is only to delay the progress of dementia, there will be significant progress in public health.
As the burden of global dementia may be high in the future, further epidemiology work is urgently needed to confirm and understand the latest findings between air pollution and dementia.
8 30 our results suggest that regional and urban background contaminants may be close-
Traffic pollutantsFuture large-
Large-scale research will need to rely on improved records and connections for the diagnosis of dementia across electronic systems, especially Alzheimer\'s disease, where multiple sources can improve the accuracy of diagnosis.
25 since exposure is lifelong and most cases are diagnosed later in life, it is better to also require historical data to better estimate cumulative exposure in previous decades.
In conclusion, our findings add to the growing evidence base for the link between air pollution and neurodegeneration and further research in this area should be encouraged.
Strengths and weaknesses while we are able to link contamination exposure to primary care records to obtain a diagnosis of dementia, there are concerns about variability in the diagnosis of dementia in primary care in the UK, A recent review concluded that the diagnosis of dementia on the primary care database may not accurately reflect the real prevalence. 25 Under-
Records are considered a common problem because the diagnosis is related to the stigma of many people, and GPs may not be willing to diagnose dementia unless highly determined.
26 A recent study in 23 practices in London, through a simple coding review, increased the prevalence of their QOF dementia registration by 9%.
26 pairs of insufficient accounts
At the time of baseline recording, we decided in advance to exclude all patients identified as living in a nursing home at the beginning of the follow-up
No matter what the diagnosis of dementiaDuring follow-
Up, we observed a roughly similar number of newly diagnosed dementia subtypes, as it is expected to be about two-
The third place in dementia is Alzheimer\'s disease, which suggests
The record of Alzheimer\'s disease in particular may be a problem. Under-
If related to key practice features such as deprivation, records in our study may be problematic, as dementia records prove to be lower in richer practices.
24 28 in our study, we observed a wide range of accident rates through practice (0. 2%–8. 4%)
Since most of the changes in air pollution are between practices, we cannot use unmeasured practice features as possible explanations for our findings.
In the UK, there are also known differences in the prevalence and diagnostic rates of dementia, of which London is the lowest 24 in the report, so we must also acknowledge that, the associations we observed may be specific in London and may not be expanded nationwide.
Another weakness of this study is the lack of historical data on exposure.
Most large-scale epidemiology studies for a long time
Long-term exposure to contamination makes it difficult to capture accurate images of longevity or cumulative exposure.
This may be related to Alzheimer\'s disease, and the pathogenesis of Alzheimer\'s disease may occur in many years.
30 we do not have any information about previous addresses or locations and the population of London is considered mobile and dynamic over time.
31 Therefore, we assume annual estimates for one year (2004)
Chief representative
Term exposure, based on the last known address of the patient in this practice.
We tested this in two ways :(1)
Sensitivity analysis based on patients who are constantly registered in practice for a long time (>10 years)
Similar findings have been made; (2)
We do have other simulated years for the next time. up (2005–2010)
These are highly relevant over time (r>0. 95)
Therefore, there is no obvious difference in the alternative analysis using them.
However, we cannot use historical factors as an explanation for our findings.
For example, recent exposure levels may be agents of other historical environmental factors associated with pollution, such as lead in gasoline, 32. Cumulative exposure is associated with a decline in cognitive ability in later years.
A new aspect of our analysis is the ability to simultaneously study the simulated effects of air and noise pollution on dementia and overcome the accepted limitations of other studies.
11 34 previous validation of the pollution models used in this study showed that roadside correlations between them were low, suggesting that independent effects of traffic pollution and road noise could be investigated.
However, a potential limitation is that our estimates of air pollution exposure may be homogeneous to some extent, if more rural geographic areas are not included, we expect to see this change across the country.
Regional contribution in London (background)PM2.
5 and 03 tend to dominate to the overall level.
However, we are able to use a decentralized model with very good resolution (20×20u2009m)
Significant changes in air pollution exposure (such as no 2) between major road and suburban background locations are estimated.
17 nevertheless, the reality is that the subtle roadside changes predicted by the model are small compared to the larger differences estimated between regions represented by GP practice (ICCs>0.
7 all air pollutants)
This shows that most simulated changes in air pollution are (practice)areas.
16 although this test has limited statistical capacity for any internal test
In this study, we found no evidence of an overall association between no 2 and pm2. 5.
Completely explained between 5
Practical differences in simulated exposure.
Finally, another limitation is incomplete information about key confusions and dependence on a domain --based measure (IMD)
Social and economic status. While mid-
Obesity in life is a risk factor for Alzheimer\'s disease, and the BMI measures we extracted near baseline showed a decreased risk of obesity in later years, a finding that has now been observed elsewhere.
35 However, further adjustments to IHD, stroke, diabetes and heart failure do not explain our findings, which will be associated with risk factors that were not measured at an early stage, including individual socio-economic conditions.
Linked to long-standing evidence of Epidemiology
The long-term concentration of the adverse effects of air pollution on health is mainly concentrated on cardiovascular diseases.
36. Describe the global burden of disease research on the effects of air pollution on a worldwide scale, taking into account a wide range of results (
IHD, stroke, lung cancer, chronic pulmonary disease)
But the results of neural degeneration were not considered.
Studies linking air pollution exposure to neurocognitive function gradually increased from observation findings in 2002, from dogs in Mexico City, 37, to larger studies evaluating cognitive decline, 38 39 and a large number of people specifically investigated the association with the diagnosis of dementia.
10 11 40-4 3a 2015 comments on the impact of long-
Duration of exposure to outdoor air pollution (15 studies)and noise (8 studies)
Regarding the cognitive and psychological functions of adults, the two exposures were shown to be related to one or several global cognitive function measures, respectively, but there was no study to consider both exposures at the same time, they stressed the need for further research.
34 The same author then performed data from the Heinz Nixdorf Recall cohort study 13 for 4086 adults using an additional calculated global cognitive score.
They concluded that air pollution and road traffic noise may have a synergistic effect on the cognitive function of adults.
Our research can consider these two measures (
Air pollution and noise at night)
While both show an independent association with dementia, any association with noise is reduced in a combined model and has marginal statistical significance.
The largest cohort study to date to investigate dementia and long-term dementia
Long-term exposure to air pollution comes from North America.
11 11 43 Chen et al 11 11 use a large Canadian population-
A basic cohort of more than 2 million adults aged 55-85 years, to determine approximately a case of dementia, has been established.
In their first analysis, 10 authors found that the risk of dementia increased as they approached a main road (adjusted HR=1. 07, 95%u2009CI 1. 06 to 1.
08, for those who live on the main traffic Road 300 m).
Urban residents, especially those living in big cities, are more closely linked.
In the subsequent analysis, they used the land
Using regression models to estimate the association between event dementia and air pollution, both were found to have a significant positive correlation.
5 and No2. the negative correlation with O3.
IQR increase of Pm2. 5 (4. 8μg/m3)
Related to 1 hour HR. 04 (95%CI 1. 03 to 1. 05)
And for NO2 (14.
2 parts per billion)
HR is estimated to be 1. 10 (95%CI 1. 08 to 1. 12).
In contrast, we estimate that the distribution of the same pollutants in Greater London is much less, and the IQRs is about 1-The size is quarterly.
Therefore, the comparison hours of similar unit changes in our study are much larger (
For example, 1g/m3 change for PM2.
Human resources will be 1.
07 to 1.
From research in Canada, this will be 1. 02 vs 1. 00).
Our estimate of pm2. 5
5 more in line with the situation found in a large study of health insurance participants in the United States
Hospitalized for dementia (HR=1. 08, 95% CI 1. 05 to 1.
11, 1g/m3 change for PM2. 5).
Chen et al speculate that a stronger association with the observed NO2 may be partly due to its better capture of fine-
Change in scale of traffic
Air pollution, while pm2. 5.
5 and ozone have a large regional component.
However, the resolution of their air pollution model is low (
1 × 1 kilometer resolution)
In our research20×20u2009m)
And may not be able to capture major emissions from road traffic.
When our model can estimate traffic
Specific ingredients of Pm2.
The effect estimate of 5,17 no2 is still high.
Some smaller studies separate Alzheimer\'s disease from dementia.
In Europe, 15-
Annual longitudinal study in northern Swedish cities found evidence that land use was positively correlated with vascular dementia, Alzheimer\'s disease, and nitrogen oxides
A regression model with a spatial resolution of 50 × 50 CUCM is used.
A comparison between participants with the highest exposure levels of residents and the lowest participants at baseline yielded similar estimates for Alzheimer\'s disease (HR=1. 38)
And vascular dementia (HR=1. 47).
There\'s a recent cohort study from Taiwan: Jungetal40 shows
Exposure to ozone and PM2.
5 was shown to increase the risk of Alzheimer\'s disease, while Chang et al found an association between dementia and no 2 and carbon monoxide.
A smaller case-control study by Wu et al linked PM10 and ozone to increased risk of Alzheimer\'s disease, and (vascular)dementia.
In our study, lower ozone was negatively correlated with the risk of dementia, mainly due to strong negative correlation with other simulated air pollutants.
45 The meaning of linking exposure to air pollution (such as no 2) to the development of dementia, especially Alzheimer\'s disease, raises many questions.
The causes of these neurodegeneration diseases are largely unclear and may be multi-factor.
While there are several ways in which toxic substances in air pollution may reach the brain, it is still speculative how and when they affect nerve degeneration. 8 30 46 traffic-
Associated Air pollution is associated with cognitive dysplasia in young children, and persistent heavy exposure in 47-year-old children may produce neuroinflammation and change the brain\'s innate immune response in early adulthood.
48 in the later years, the risk of accelerated decline in cognitive ability may involve gene-environmental interactions, such as interactions with Apo E, and 49 evidence is derived from the findings of mouse neurotoxic experiments.
50 our observations of the association between air pollution in older people living in Greater London and the diagnosis of new dementia contrast with earlier analyses of these data that fail to show between air pollution and heart and lung outcomes.
This suggests that there may be a geographical pattern specific to dementia and potential Alzheimer\'s disease, which needs to be further explored nationwide.
In the Ontario cohort, Chen et al estimated 6.
1% of them had dementia caused by air pollution.
11 in our study, all patients were theoretically transferred to the bottom of 20% of NO2 exposure, resulting in a attributable score of 7% (data not shown).
Although this will be smaller than the previous population (PAF)
The UK estimates of dementia for a range of independent risk factors, such as high blood pressure or obesity, 5 even small platelet activation factors for dementia can have an impact, environmental risk factors such as air pollution are more likely to be changed at the population level.
30 even if the effect is only to delay the progress of dementia, there will be significant progress in public health.
As the burden of global dementia may be high in the future, further epidemiology work is urgently needed to confirm and understand the latest findings between air pollution and dementia.
8 30 our results suggest that regional and urban background contaminants may be close-
Traffic pollutantsFuture large-
Large-scale research will need to rely on improved records and connections for the diagnosis of dementia across electronic systems, especially Alzheimer\'s disease, where multiple sources can improve the accuracy of diagnosis.
25 since exposure is lifelong and most cases are diagnosed later in life, it is better to also require historical data to better estimate cumulative exposure in previous decades.
In conclusion, our findings add to the growing evidence base for the link between air pollution and neurodegeneration and further research in this area should be encouraged. References1.
National Bureau of Statistics.
Deaths registered in England and Wales (Series DR): 2015. 2016 main-points (Accessed 06-10-17. ). 2.
Newton JN, Briggs AD, Murray CJ and others.
Health changes in the UK, by analysis of poor areas and regions in the UK, 1990-
2013: systematic analysis of the Global Burden of Disease Study 2013. Lancet 2015; 386:2257–74. doi:10. 1016/S0140-6736(15)00195-
6openurlcrossrefpmed3.
Prince M. , Bryce R. , year E.
Global prevalence of dementia: systematic review and meta-analysis.
Alzheimer\'s disease 2013; 9:63–75. doi:10. 1016/j. jalz. 2012. 11.
The openurlcrosspubmedweb of 007 Science 4.
Brookmeyer R, Johnson E, Ziegler-Graham K and others.
Predict the global burden of Alzheimer\'s disease.
Alzheimer\'s disease 2007; 3:186–91. doi:10. 1016/j. jalz. 2007. 04.
Openurlcrosspubmedweb of 381 science.
Norton, Matthews, Barnes DE and others.
Potential for primary prevention of Alzheimer\'s disease: Population Analysisbased data.
The Lancet and neurology 2014; 13:788–94. doi:10. 1016/S1474-4422(14)70136-
Xopenurlcross Web Science6.
Lakkillin LOJ, Starr JM, Shiue IJ, etc.
Environmental risk factors for dementia: a systematic review.
BMC Geriatr 2016; 16:28. doi:10. 1186/s12877-016-0342-yOpenUrl7.
Cohen AJ, Brewer M, Burnett R, et al.
Estimates and 25-
Annual trend of global burden of disease caused by environmental air pollution: Analysis of Global Burden of Disease research data in 2015. Lancet 2017; 389:1907–18. doi:10. 1016/S0140-6736(17)30505-6 openurlpubmed8.
Yang W Morton PV
Air pollution, oxidative stress and Alzheimer\'s disease.
Environmental Public Health 2012; 2012:1–9. doi:10.
1155/2012/472751openur9.
Power MC, yada SD, Yanosky JD, etc.
Exposure to air pollution is a potential contributor to cognitive function, cognitive decline, brain imaging, and dementia: a systematic review of epidemiology studies.
Neurotoxicology 201656:235–53. doi:10. 1016/j. neuro. 2016. 06. 004OpenUrl10.
Chen H, Kwong JC, Copes R, etc.
Incidence of dementia, Parkinson\'s disease and MS living near the main road: population-
Based on queue research. Lancet 2017; 389:718–26. doi:10. 1016/S0140-6736(16)32399-6OpenUrl11.
Chen H, Kwong JC, Copes R, etc.
Exposure to ambient air pollution and dementia incidence: population-
Based on queue research.
Environment 2017; 108:271–7. doi:10. 1016/j. envint. 2017. 08. 020OpenUrl12. ↵Calderón-
Gareth denas of VillarrealRíos R.
Living in the vicinity of heavy traffic, air pollution and dementia. Lancet 2017; 389:675–7. doi:10. 1016/S0140-6736(16)32596-XOpenUrl13.
Jobtzivian L, Jokisch M, progress of work, etc.
Long Association
Long-term exposure of air pollution and road traffic noise with cognitive function
Analysis of effect metric modification.
Environment 2017; 103:30–8. doi:10. 1016/j. envint. 2017. 03. 018OpenUrl14.
Herarrett E, Galach AM, Bhaskaran K, etc.
Introduction to Data Resources: Data link for Clinical Practice Research (CPRD).
On 2015; 44:827–36. doi:10.
Openurlcrosspubmed15 1093/ije/dyv098.
Community and local government departments.
The deprivation index was 2010.
London: Department of Community and Local Government, 2011. (
Visit December 28, 2015). 16.
Carey carey IM, Anderson HR, Atkinson RW, etc.
Traffic pollution and incidence of heart and lung outcomes in the London adult cohort.
Occup Environ Med 2016; 73:849–56. doi:10. 1136/oemed-2015-
103531 OpenUrlAbstract/free full text text17.
Kitbeevers SD, Kitwiroon N, Williams ML, etc.
The air pollution diffusion model predicted by human exposure in London.
J. Environmental Epidemiology of Sci World Expo 2013; 23:647–53. doi:10. 1038/jes. 2013.
6openurlcrossrefubmed18.
Dajnak D. Beevers SD
Supplementary documents for traffic projects.
Air pollution model
Description, evaluation and output of KCLurban model. 2015 (
Visit February 14, 2018). 19.
Gulliver J, Morley D, Vienna D, etc.
An open development
Road traffic noise source model for exposure assessment.
Modeling and software for environment 201574:183–93. doi:10. 1016/j. envsoft. 2014. 12. 022OpenUrl20.
Hanfecht D, haneshal, Morley D, etc.
Temporal and spatial association of road traffic noise and air pollution in London: Significance of epidemiology studies.
Environment 2016; 88:235–42. doi:10. 1016/j. envint. 2015. 12. 001OpenUrl21. ↵NHS Digital.
Quality and results framework: NHS Digital, 2016. (
Visit December 06, 2017). 22.
Alimcarey IM, Shah SM, Harris T, etc.
A new simple primary care incidence score predicts mortality better than the Charlson index and better explains the change in practice.
J. Clin Epidemiol 2013; 66:436–44. doi:10. 1016/j. jclinepi. 2012. 10.
012openurlcross refpmed23. ↵Austin PC.
Multi-level Survival Analysis Tutorial: methods, models, and applications.
The Int Stat Rev 2017; 85:185–203. doi:10. 1111/insr. 12214OpenUrl24.
If, Lord PA, Farragher TM.
Changes in the diagnosis of dementia in the UK and their association with general practice features.
Prim Health Care Res Dev 2017; 18:235–41. doi:10.
1017/s1463423617. 07xopenurl25.
Sibsibbett RA, Russ TC, Deary IJ, etc.
Using available data from UK longitudinal and cohort studies to determine dementia: a systematic review of the method.
BMC Psychiatry 201717:17. doi:10. 1186/s12888-017-1401-4OpenUrl26.
Russell P, bannaji, Watt J, et al.
Improving the identification of dementia patients in primary care: an assessment of the impact of the code guide for primary care dementia on the identification prevalence. BMJ Open 2013; 3:e004023. doi:10. 1136/bmjopen-2013-00402327.
Alzheimer\'s Association
Dementia in the UK: update.
2nd edn: Alzheimer\'s Association, 2014. (
Visit April 10, 2018). 28.
Connconnolly A, Gaehl E, Martin H, etc.
Underestimation of dementia in primary care: Changes in observed prevalence and comparison with expected prevalence.
2011 of the elderly health; 15:978–84. doi:10. 1080/13607863. 2011.
Openurlcrosspubmedweb of 596805 Science 29.
Ministry of Health.
A national report on dementia care and support in the UK. 2013 (
Visit April 10, 2018). 30. ↵Weuve J.
Invited comments: how exposure to air pollution may affect the risk of dementia, and what is the comment of Epidemiology on this.
Am J Epidemiol 2014; 180:367–71. doi:10.
English Channel 1093 Al Jazeera/kwu153openurlcrossrefpmed31.
Perkins J. Bell
Imagine the city of the future: London 2062.
London: Publishing House everywhere. 32.
MAS laidlaw MAS, Poropat AE, Ball A, etc.
The incidence of lead and dementia in contact with gasoline increased. Lancet 2017; 389:2371–2. doi:10. 1016/S0140-6736(17)31466-6OpenUrl33.
Bakbakulski KM, Rozek LS, Dolinoy DC, etc.
Environmental exposure to Alzheimer\'s disease and lead: epidemiology evidence and potential role of epigenetic.
Alzheimer\'s disease Res 2012; 9:563–73. doi:10.
Openurlcrosspubmedweb of 2174/156720512800617991 Science 34.