Section Analysis
Introduction
Key Aspects
- Alzheimer's Disease Mortality Trends: Alzheimer's disease mortality has doubled in the last thirty years and is projected to rise with an aging population. This highlights the urgency for effective treatments or preventative measures, which are currently lacking despite extensive research efforts.
- Hippocampus and Spatial Processing: The hippocampus, a brain region crucial for spatial memory and navigation, is significantly affected in the early stages of Alzheimer's disease. Notably, studies have shown that London taxi drivers exhibit enhanced hippocampal function, potentially due to their extensive spatial processing demands.
- Research Hypothesis: The study hypothesizes that occupations requiring frequent spatial and navigational processing, such as taxi and ambulance driving, may be associated with a reduced risk of Alzheimer's disease mortality. This is based on the observed hippocampal changes in London taxi drivers and the role of the hippocampus in Alzheimer's disease pathology.
- Study Objective: The researchers aim to investigate this hypothesis using population-based mortality data from the United States. This data includes newly linked occupational information, allowing for a comparison of Alzheimer's disease mortality across different professions, particularly focusing on taxi and ambulance drivers.
- Methodological Approach: The study utilizes data from the National Vital Statistics System, a comprehensive registry of deaths in the US from 2020 to 2022. Death certificates provide information on the underlying cause of death, including Alzheimer's disease, and sociodemographic details of the deceased. Importantly, these certificates also include the individual's usual occupation, which is coded according to standardized US Census Bureau classifications.
Strengths
-
Clear Research Question
The introduction clearly establishes the research question by linking the increasing prevalence of Alzheimer's disease, the role of the hippocampus, and the potential protective effect of spatial processing in certain occupations.
"occupations that demand frequent spatial processing, such as taxi driving, may be associated with decreased Alzheimer’s disease mortality." (Page 1) -
Well-Defined Hypothesis
The hypothesis is clearly stated and directly relates to the research question, proposing that taxi and ambulance driving might be associated with a reduced burden of Alzheimer's disease mortality.
"occupations such as taxi driving and ambulance driving (but not emergency medical technicians (or EMTs)), which demand frequent real time spatial and navigational processing, might be associated with a reduced burden of Alzheimer’s disease mortality compared with other occupations." (Page 1) -
Logical Flow and Context
The introduction effectively builds a logical flow, starting with the problem of Alzheimer's disease, introducing the relevant background on the hippocampus, and then presenting the research question and hypothesis within this context.
"Despite decades of research, an important gap remains in the development of definitive treatment or prevention strategies." (Page 1)
Suggestions for Improvement
-
Clarify Distinction Between EMTs and Ambulance Drivers
This medium-impact improvement would enhance the clarity of the research focus. The Introduction section is the appropriate place to define the specific occupational groups being studied and differentiate them from related but distinct roles. This is crucial for understanding the study's scope and rationale. Elaborating on this distinction would strengthen the paper by providing a more precise definition of the target population and highlighting the unique aspects of ambulance driving that are hypothesized to be relevant to Alzheimer's disease risk. This clarification would also help readers understand why EMTs are excluded from the primary analysis. Ultimately, providing a clearer explanation of the differences between ambulance drivers and EMTs would improve the study's focus and help readers better understand the specific occupational characteristics being investigated.
"occupations such as taxi driving and ambulance driving (but not emergency medical technicians (or EMTs))" (Page 1)Implementation: 1. Briefly describe the typical duties of ambulance drivers, emphasizing their navigational responsibilities. 2. Contrast these duties with those of EMTs, highlighting the differences in their roles and training. 3. Explain why the study focuses on ambulance drivers and not EMTs, linking this decision to the hypothesis about spatial processing.
-
Expand on the Significance of US Data
This medium-impact improvement would strengthen the justification for using US mortality data. The Introduction is the ideal place to explain the advantages and relevance of the chosen dataset, especially since it's a novel aspect of this study. Expanding on this would enhance the paper by highlighting the unique contribution of this research and demonstrating the value of leveraging newly available occupational data in the US. This would also provide context for readers unfamiliar with the US National Vital Statistics System. Ultimately, providing a more detailed explanation of the significance of using US mortality data would strengthen the study's rationale and highlight its innovative approach.
"To investigate this possibility, we leveraged population based US mortality data, which include newly linked information on the occupation of people who had died" (Page 1)Implementation: 1. Briefly describe the National Vital Statistics System and its recent inclusion of occupational data. 2. Explain the advantages of using this large, population-based dataset for studying occupational mortality patterns. 3. Highlight the novelty of this approach and its potential to provide new insights into the relationship between occupation and Alzheimer's disease.
-
Briefly Mention Potential Mechanisms
This low-impact improvement would provide a more complete overview of the research topic. While the Introduction should primarily focus on the research question and hypothesis, briefly mentioning potential mechanisms linking spatial processing to Alzheimer's disease risk would enhance the reader's understanding. This addition would strengthen the paper by providing a more comprehensive introduction to the topic and hinting at the potential biological underpinnings of the hypothesized relationship. This would also help readers appreciate the broader implications of the study. Briefly alluding to potential mechanisms, such as neuroplasticity or cognitive reserve, would provide a more complete picture of the research area without delving into excessive detail in the Introduction.
"The hippocampus is the brain region involved in both the creation of cognitive spatial maps and the development of Alzheimer’s disease" (Page 1)Implementation: 1. Add a sentence or two suggesting potential mechanisms, such as the impact of spatial processing on hippocampal structure and function. 2. Briefly mention concepts like neuroplasticity or cognitive reserve, linking them to the hypothesis. 3. Avoid detailed explanations, as these are better suited for the Discussion section.
Methods
Key Aspects
- Data Source: National Vital Statistics System: The study utilizes mortality data from the National Vital Statistics System, a comprehensive, population-based registry of all deaths in the United States from January 1, 2020, to December 31, 2022. This system collects data from death certificates, which include the underlying cause of death, coded according to the International Classification of Diseases, 10th revision (ICD-10), and sociodemographic information such as age, sex, race, ethnicity, and educational attainment.
- Occupational Data Coding: Death certificates include a field for the decedent's usual occupation, defined as the occupation in which they spent most of their working life. Starting in 2020, these occupational narratives were coded to standardized 2010 US Census Bureau occupation codes in collaboration with the National Institute for Occupational Safety and Health. This standardized coding allows for systematic analysis of mortality patterns across different occupations.
- Study Population and Occupational Groups: The study population comprised deceased adults aged 18 years and older with reported occupational information, covering approximately 98% of the US population during the study period. The researchers analyzed 443 occupational groups, with a specific focus on taxi drivers and ambulance drivers due to their extensive navigational demands. Other transportation-related occupations, such as bus drivers, aircraft pilots, and ship captains, were included as a comparison group.
- Outcome Measure: Alzheimer's Disease Mortality: The primary outcome was the percentage of deaths for each occupation where Alzheimer's disease (ICD-10 code G30) was listed as the underlying cause of death. This measure was used to assess the relative burden of Alzheimer's disease mortality across different occupations.
- Statistical Analysis and Adjustment: The researchers calculated the percentage of deaths due to Alzheimer's disease and the mean age at death for each occupation. They used multivariable logistic regression to estimate risk-adjusted percentages of deaths from Alzheimer's disease, controlling for age at death, sex, race, ethnicity, and educational attainment. This adjustment was crucial to account for potential confounding factors and isolate the association between occupation and Alzheimer's disease mortality.
Strengths
-
Comprehensive Data Source
The use of the National Vital Statistics System provides a large, population-based dataset, enhancing the generalizability of the findings.
"Mortality data were obtained from the National Vital Statistics System, a population based registry of all deaths in the US" (Page 1) -
Standardized Occupational Coding
The coding of occupations to standardized US Census Bureau categories allows for systematic comparison across different professions.
"occupation narratives from death certificates were coded to standardized 2010 US Census Bureau occupation codes" (Page 1) -
Appropriate Outcome Measure
Focusing on Alzheimer's disease as the underlying cause of death provides a clear and relevant outcome for investigating the research question.
"The primary outcome was the percentage of deaths for each occupation with underlying cause of death from Alzheimer's disease" (Page 2) -
Adjustment for Confounders
The use of multivariable logistic regression to adjust for demographic factors strengthens the internal validity of the study by controlling for potential confounders.
"adjusting for age at death (indicator variables for each year of age), sex (male or female), race and ethnic group" (Page 2)
Suggestions for Improvement
-
Clarify Occupational Data Collection
This medium-impact improvement would enhance the transparency and reproducibility of the study. The Methods section is the appropriate place to detail the process of occupational data collection and coding, as it directly impacts the validity of the study's findings. Providing more information on this process would strengthen the paper by allowing readers to better understand the potential limitations and biases associated with the occupational data. This would also enable other researchers to replicate the study more effectively. Ultimately, a more detailed description of the occupational data collection and coding process would improve the study's methodological rigor and enhance its contribution to the field.
"Starting in 2020, occupation narratives from death certificates were coded to standardized 2010 US Census Bureau occupation codes" (Page 1)Implementation: 1. Describe how the "usual occupation" was determined and recorded on death certificates. 2. Explain the role of the funeral director and the decedent's informant in providing occupational information. 3. Provide details about the training and guidelines given to individuals responsible for coding occupations. 4. Discuss any quality control measures implemented to ensure the accuracy and consistency of occupational coding.
-
Address Potential for Misclassification
This medium-impact improvement would enhance the study's robustness by acknowledging and addressing a potential source of bias. The Methods section is the appropriate place to discuss potential limitations in data classification, especially since it directly relates to the study's main exposure variable. Addressing this limitation would strengthen the paper by demonstrating the authors' awareness of potential biases and their efforts to mitigate them. This would also provide a more nuanced interpretation of the findings and highlight areas for future research. Ultimately, acknowledging and discussing the potential for misclassification of occupations would improve the study's transparency and contribute to a more accurate understanding of the relationship between occupation and Alzheimer's disease mortality.
"reported usual occupation (the occupation in which the decedent spent most of their working life)" (Page 1)Implementation: 1. Acknowledge that using "usual occupation" from death certificates may not perfectly capture an individual's lifetime occupational exposure. 2. Discuss the possibility that individuals may have changed occupations throughout their working life. 3. Explain how this potential misclassification could bias the results, and in what direction. 4. Suggest ways to address this limitation in future studies, such as using more detailed occupational histories.
-
Justify Exclusion Criteria
This low-impact improvement would provide greater clarity regarding the study population. The Methods section should clearly explain the rationale behind excluding specific groups, ensuring transparency in the selection process. This would strengthen the paper by providing a more complete understanding of the study population and the reasons for its composition. This detailed justification would also allow readers to better assess the generalizability of the findings. Providing a clear justification for each exclusion criterion would enhance the methodological rigor of the study.
"We also excluded students attending high school or college (occupation code: 9070), and occupations with fewer than 250 overall deaths per year" (Page 2)Implementation: 1. Explain why individuals with unknown occupational data were excluded, including the potential impact of missing data on the analysis. 2. Justify the exclusion of students, linking it to the focus on occupational exposure during working life. 3. Elaborate on the rationale for excluding occupations with fewer than 250 deaths per year, explaining how this threshold was determined and its potential impact on the results.
Results
Key Aspects
- Study Population Demographics: The study encompassed 8,972,221 individuals with recorded occupational data. The demographics reveal that among the selected navigational occupations, ambulance drivers had a mean age at death of 64.2 years (standard deviation 14.7), while taxi drivers had a mean age of 67.8 years (standard deviation 14.5). These groups predominantly comprised men, except for bus drivers, where the gender distribution was more balanced. Educational attainment varied, with most individuals in navigational occupations having a high school education or less, except for aircraft pilots, who generally had higher educational levels.
- Alzheimer's Disease Mortality Rates: The unadjusted percentage of deaths attributed to Alzheimer's disease was 1.03% among taxi drivers and 0.74% among ambulance drivers. In contrast, the rates were significantly higher for bus drivers (3.11%), aircraft pilots (4.57%), and ship captains (2.79%). Notably, the overall percentage of deaths due to Alzheimer's disease in the entire study population was 3.88%.
- Risk-Adjusted Analysis: After adjusting for age at death, sex, race, ethnicity, and educational attainment using multivariable logistic regression, ambulance drivers and taxi drivers still exhibited the lowest adjusted percentages of deaths from Alzheimer's disease, at 0.91% (95% CI 0.35% to 1.48%) and 1.03% (95% CI 0.87% to 1.18%), respectively. These findings suggest a potential association between occupations requiring frequent spatial navigation and a lower risk of Alzheimer's disease mortality.
- Comparison with Other Transportation Occupations: In contrast to taxi and ambulance drivers, other transportation-related occupations with less reliance on real-time navigation, such as bus drivers, aircraft pilots, and ship captains, did not show a similar pattern of reduced Alzheimer's disease mortality. Bus drivers had an adjusted Alzheimer's disease mortality rate of 1.65%, pilots 2.34%, and ship captains 2.12%, indicating that the observed trend might be specific to occupations with high navigational demands.
- Sensitivity Analyses: The researchers conducted four sensitivity analyses to assess the robustness of their findings. These included restricting the analysis to individuals who died after age 60, broadening the definition of Alzheimer's disease mortality to include contributing causes, evaluating other transportation occupations, and examining other forms of dementia. The results remained consistent across these analyses, supporting the main findings.
Strengths
-
Clear Presentation of Descriptive Statistics
The Results section effectively presents descriptive statistics, including demographic characteristics and unadjusted Alzheimer's disease mortality rates for each occupational group, providing a clear overview of the study population and initial findings.
"Of all people studied, 3.88% (348328/8972221) were identified as having an underlying cause of death from Alzheimer's disease." (Page 3) -
Well-Organized Tables
Tables 1 and 2 are well-organized and effectively summarize key data, including demographic characteristics, unadjusted and adjusted Alzheimer's disease mortality rates, and odds ratios, facilitating easy comparison across occupational groups.
"Table 2 | Deaths attributed to Alzheimer's disease by occupation" (Page 2) -
Appropriate Use of Visualizations
Figure 1 effectively visualizes the relationship between Alzheimer's disease mortality and age at death, as well as the risk-adjusted percentages and odds ratios, enhancing the interpretability of the results.
"Fig 1 | Mortality from Alzheimer's disease among ambulance drivers, taxi drivers, and other occupations." (Page 3) -
Comprehensive Sensitivity Analyses
The inclusion of four sensitivity analyses demonstrates the robustness of the findings and addresses potential concerns about the study's limitations.
"We performed four sensitivity analyses to assess the robustness of our comparisons." (Page 3)
Suggestions for Improvement
-
Provide Context for Odds Ratios
This medium-impact improvement would enhance the reader's understanding of the magnitude and significance of the observed associations. While the Results section presents adjusted odds ratios, providing more context for interpreting these values, especially for readers less familiar with logistic regression, would be beneficial. This is important in this section because it is where the primary statistical findings are presented. Adding this context would strengthen the paper by making the results more accessible and impactful. It would also help readers appreciate the practical significance of the findings in relation to the reference group (chief executives). Ultimately, providing a clearer explanation of the odds ratios would improve the interpretability of the results and enhance their contribution to the field.
"Adjusted odds ratios were calculated using chief executives (US Census Bureau occupation code 0010) as an arbitrary reference group." (Page 2)Implementation: 1. Briefly explain the interpretation of odds ratios, particularly values less than 1, which indicate a lower risk compared to the reference group. 2. Provide a statement about the magnitude of the observed effect sizes, highlighting the practical significance of the lower odds ratios for taxi and ambulance drivers. 3. Consider adding a sentence explaining that an odds ratio of 0.50, for example, suggests that the odds of death from Alzheimer's disease are 50% lower in that group compared to the reference group.
-
Clarify the Choice of Reference Group
This low-impact improvement would provide greater transparency regarding the methodological choices made in the study. While the use of chief executives as a reference group is mentioned, briefly justifying this choice in the Results section would be helpful. This is relevant to this section as it directly affects how the results are interpreted. This clarification would strengthen the paper by providing a more complete understanding of the rationale behind the chosen reference group. It would also help readers assess the appropriateness of this choice for comparing occupational risks. Briefly explaining why chief executives were chosen as the reference group would enhance the methodological clarity of the study.
"Adjusted odds ratios were calculated using chief executives (US Census Bureau occupation code 0010) as an arbitrary reference group." (Page 2)Implementation: 1. Add a sentence explaining why chief executives were chosen as the reference group. 2. Briefly mention factors considered when selecting this group, such as their likely different exposure to navigational demands compared to taxi and ambulance drivers. 3. State that this choice allows for a comparison of Alzheimer's disease mortality between groups with potentially contrasting occupational characteristics.
-
Report Confidence Intervals for Unadjusted Percentages
This low-impact improvement would provide a more complete picture of the precision of the unadjusted estimates. While the Results section reports unadjusted percentages of deaths from Alzheimer's disease, including confidence intervals would enhance the interpretation of these values. This is relevant to this section because it presents the initial, unadjusted findings. Adding confidence intervals would strengthen the paper by allowing readers to assess the variability around the unadjusted estimates. It would also provide a more nuanced understanding of the differences between occupational groups before adjustment. Including confidence intervals for the unadjusted percentages would improve the statistical rigor of the results presentation.
"The unadjusted percentage of deaths from Alzheimer's disease was 1.03% (171/16658) among taxi drivers" (Page 3)Implementation: 1. Calculate and report the 95% confidence intervals for the unadjusted percentages of deaths from Alzheimer's disease in Table 2. 2. Briefly explain that these confidence intervals provide a range within which the true population value is likely to lie. 3. Consider adding a footnote to Table 2 explaining how these confidence intervals were calculated.
Non-Text Elements
Table 1 | Characteristics of decedents with reported navigational occupation
- Purpose of the table: This table shows the demographic and socioeconomic characteristics of people who have died and whose occupations were related to navigation, such as taxi drivers, ambulance drivers, and pilots. It breaks down the data by different job types to compare these groups. Essentially, it sets the stage to understand if the people in these jobs are similar or different in terms of age, sex, race, and education.
- Structure of the table: The table is organized with different navigational occupations listed in columns, such as "Ambulance drivers," "Taxi drivers and chauffeurs," "Aircraft pilots," and so on. Each column represents a specific occupational group. There is also a column for "All other occupations" which serves as a comparison group. The rows of the table list various characteristics like age, sex, race and ethnicity, and educational attainment. The values in each cell represent either the average value (for age) or the number and percentage of individuals (for categorical variables like sex or education level) within each occupational group that share a particular characteristic.
- Definition of key variables: The table includes several key variables. "Age, mean, years (SD)" refers to the average age at death for each group, with SD standing for standard deviation, which is a measure of how spread out the ages are around the average. "Sex, no. (%)" indicates the number and percentage of males and females in each group. "Race and ethnicity, no. (%)" breaks down each group by racial and ethnic categories, such as Asian or Pacific Islander, Black, Hispanic, and others, showing the number and percentage in each category. "Educational attainment, no. (%)" shows the number and percentage of individuals in each group who have achieved different levels of education, from less than high school to a graduate degree. It also includes a category "Unknown" for cases where the educational level is not reported.
- Technical terms specific to this table: The term "decedents" is used to refer to the people who have died, and their data is being analyzed. "Navigational occupation" refers to jobs that involve navigating or directing a vehicle as a primary part of the work, like driving a taxi or piloting an aircraft. The phrase "reported navigational occupation" implies that the occupation was recorded on official documents, likely death certificates. "GED" stands for General Educational Development, which is a set of tests that, when passed, certify that the test taker has met high-school level academic skills.
- Data Representation: The table appropriately represents demographic and socioeconomic data for different occupational groups. However, the validity of comparisons across groups depends on the accuracy and completeness of the underlying data from the National Vital Statistics System. Potential biases in reporting or coding occupations could affect the validity of these comparisons.
- Variable Selection: The selected variables (age, sex, race and ethnicity, educational attainment) are standard demographic descriptors and are relevant for controlling for potential confounding factors in the subsequent analyses. The inclusion of 'Unknown' categories for some variables is a good practice as it acknowledges missing data, but the impact of these missing data on the overall analysis should be considered.
- Statistical Descriptors: The use of mean and standard deviation for age is appropriate for summarizing continuous data. The use of counts and percentages for categorical variables is standard and suitable for this type of data. The table provides a clear descriptive summary of each group, which is crucial for understanding the composition of each occupational group before proceeding to further statistical analysis.
- Clarity of Column Headers: The column headers are clear and self-explanatory, such as 'Ambulance drivers,' 'Taxi drivers and chauffeurs,' etc. The use of 'n=' to indicate the sample size for each group is standard and well-understood in scientific literature.
- Readability of Row Labels: The row labels for demographic characteristics are straightforward. However, the term 'decedents' might not be immediately clear to all readers, although it is technically correct.
- Use of Abbreviations: Abbreviations like 'SD' for standard deviation and 'GED' for general educational development are used. While 'SD' is commonly understood, 'GED' is explained in a footnote, which is helpful. However, the footnote is placed far below the table which could hinder immediate understanding.
- Overall Table Layout: The table is well-organized, with a clear distinction between different occupational groups and demographic categories. However, the table is quite wide, which might make it difficult to view on smaller screens or when printed.
- Footnote Placement and Clarity: The footnote explaining 'GED' is appropriate but could be more effective if placed closer to the relevant row or if 'GED' was spelled out in the row label itself. Additionally, the table could benefit from a brief introductory sentence or note explaining the source of the data (National Vital Statistics System) and the years covered.
Table 2 | Deaths attributed to Alzheimer's disease by occupation
- Purpose and content of the table: This table presents data on the number of deaths attributed to Alzheimer's disease across different occupations. It aims to compare the prevalence of Alzheimer's-related deaths among various professions, particularly those involving navigation. The table lists the total number of deaths within each occupation, the number of those deaths caused by Alzheimer's, the raw percentage of Alzheimer's deaths, and an adjusted percentage that accounts for factors like age, sex, race, and education. It also includes an adjusted odds ratio, which is a statistical measure that quantifies the strength of association between Alzheimer's disease and each occupation compared to a reference group (chief executives in this case).
- Explanation of key columns: "Deaths, no." represents the total number of deceased individuals within each listed occupation. "Alzheimer's deaths, no." is the count of those deaths specifically attributed to Alzheimer's disease. "Alzheimer's deaths, %" is a simple calculation of the percentage of deaths due to Alzheimer's within each group. "Alzheimer's deaths, adjusted % (95% CI)" is a more complex measure. The 'adjusted %' accounts for differences in age, sex, race, ethnic group, and education levels across the occupations, providing a fairer comparison. The "95% CI" stands for the 95% confidence interval, which is a range of values that likely contains the true adjusted percentage 95% of the time if we were to repeat the study many times. "Adjusted odds ratio (95% CI)" compares the odds of dying from Alzheimer's in each occupation to the odds in a reference group (chief executives), after adjusting for demographic factors. An odds ratio of 1 means no difference in odds, less than 1 means lower odds, and greater than 1 means higher odds. The 95% CI for the odds ratio similarly gives a range within which the true odds ratio likely falls.
- Occupations compared: The table focuses on several occupations: ambulance drivers, taxi drivers and chauffeurs, bus drivers, chief executives, ship captains, and aircraft pilots. These occupations are chosen to explore the hypothesis that jobs requiring frequent spatial and navigational processing, like those of taxi and ambulance drivers, might be associated with a lower risk of Alzheimer's disease. Chief executives are used as a comparison group because they represent a different type of occupation, presumably with different cognitive demands. The adjusted odds ratios are calculated using chief executives as a reference, which allows for a comparison of each occupation's Alzheimer's mortality risk relative to this group.
- Appropriateness of Adjustments: The adjustment for age, sex, race, ethnic group, and educational attainment is crucial for a valid comparison of Alzheimer's disease mortality across occupations. These factors are known to be associated with both Alzheimer's risk and occupational choice. The use of logistic regression to perform these adjustments is a standard and appropriate statistical method.
- Choice of Reference Group: Using chief executives as the reference group for calculating adjusted odds ratios is an arbitrary but acceptable choice. However, it's important to note that the interpretation of the odds ratios depends on this choice. A different reference group could yield different odds ratios, although the overall pattern of results would likely be similar.
- Statistical Significance: The inclusion of p-values helps to assess the statistical significance of the observed differences in Alzheimer's mortality. However, the reliance on a strict p-value threshold (e.g., p < 0.05) for determining significance can be misleading, especially when multiple comparisons are made. The 95% confidence intervals provide a more informative measure of the precision of the estimates.
- Limitations of Observational Data: The study is based on observational data, which limits the ability to draw causal inferences. While the adjustments attempt to control for confounding factors, there may be unmeasured confounders that influence both occupational choice and Alzheimer's risk. The observed associations, therefore, do not necessarily imply a causal relationship.
- Clarity of Column Headers: The column headers are mostly clear, but "Alzheimer's deaths, adjusted % (95% CI)" could be more descriptive. For instance, it could be expanded to "Adjusted Percentage of Deaths from Alzheimer's Disease (95% Confidence Interval)," although this would make the header quite long.
- Use of Abbreviations and Technical Terms: The table uses abbreviations like "no." for number and "CI" for confidence interval, which are standard in scientific tables. However, the meaning of "adjusted" might not be immediately obvious to all readers. A brief explanation in the table footnote would be helpful.
- Presentation of Odds Ratios: The presentation of adjusted odds ratios with 95% confidence intervals is appropriate and allows for a clear comparison of Alzheimer's mortality risk across occupations. The use of chief executives as a reference group is explicitly stated in the table footnote, which is good practice.
- Overall Table Structure: The table is well-organized and easy to follow. The occupations are listed in a logical order, and the different measures of Alzheimer's mortality are clearly presented. However, the table could benefit from a more descriptive title that highlights the main finding or the purpose of the comparison.
- Accessibility to Non-Expert Readers: While the table is technically sound, it might be challenging for non-expert readers to fully understand the meaning of adjusted percentages and odds ratios. A brief explanation of these concepts in the table footnote or in the main text would improve the table's accessibility.
Fig 1 | Mortality from Alzheimer's disease among ambulance drivers, taxi...
Fig 1 | Mortality from Alzheimer's disease among ambulance drivers, taxi drivers, and other occupations. Risk adjusted percentages and mortality odds ratios were adjusted for age at death, sex, race, ethnic group, and educational attainment using logistic regression. In the bottom graph, a logarithmic scale was used for the y axis to allow for accurate visual comparison of effect sizes between occupations, as the logarithmic scale equalizes the distances between ratios and their reciprocals. Adjusted odds ratios were calculated using chief executives (US Census Bureau occupation code 0010) as an arbitrary reference group
- Overall purpose of the figure: This figure visually compares the death rates from Alzheimer's disease among different occupations, focusing on ambulance drivers and taxi drivers. It uses graphs to show how these occupations relate to Alzheimer's mortality after taking into account various factors like age, sex, race, ethnicity, and education level. The main idea is to see if people in jobs that involve a lot of navigation, like taxi and ambulance drivers, have different rates of Alzheimer's deaths compared to other jobs.
- Description of the top graph: The top graph plots the percentage of deaths due to Alzheimer's disease on the vertical axis (y-axis) against the average age at death for each occupation on the horizontal axis (x-axis). Each point on the graph represents a different occupation. The percentage of Alzheimer's deaths has been 'risk-adjusted,' meaning that the researchers have used a statistical method called logistic regression to account for differences in age, sex, race, ethnic group, and education across the occupations. This adjustment allows for a fairer comparison of Alzheimer's mortality rates. Occupations are color-coded: blue dots represent ambulance drivers, orange dots represent taxi drivers, and gray dots represent all other occupations. This graph helps to visualize whether occupations with similar average ages at death have different rates of Alzheimer's deaths.
- Description of the bottom graph: The bottom graph shows the 'risk-adjusted Alzheimer's mortality odds ratio' on the y-axis, again plotted against different occupations. The odds ratio is a way of comparing the odds of dying from Alzheimer's in one group to the odds in another group. Here, the odds ratios are adjusted for the same factors as in the top graph (age, sex, race, ethnicity, and education). The y-axis uses a logarithmic scale, which means that equal distances on the axis represent equal ratios, making it easier to compare the relative differences in odds ratios between occupations. The graph also uses color-coding to differentiate between ambulance drivers (blue), taxi drivers (orange), and other occupations (gray). The adjusted odds ratios are calculated relative to a reference group, which in this case is chief executives. An odds ratio of 1 would mean that the odds of dying from Alzheimer's are the same as for chief executives, while an odds ratio less than 1 means the odds are lower, and greater than 1 means the odds are higher.
- Explanation of logistic regression and its use here: Logistic regression is a statistical method used to analyze the relationship between a dependent variable (in this case, whether someone died from Alzheimer's disease) and one or more independent variables (like age, sex, occupation, etc.). It's particularly useful when the dependent variable is binary, meaning it can only take two values (e.g., yes or no, died from Alzheimer's or not). In this study, logistic regression is used to adjust the percentages and odds ratios for the effects of age, sex, race, ethnic group, and education. This is important because these factors might influence both the likelihood of dying from Alzheimer's and the type of occupation someone has. By adjusting for these factors, the researchers can get a clearer picture of the relationship between occupation and Alzheimer's mortality.
- Explanation of logarithmic scale in the bottom graph: A logarithmic scale is a way of displaying numerical data over a very wide range of values in a compact way. In this graph, a logarithmic scale is used on the y-axis to represent the odds ratios. This means that instead of each unit on the y-axis representing an increase of 1, each unit represents a multiplication by a constant factor (e.g., each unit could represent a doubling or tripling of the odds ratio). The reason for using a logarithmic scale here is that it makes it easier to visually compare ratios that are much smaller than 1 with ratios that are much larger than 1. If a regular linear scale were used, the differences between very small odds ratios would be hard to see. The caption explains that the logarithmic scale 'equalizes the distances between ratios and their reciprocals,' which means that, for example, an odds ratio of 2 (double the odds) is the same distance from 1 as an odds ratio of 0.5 (half the odds) on the logarithmic scale.
- Appropriateness of Risk Adjustment: Adjusting for age, sex, race, ethnic group, and educational attainment is crucial for a valid comparison of Alzheimer's mortality across occupations. These factors are known confounders, as they are associated with both Alzheimer's risk and occupational choice. The use of logistic regression for this adjustment is a standard and appropriate statistical method.
- Use of Logarithmic Scale: The use of a logarithmic scale for the odds ratio is appropriate and justified. It allows for a more accurate visual comparison of effect sizes, especially when dealing with ratios that span several orders of magnitude. The explanation in the caption is accurate and helpful.
- Choice of Reference Group for Odds Ratios: Using chief executives as the reference group is an arbitrary but acceptable choice. It's important to keep in mind that the interpretation of the odds ratios depends on this choice. However, the overall pattern of results would likely be similar regardless of the reference group used.
- Limitations of Visualizations: While the graphs are informative, they do not provide precise numerical estimates of the adjusted percentages or odds ratios. Readers need to refer to Table 2 for these values. Additionally, the graphs do not show the confidence intervals around the estimates, which are important for assessing the statistical significance of the observed differences.
- Clarity of Axis Labels and Legends: The axis labels are generally clear, but the y-axis label in the bottom graph ("Risk-adjusted Alzheimer's mortality odds ratio") could be more descriptive. For instance, it could be expanded to "Risk-Adjusted Odds Ratio of Death from Alzheimer's Disease (Relative to Chief Executives)." The use of color-coded dots to represent different occupations is effective, and the legend is clear.
- Effectiveness of Logarithmic Scale Explanation: The explanation of the logarithmic scale in the caption is accurate and relatively easy to understand. However, the concept of a logarithmic scale might still be challenging for some readers. A brief explanation in the main text or a visual aid in the figure itself could further improve understanding.
- Overall Clarity of the Figure: The figure is well-designed and effectively conveys the main findings of the study. The use of two complementary graphs provides a comprehensive picture of the relationship between occupation and Alzheimer's mortality. However, the figure might be challenging for readers who are not familiar with the concepts of risk adjustment, odds ratios, and logarithmic scales.
- Accessibility to Non-Expert Readers: The figure is somewhat accessible to non-expert readers, especially the top graph. However, the bottom graph and the concepts of odds ratios and logarithmic scales might be difficult to grasp without some background knowledge in statistics. Providing a more detailed explanation of these concepts in the main text or in a supplementary material could improve the figure's accessibility.
- Caption Clarity and Completeness: The caption is detailed and provides most of the necessary information to interpret the figure. However, it could explicitly state that each dot in the graphs represents an occupation. Also, while it mentions the adjustment variables, it could briefly reiterate why these adjustments are necessary.
Fig 2 | Mortality from Alzheimer's disease as underlying or contributing cause...
Fig 2 | Mortality from Alzheimer's disease as underlying or contributing cause of death by occupation. Risk adjusted mortality odds ratios were adjusted for age at death, sex, race, ethnic group, and educational attainment using logistic regression A logarithmic scale was used for the y axis to allow for accurate visual comparison of effect sizes between occupations, as the logarithmic scale equalizes the distances between ratios and their reciprocals. Adjusted odds ratios were calculated using chief executives (US Census Bureau occupation code 0010) as an arbitrary reference group
- Purpose of the figure: This figure explores how often Alzheimer's disease is listed as a cause of death, either the main cause or a contributing factor, across different jobs. It specifically looks at whether being an ambulance driver or a taxi driver is associated with a different likelihood of having Alzheimer's mentioned on one's death certificate compared to other occupations.
- What the graph shows: The graph plots occupations along the horizontal axis (x-axis) and a measure called the "risk-adjusted mortality odds ratio" on the vertical axis (y-axis). This odds ratio is a way of comparing the odds of having Alzheimer's listed on the death certificate for people in different jobs, after adjusting for factors like age, sex, race, ethnicity, and education level. The adjustment is done using a statistical technique called logistic regression, which helps to ensure that the comparisons are fair. A logarithmic scale is used on the y-axis. This type of scale is useful for displaying data that covers a wide range of values. On a logarithmic scale, equal distances represent equal *ratios* rather than equal *differences*. This allows for easier visual comparison of the odds ratios, especially when some are very small and others are very large. The graph uses different colors to mark different types of occupations: taxi drivers are shown in orange, ambulance drivers in blue, and all other occupations in gray.
- Explanation of "underlying or contributing cause of death": When someone dies, there can be multiple factors contributing to their death. The "underlying cause" is the main disease or condition that initiated the sequence of events leading to death. A "contributing cause" is any other disease or condition that also played a role but was not the primary cause. In this study, the researchers are looking at death certificates where Alzheimer's disease is listed as either the underlying cause or a contributing cause. This means they are considering cases where Alzheimer's was a significant factor in the person's death, even if it wasn't the sole cause.
- Explanation of "risk-adjusted mortality odds ratio": An odds ratio is a way to compare the odds of an event happening in one group to the odds of it happening in another group. In this case, the event is having Alzheimer's disease listed on the death certificate. The "mortality odds ratio" compares the odds of this event for people in different occupations. "Risk-adjusted" means that the odds ratios have been adjusted to account for differences in factors like age, sex, race, ethnicity, and education level between the groups. This adjustment is important because these factors can influence both a person's risk of developing Alzheimer's and the type of job they have. By adjusting for these factors, the researchers can get a clearer picture of the relationship between occupation and Alzheimer's mortality. The adjusted odds ratios are calculated relative to a reference group, which in this case is chief executives. An odds ratio of 1 means that the odds of having Alzheimer's on the death certificate are the same as for chief executives. An odds ratio less than 1 means the odds are lower, and an odds ratio greater than 1 means the odds are higher.
- Use of logarithmic scale: A logarithmic scale is used on the y-axis to make it easier to compare odds ratios that are very different in size. On a logarithmic scale, the distance between 1 and 2 is the same as the distance between 10 and 20, or between 100 and 200. This is because each step on the scale represents a multiplication by a constant factor (e.g., doubling). The caption explains that the logarithmic scale "equalizes the distances between ratios and their reciprocals." This means that an odds ratio of 2 (double the odds) is the same distance from 1 as an odds ratio of 0.5 (half the odds) on the logarithmic scale.
- Appropriateness of Adjustment: Adjusting for age, sex, race, ethnic group, and educational attainment is essential for a valid comparison of Alzheimer's mortality across occupations. These are well-established confounders in epidemiological studies, influencing both Alzheimer's risk and occupational choices. The use of logistic regression for this adjustment is statistically sound.
- Inclusion of Underlying or Contributing Cause: Considering Alzheimer's as both an underlying and contributing cause of death broadens the scope of the analysis and provides a more comprehensive picture of the disease's impact. This approach is justified given the complex and multifactorial nature of Alzheimer's disease and its contribution to mortality.
- Use of Logarithmic Scale: The application of a logarithmic scale for the y-axis is appropriate for visualizing odds ratios, especially when they vary widely. This allows for a more accurate visual comparison of effect sizes across different occupations. The explanation provided in the caption regarding the equalization of distances between ratios and their reciprocals is accurate.
- Choice of Reference Group: The use of chief executives as a reference group for calculating adjusted odds ratios is arbitrary but does not invalidate the findings. The interpretation of the odds ratios is contingent on this choice, but the relative comparisons between other occupations remain informative.
- Clarity of Axis Labels: The y-axis label, "Risk-adjusted mortality odds ratio," could be more descriptive. While technically accurate, it might not be immediately clear to all readers. An alternative could be "Adjusted Odds Ratio of Alzheimer's Disease on Death Certificate (Relative to Chief Executives)." The x-axis is implicitly understood to represent different occupations, but this could be stated explicitly.
- Effectiveness of Color Coding: The use of different colors to represent taxi drivers, ambulance drivers, and other occupations is effective for visual differentiation. The legend, although not shown in the provided text, is presumably clear and easy to understand.
- Explanation of Logarithmic Scale: The explanation of the logarithmic scale in the caption is accurate but could be challenging for readers unfamiliar with this concept. A more intuitive explanation or a visual aid within the figure could enhance understanding.
- Overall Clarity and Accessibility: The figure effectively communicates the main findings of the sensitivity analysis, showing that the pattern observed in the primary analysis holds when considering Alzheimer's as an underlying or contributing cause of death. However, the figure's reliance on the concept of odds ratios and the use of a logarithmic scale might make it less accessible to readers without a strong statistical background.
- Caption Completeness: The caption provides a comprehensive description of the figure's components and the methods used. However, it could benefit from a brief introductory sentence summarizing the figure's main message or purpose. Additionally, while it mentions the adjustment variables, it could briefly reiterate why these adjustments are crucial for the analysis.
Fig 3 | Mortality from Alzheimer's disease among bus drivers, aircraft pilots,...
Fig 3 | Mortality from Alzheimer's disease among bus drivers, aircraft pilots, ship captains, and other occupations. Risk-adjusted percentages and mortality odds ratios were adjusted for age at death, sex, race, ethnicity, and educational attainment using logistic regression. In the bottom graph, a logarithmic scale was used for the y axis to allow for accurate visual comparison of effect sizes between occupations, as the logarithmic scale equalizes the distances between ratios and their reciprocals. Adjusted odds ratios were calculated using chief executives (US Census Bureau occupation code 0010) as an arbitrary reference group