辅导 P6110: Statistical Computing with SAS Final Project调试Haskell程序

P6110: Statistical Computing with SAS

Final Project

Due: Wednesday, 04/30/2025 @5:00p EST

This project deals with data collected with the intention of determining the risk factors and social determinants for cognitive impairment and early-stage dementia in 266 Philadelphia adults aged 50-64 (at recruitment). Exclusion criteria include: 1. history of cancer other than non-melanoma skin cancer, 2. presence of a clinical diagnosis of dementia, and 3. visual, hearing, or physical impairment the precludes active participation in study and precludes ability to complete study questionnaires.

Participants underwent phlebotomy, measurement of vital signs and anthropometric measures, interviews of medical history, cognitive assessments, and interviews of social determinants of health and mental/physical conditions. Given the depth of this dataset ,  investigators have posed a few potential questions.

In each prompt description, a few relevant questions have been included to help guide your analysis. You should consider these questions carefully as you prepare your results for your submission, though you are free to conduct additional analysis (which may include any other variables in the dataset) if you feel it may help inform. your primary aim.

Prompt One – Associations Between Diabetes and Cognitive Impairment in Late Middle-Aged Adults

Eleven percent of people aged 65 years and older have dementia in the United States (US), and the most common cause is Alzheimer’s disease (AD), followed by vascular dementia. One of the consistent risk factors reported for dementia, including AD and vascular dementia, is type 2 diabetes, and its antecedent, pre-diabetes. The role of type 2 diabetes and pre-diabetes in dementia is of significant public health importance because approximately 12% of the US adult population has type 2 diabetes (30.2 million) while another 33% has pre-diabetes (84.1 million). Decades of advances in AD research, particularly in AD biomarkers, have led to the dominance of brain amyloid and neurodegeneration as neuropathological constructs.

A cross-sectional study was conducted to determine the relationship between type 2 diabetes and AD neuropathology. Cognitive impairment is measured by: 1. Brain amyloid plaques, measured via PET scans, as amyloid beta (Aβ) global standardized uptake value ratio (SUVR) in each lobe and hemisphere. Mean Aβ burden should be derived as the average of these measures, and 2. Neurodegeneration, measured via  MRI as cortical thickness (mm) in signature regions of the brain related to Alzheimer’s disease.

According to the American Heart Association (AHA), diabetes categories are normal glucose tolerance (NGT; defined by hemoglobin A1c [HbA1c] < 5.7%), pre-diabetes (HbA1c 5.7 to 6.4%), and type 2 diabetes (HbA1c > 6.4%). Patients are also considered to have diabetes if they have ever received a clinical diagnosis of type 2 diabetes in the  past, regardless of current HbA1c.

Questions to Consider

1.  Primary Aim: Are there differences in mean Aβ levels among diabetes categories?

2.  Clinically, Aβ SUVR is often used to determine Aβ positivity (Aβ >1.15). Is diabetes associated with Aβ positivity?

3.  Are there differences in physical neurodegeneration (as outlined above) among diabetes categories?

4.  Is there an association between your outcomes and continuous HbA1c?

5.  Does patient use of diabetes medication affect these associations?

6.  Are these effects consistent across other criteria for determining diabetes categories (see Metadata)?

7.  Are your effects consistent after adjusting for potential confounders? Thoughtfully consider which covariates to include in your model. In past studies, evidence has shown that age, sex, depression, low physical activity, and adiposity (high waist circumference) may be risk factors for vascular dementia.

Prompt Two – Sex and Genetic Risk Factors for Cognitive Impairment in Late Middle-Aged Adults

Eleven percent of people aged 65 years and older have dementia in the United States (US), and the most common cause is Alzheimer’s disease (AD), followed by vascular dementia. This prevalence is higher for females compared to males, the causes for which are unclear but could relate to differences in brain structure, biochemistry, function, and susceptibility to developing AD in response to genetic factors. Decades of advances in AD research, particularly in AD biomarkers, have led to the dominance of brain amyloid and neurodegeneration as neuropathological constructs.

A cross-sectional study was conducted to determine the relationship between sex/genetic risk factors and AD neuropathology. Cognitive impairment is measured by:

1. Brain amyloid plaques, measured via PET scans, as amyloid beta (Aβ) global standardized uptake value ratio (SUVR) in each lobe and hemisphere. Mean Aβ burden should be derived as the average of these measures, and 2. Neurodegeneration, measured via MRI as cortical thickness (mm) in signature regions of the brain related to Alzheimer’s disease. APOE, one the strongest genetic determinant of AD, has genotypes derived from its two SNPs (rs429358 & rs7412). The presence of an ε4 allele in an APOE genotype is considered a risk factor for AD.

Questions to Consider

1.  Primary Aim: Are there differences in mean Aβ levels between APOE-ε4 carrier statuses or sexes?

2.  Clinically, Aβ SUVR is often used to determine Aβ positivity (Aβ >1.15). Are APOE-ε4 carrier statuses or sex associated with Aβ positivity?

3.  Are there differences in physical neurodegeneration (as outlined above) between APOE-ε4 carrier statuses or sexes?

4.  What is the distribution of APOE-ε4 between sexes? Do genetic risk factors for your outcomes behave differently between sexes?

5.  Are your effects consistent after adjusting for potential confounders? Thoughtfully consider which covariates to include in your model. In past studies, evidence has shown that age, sex, depression, low physical activity, and adiposity (high waist circumference) may be risk factors for vascular dementia.

Prompt Three - Associations Between Metabolic Deficiencies and Cognitive Impairment in Late Middle-Aged Adults

Eleven percent of people aged 65 years and older have dementia in the United States (US), and the most common cause is Alzheimer’s disease (AD), followed by vascular dementia. Metabolic syndrome (MetS) is defined as a cluster of cardiometabolic conditions, including elevated blood pressure, excess central adiposity, high levels of triglycerides, and abnormal cholesterol. MetS is a known risk factor for diabetes, cardiovascular disease, and all-cause mortality, as well as cognitive impairment, vascular dementia, and dementia due to Alzheimer’s disease.

A cross-sectional study was conducted to determine the relationship between MetS and memory function. Memory function is measured by: 1. Total words recalled on the selective reminding (SRT) test (your primary outcome), and 2. Percentile rank on other memory tests, including the C-F-L Letter Fluency test, the Category Fluency Test, and  the Color Trails tests. The presence of MetS/a MetS risk score can be derived from its  five components (see Metadata).

Questions to Consider

1.  Primary Aim: Are there differences in SRT scores by MetS status?

2.  Are there differences in other neuropsychological exams by MetS status?

3.  Are your results consistent when using MetS risk score as a continuous exposure rather than categorical MetS status?

4.  Are there different associations with your outcomes when testing your MetS   scores individually? Are some more strongly related to your primary outcome than others?

5.  Are your effects consistent after adjusting for potential confounders? Thoughtfully consider which covariates to include in your model. In past studies, evidence has shown that age, sex, education, depression, and low physical activity may be risk factors for depreciated memory function.

Submission Requirements

You will be submitting two (2) files for your project: