Nurses' Health Study Insights

Podcast

Play an AI-generated podcast conversation about this lesson

Questions and Answers

What was the primary purpose of the Nurses’ Health Study?

To assess the levels of pesticides in the environment

To investigate the incidence of breast cancer in a specific population (correct)

To analyze dietary habits among women

To evaluate the effectiveness of cancer treatments

Why was a case-control study proposed instead of analyzing all blood samples?

It is a faster method for determining pesticide exposure

There were not enough women diagnosed with breast cancer

A cohort study does not provide enough data

Quantifying pesticide levels in blood is too expensive (correct)

What was the significance of having 1,439 incident cases of breast cancer during the follow-up?

It represented the total number of women studied

It helped in forming the hypothesis about pesticide exposure (correct)

It suggested that monitoring breast cancer is unnecessary

It indicated that breast cancer was not a prevalent disease

What proportion of women in the study had high exposure levels to pesticides?

15% Signup and view all the answers

What is the rationale for sampling twice the number of cases in a case-control study?

It creates a more balanced sample between cases and controls Signup and view all the answers

How many total women were included in the case-control study nested within the Nurses’ Cohort?

4,317 Signup and view all the answers

What does an odds ratio of 7.63 indicate about the relationship between protein deficiency and Kwashiorkor?

Protein deficiency is associated with an increased risk of Kwashiorkor. Signup and view all the answers

What is a key advantage of conducting a case-control study compared to a cohort study?

It is less costly and more efficient Signup and view all the answers

What was the initial hypothesis tested in the Nurses’ Health Study?

Women with high pesticide exposure have an increased risk of breast cancer Signup and view all the answers

In the case-control study, how many cases had protein deficient diets?

212 Signup and view all the answers

What is the formula used to calculate the odds ratio in case-control studies?

Odds Ratio = (a/b) / (c/d) Signup and view all the answers

If 'a' represents the number of cases exposed to the risk factor, which value corresponds to 'b' in the data table?

72 Signup and view all the answers

What are unexposed cases assigned to if they are at risk for Kwashiorkor?

c Signup and view all the answers

What does the odds ratio represent in a case-control study?

The odds of disease occurrence based on exposure status. Signup and view all the answers

If 88 unexposed individuals did not have protein deficiency, what is the total of individuals unexposed to the risk factor?

228 + 88 Signup and view all the answers

How many total individuals were included in the case-control study?

600 Signup and view all the answers

What is the primary focus of a case-control study?

Classify participants according to their outcome status Signup and view all the answers

In the study of physical activity and lung cancer, what was the significant factor assessed from 20 years ago?

Prior physical activity Signup and view all the answers

What was the main finding regarding the odds ratio for lung cancer in relation to physical activity?

Decreased odds of lung cancer with increased physical activity Signup and view all the answers

What is the purpose of matching controls in a case-control study?

To equate selected variables such as age and sex Signup and view all the answers

What hypothesis is illustrated related to pesticide exposure and breast cancer?

High pesticide exposure is associated with an increased risk of breast cancer Signup and view all the answers

What would most likely be a limitation of a case-control study?

It cannot establish cause-and-effect relationships directly Signup and view all the answers

Why might researchers prefer a case-control study over a cohort study in some scenarios?

Case-control studies are quicker and less expensive to conduct Signup and view all the answers

What classifies someone as a 'case' in a case-control study?

A participant with the health outcome being studied Signup and view all the answers

What is the overall measurement error for people with asthma in the study?

20% Signup and view all the answers

What type of misclassification is observed in the study?

Differential misclassification Signup and view all the answers

What is the misclassified incidence rate ratio calculated in the study?

1.49 Signup and view all the answers

How many individuals in total were classified as 'not exposed' to asthma in the misclassified data?

1000 Signup and view all the answers

In the context of the misclassification described, what is the expected outcome for healthy children?

10% misclassification leading to incorrect exposure classification Signup and view all the answers

What is the primary purpose of a matched design in studies?

To make groups comparable regarding extraneous factors Signup and view all the answers

In matched case-control studies, what is the main characteristic that is matched between cases and controls?

Outcome status Signup and view all the answers

What type of matching involves pairing participants one-to-one?

Individual matching Signup and view all the answers

What is a significant limitation of matching in cohort studies?

It can be impractical and expensive Signup and view all the answers

In the context of analysis for matched designs, what is the basic unit of analysis?

The pair of case and control Signup and view all the answers

What is the primary purpose of selecting controls in a case-control study?

To estimate the exposure distribution in the source population Signup and view all the answers

Which of the following factors is crucial to consider when selecting matching characteristics?

Identifying potential confounders Signup and view all the answers

Which of the following is NOT a necessary requirement for the selection of controls?

Controls must be selected based on their disease history Signup and view all the answers

Which method is considered the least favorable for finding controls in a case-control study?

Convenience-based controls Signup and view all the answers

What is the primary difference between matching in case-control studies and cohort studies?

Cohort studies match exposed to unexposed individuals Signup and view all the answers

What statistical method is typically used in the analysis of matched designs?

Paired analysis Signup and view all the answers

In a case-control study, how should controls be selected in relation to their exposure status?

Controls must be selected independently of their exposure status Signup and view all the answers

What is a philosophical question to consider when selecting controls in case-control studies?

Would the controls have been identified as cases if they had experienced the outcome? Signup and view all the answers

What is a major advantage of using population-based controls?

They provide a representative sample from the general population Signup and view all the answers

Which of the following is an example of a source for population-based controls?

Driver's license holders Signup and view all the answers

What is a significant challenge when selecting controls in case-control studies?

Ensuring the controls are representative of the source population Signup and view all the answers

Study Notes

Session Information

Session 7, Design 2: Cohort, Case-control Studies
Date: November 1, 2024
Course: HS2801A: Research Methods in Health Sciences
Instructor: Dr. Afshin Vafaei
Course Term: Fall 2024

Announcements

Midterm marks available on OWL (out of 75 total)
- Sections: Fill in the blanks/Calculations (out of 25), MCQ (out of 50)
Exam review today from 12:00 to 12:30 in class
In-person office hours today from 1:00 to 3:00 in HSB Room 215 (for questions)
Online meetings not possible

Final Exam

Date: Friday, December 13, 2024
Time: 7:00 pm to 10:00 pm
Locations: Alumni Hall 15, Alumni Hall 201
Makeup exam date: TBD; January 2025
Lists of last names and corresponding locations provided

Today's Class

Epidemiological (population health & clinical) Studies
- Biological Studies
- Physiological, direct cell observation
Observational Studies
- Descriptive studies
  - Case reports
  - Case series
  - Ecological
  - Cross-sectional
- Analytic studies
  - Case-control
  - Cohort
Experimental Studies
- Randomized Controlled Trial
- Community Trial

Cohort Studies

Definition: A group of individuals followed over time
Key Characteristics
- Participants must be free from the outcome at the beginning of the study
- Exposure and comparison groups must be selected
- Participants must be followed up
- Outcome status must be determined
Exposure determined before the outcome

The Big Picture of a Cohort Study

Cohort followed over time (past to future)
Exposure is determined at the start of the study
Compare the incidence of the disease in the exposed and unexposed groups
More expensive and time-consuming but less vulnerable to information bias

Schematic of a Cohort Study

Participants classified by exposure (observed, not assigned)
Followed until outcome occurs
Different categories for exposed/unexposed individuals with/without outcome

Using a Real Study to Explore Features of Cohort Design

Study purpose: Exploring the relationship between working night shifts and breast cancer
Participants: 78,562 female nurses aged 30-55 with no history of cancer
Exposure: Years of night shift work (obtained via questionnaire in 1988)
Outcome: Incidences of breast cancer (between 1988 and 1998), nonfatal cases by medical records and the National Death Index (fatal cases and deaths from other reasons), and death reported to next of kin

Example #2

Study of multimorbidity and depression in older adults
Aims to assess whether multimorbidity is an independent risk factor for the development of depression
Stratified analyses by region (Canada; Latin America; Albania)
Indicates that multimorbidity does not appear to increase risk of developing depression in older adults between 65-74.

Timing: Prospective vs. Retrospective Cohort Studies

Prospective: Exposure measured before outcome, follow-up after start of study
Retrospective: Exposure measured after study started, study started after outcome

Prospective vs. Retrospective Studies

Prospective: Investigators assemble cohort and follow participants in real-time
Retrospective: Investigators use historical records and link exposure to outcome

Retrospective Cohort Study

Cheaper, faster, and more efficient for diseases with long latency periods
Potential for exposure and confounder data being inadequate
High vulnerability to bias
Need for established recording systems (e.g., military, occupational, birth cohorts)

A Retrospective Cohort Study - Example

Study the relationship between tire manufacturing chemicals and risk of death

Two Questions - Exposure and Disease

Exploring if an occupational chemical exposure causes skin rash or cancer

Incidence (rate) Ratio and Relative Risk

Comparison of incidence rates in two populations/subpopulations
Ratio of incidence rates (exposed/unexposed groups)
Measure of exposure risk and association between exposure and outcome

Example #1 - Incidence Rate Ratio

Study of respiratory infection in pesticide-exposed plant workers
Incidence rate ratio between exposed/unexposed groups

Time of Measurement

Time of outcome measurement is crucial in cohort studies

Common Baseline Time Point & Outcome

Examples of various outcomes recorded at the same point in time

Rolling Entry Point (Variable Baseline)

Outcome measured over time
Participants join the study at various points (different durations, exposure periods)
Person-time: individuals are in the study for various durations; contribute differently to the study

Example: Prospective Cohort Study with Rolling Entry

Example describing the years of follow-up for breast cancer in participants

Cohort Studies - Advantages & Disadvantages

Advantages:
- Valuable when exposure is rare
- Can examine multiple effects of a single exposure
- Easier to determine the temporal relationship between exposure and outcome
- Allows measurement of incidence
Disadvantages:
- Validity affected by losses to follow-up (selection bias)
- Other factors may not be distributed evenly between exposure groups (confounding)
- Inefficient for evaluation of rare diseases
- Expensive and time-consuming (need for large numbers and long follow-up)
- If retrospective, requires good records

Observational Studies - Case-Control Studies

Used when cohort studies are difficult or impractical

Recall Cohort Studies Limitations

Difficulty when large numbers of subjects need lengthy follow-up
Potential for loss to follow for prospective cohort studies to undermine validity
Not suited for rare or long-latency diseases
Exposure data may be expensive to obtain
Solutions: Conduct a case-control study

My Thoughts - Case-Control vs. Cohort Study

Case-Control can efficiently answer study questions regarding exposure/outcome relationships, including rare outcomes
Prospective cohort studies are time-consuming and expensive, while Case-Control can answer the same question in a shorter time and with fewer resources

The Design of a Case-Control Study

Classifying participants (Cases and Controls) & determining exposure status
Direction of inquiry (past exposures)
Defining study population and exposure status

Case-Control Design

Sampling Participants
Historical exposure & outcome occurrence
Identifying participants' exposure status

Cohort Study vs. Case-Control Study

Cohort: Participants classified by exposure, followed until outcome. Compare prior exposures over time to outcome.
Case-Control: Identify cases and controls based on outcome and then determine their exposure status.

Example: Protein Deficiency and Kwashiorkor

Case-control study of Kenyan children
Examined protein deficiency diets in cases with Kwashiorkor relative to controls

Odds Ratio Calculation in Case-Control Study

Calculation of odds ratio from tables

Selection of Cases

Very clear case definition needed
Ideally, case selection involves sampling cases within a source population
Includes all individuals in the source population who meet disease criteria; or, if there are constraints, random sampling

Controls Definition and Purpose

A sample of the source population that produced the cases
Purpose: Estimating exposure distribution in the source population that produced the cases
Knowing the exposure prevalence helps determine the relationship between exposure and outcome

Selection of Controls

Crucial element in case-control studies
Controls must come from the same source population as the cases;
Representative sample necessary
Control selection must be independent of exposure

Where to Find Controls

Population-based, nested, hospital/clinic-based, or family/friend-based

Population-Based Controls

Selecting controls from the general population
Useful when cases are from well-defined geographic regions
Sources: Random digit dialing, phone/internet subscribers, residency lists, voter registration lists, driver's license holders
Benefit: representative source population
Drawback: time consuming, hard to inspire participation, recall past exposure

Nested Controls

Selecting controls from an existing cohort
Controls from a clearly defined source population; already enrolled; willing participants
Benefit: clearly defined source
Drawback: restricted to cohort members; limited hypothesis that can be studied

Hospital- or Clinic-Based Controls

Controls selected from patients in hospitals or clinics
Choose controls with a disease different from the cases
Should be from the same catchment area (same source)
Should be unrelated to the risk factor being studied

Hospital- or Clinic-Based Controls - Example

Example: A case-control study of smoking and myocardial infarction

The Matched Design

Explanation of the concept

Definition/Purpose - Matching

Process of making the study group and comparison group comparable with respect to extraneous factors
Techniques ensure groups are similar to account for confounding
Like randomization; only difference is exposure status.

Matching in Case-Control Studies

Important to match for factors that might be confounding
Making exposure group similar (other factors)
Unit is a pair (case and its matched controls) and paired analysis
Statistical models are well-developed

Types of Matching

Individual matching, performed by participant, each case is matched to a control
Frequency matching, similar distribution of confounders in groups

Matching in Cohort Studies

An effort to mimic randomization
Less common, much less expensive
Sometimes impractical, may require control

Matching in Case-Control Studies (Analysis)

Unit of analysis is a pair (case & its matched control)
Paired analysis & well-developed statistical models

The Unmatched Contingency Table

Table format for cases and controls, showing high/low fat exposures

Odds Ratio - Unmatched Situation

Calculating odds ratio for exposure and metabolic syndrome

Matched Example

Cases are exposed or not. Controls are matched exposed/not exposed in the same group.

Case-Control Studies - Advantages & Disadvantages

Advantages:
- Efficient in terms of time, money, and effort
- Useful for diseases with long latency
- Optimal design for rare diseases
- Can examine multiple exposures
Disadvantages:
- Exposure assessed after outcome development (temporal ambiguity)
- May be uncertain about temporal sequence between exposure and disease
- Recall bias is possible
- Selection bias possible in control groups
- Can only measure odds ratio from the data , not incidence

Correct Measures of Association for Various Designs

Association measure specific to each design (prevalence for cross-sectional, odds ratio for case-control, incidence for cohort)

Midterm Review

Indicates impending review of study content

Odd and Proportion Conversions

Method for converting odds to proportion and vice versa.

Example: Prevalence and Incidence Calculation (Gastroenteritis)

Calculation of prevalence and incidence of gastroenteritis in a sample group

Cross-Sectional Study

Determining the relative risk (effect estimate) of exposure to second-hand smoke and development of asthma.
Exposure is measured and compared to the incidence of the disease.
Prevalence of exposure in the total population is given.

Differential Misclassification

Misclassifications of exposed and unexposed individuals for study factors.
Effect and calculation of misclassified incidence rate ratio

Next Session (November 8, 2024)

Experimental studies topics
Chapter 12 of textbook for quiz 3
In-class format, open book format quiz

Studying That Suits You

Use AI to generate personalized quizzes and flashcards to suit your learning preferences.

Description

Test your knowledge on the Nurses' Health Study and its case-control study aspect. This quiz covers the primary purposes, key findings, and statistical methods involved in the study. Gain a deeper understanding of the implications related to diet and health outcomes.