Case-Control Study

How Case-Control Studies Work

Study Structure

Case-control studies work backward from outcome:

1. Identify cases - People with the outcome of interest
2. Select controls - People without the outcome
3. Look backward - Assess past exposures
4. Compare exposure rates - Cases vs controls
5. Calculate odds ratio - Strength of association

Visual Representation

Present (Outcome Known)          Past (Exposure Unknown)
        |                                |
   +---------+                     +----------+
   |  Cases  | -----------------> | Exposed? |
   | (have   |                     |   Y/N    |
   | outcome)|                     +----------+
   +---------+
        |
   +---------+                     +----------+
   | Controls| -----------------> | Exposed? |
   | (no     |                     |   Y/N    |
   | outcome)|                     +----------+
   +---------+

Relevance to Peptides

Safety Signal Investigation

Case-control studies excel at investigating rare adverse events:

Example: Thyroid Cancer and GLP-1 Agonists

• Cases: Patients with thyroid cancer
• Controls: Matched patients without thyroid cancer
• Exposure: Prior GLP-1 agonist use
• Question: Is GLP-1 use associated with thyroid cancer risk?

Peptide Research Applications

Application	Cases	Controls
Pancreatitis risk	Patients with pancreatitis	Matched without
Gallbladder disease	Cholecystectomy patients	Surgery-free patients
Injection site reactions	Patients with reactions	Patients without
Rare metabolic effects	Affected individuals	Unaffected individuals

Post-Marketing Surveillance

When rare adverse events are reported after peptide approval:

• Case-control studies quickly investigate signals
• Much faster than waiting for cohort data
• Cost-effective for rare outcomes
• Can use existing medical records

Advantages and Limitations

Advantages

Benefit	Explanation
Efficient for rare outcomes	Don’t need huge cohorts
Fast and inexpensive	No long follow-up needed
Multiple exposures	Can study many risk factors
Existing data	Can use medical records
Ethical for harmful exposures	No deliberate exposure

Limitations

Challenge	Impact
Recall bias	Cases may remember differently
Selection bias	Control selection critical
Temporal ambiguity	Hard to establish sequence
No incidence rates	Can’t calculate absolute risk
Confounding	Difficult to fully control

Control Selection

Matching Strategies

Proper control selection is critical:

Strategy	Description	Use When
Individual matching	Each case paired with control	Small studies
Frequency matching	Groups balanced on factors	Larger studies
Hospital controls	Other patients at same facility	Hospital-based cases
Population controls	Random community sample	Population-based cases

Common Matching Variables

• Age
• Sex
• Geographic location
• Time period
• Healthcare utilization

Interpreting Results

The Odds Ratio

Case-control studies report odds ratios (OR):

OR Value	Interpretation
OR = 1.0	No association
OR > 1.0	Increased odds with exposure
OR < 1.0	Decreased odds with exposure
OR = 2.0	Twice the odds
OR = 0.5	Half the odds

Confidence Intervals

• CI includes 1.0: Not statistically significant
• CI excludes 1.0: Statistically significant
• Narrow CI: More precise estimate
• Wide CI: Less precise estimate

Case-Control vs Other Designs

Feature	Case-Control	Cohort	RCT
Direction	Backward	Forward	Forward
Outcome	Selected on	Measured	Measured
Best for	Rare outcomes	Rare exposures	Efficacy
Time	Short	Long	Medium
Cost	Low	High	High
Bias risk	Higher	Medium	Lower

Frequently Asked Questions

When should case-control studies be used for peptide research?

Case-control designs are ideal when investigating rare outcomes like uncommon adverse events, when quick answers are needed (safety signals), when the condition takes a long time to develop, or when prospective studies would be too expensive. They’re commonly used in post-marketing safety surveillance.

Can a case-control study prove a peptide causes harm?

No, case-control studies cannot prove causation. They identify associations that suggest a relationship between exposure and outcome. The retrospective design and potential for bias limit causal inference. However, strong, consistent associations across multiple well-designed studies provide compelling evidence.

What is the difference between odds ratio and relative risk?

Odds ratio compares the odds of exposure between cases and controls; relative risk compares the probability of outcome between exposed and unexposed. For rare outcomes, OR approximates RR. For common outcomes, OR overestimates the strength of association compared to RR.