Chapter 6: Finding the Cause – Analytical Study Designs

Analytical Studies go beyond just describing a health event (descriptive studies); they look for cause-and-effect relationships using a comparison group. The investigator does not assign the exposure (no randomization); they just carefully measure what people naturally choose to do.

1. The Direction of Analytical Studies

Analytical studies are categorized by which direction they look:

• Cohort Study: Proceeds from Exposure – Outcome. (We start with the risk factor and look forward to see who gets sick).
• Case-Control Study: Proceeds from Outcome – Exposure. (We start with people who are sick and look backward to see what risk factors they had).

---

2. Cohort Studies: Exposure to Outcome (Forward Look)

A Cohort is simply a group of individuals sharing a common characteristic (like a group of children who all start using SuperPaste today).

The Cohort Design (The Logic)

1. Select Cohorts: We select a group that is Exposed (e.g., kids who naturally choose to drink sugary sodas every day) and a group that is Unexposed (kids who don’t drink sodas).
2. Follow Up: We follow both groups over time (e.g., for three years).
3. Measure Outcome: We count how many kids in each group develop dental caries (the Incidence).
4. Compare: We compare the Incidence of caries in the Exposed group versus the Unexposed group.

Type of Cohort Study	When the Study Starts (Relative to Caries/Soda)
Prospective Cohort	Exposure (soda habit) and Outcome (caries) have NOT YET occurred. We start now and follow the kids into the future.
Retrospective Cohort	Exposure and Outcome have ALREADY occurred. We look at old records (e.g., old clinic charts from 10 years ago) to see who drank soda and who got caries.
Bidirectional/Ambispective	The Exposure (soda habit) has occurred, but the Outcome (caries) has not. We use existing records to define the exposure and then follow them forward.

The Analysis: Relative Risk (RR)

Since we start with healthy people and count how many develop the disease, Cohort Studies allow us to calculate Incidence and the Relative Risk (RR):

RR Value	Interpretation
RR = 1	No association. Soda drinking does not affect caries risk.
RR > 1	Positive association (Risk Factor). Soda drinkers are more likely to get caries.
RR < 1	Negative association (Protective Factor). Soda drinkers are less likely to get caries.

Strengths & Weaknesses

• Strength: Temporality is certain. We know the Exposure (soda) happened before the Outcome (caries). We can also calculate Incidence directly.
• Weakness: Large sample size needed, expensive, and takes a long time (long follow-up), making it bad for rare diseases.

---

3. Case-Control Studies: Outcome to Exposure (Backward Look)

Case-Control studies start with the disease and look backward in time. This is the opposite logic of a Cohort Study.

The Case-Control Design (The Logic)

1. Select Groups: We select Cases (children who HAVE severe dental caries) and Controls (children who DO NOT HAVE severe dental caries).
2. Measure Exposure (Backward): We interview both groups (or check their old records) to find out their Past Exposures (e.g., “Did you drink sugary soda 5 years ago?”).
3. Compare: We see if the Odds of Exposure (soda drinking) are higher in the Cases than in the Controls.

Selection Component	Important Consideration in Dentistry
Cases	Should be clearly defined. It’s often better to use Incident Cases (newly diagnosed) rather than Prevalent Cases (already existing), because prevalent cases may be related to survival (who lives long enough to be included) rather than just the development of the disease.
Controls	Must represent the background rate of exposure in the population from which the cases came. (If our cases came from one clinic, our controls should also come from that same clinic).

The Analysis: Odds Ratio (OR)

In Case-Control Studies, we cannot calculate Incidence. Instead, we calculate the Odds Ratio (OR), which is an estimate of the Relative Risk.

• The interpretation of the OR is the same as RR: OR>1 means a positive association.

Strengths & Weaknesses

• Strength: Quick to conduct, inexpensive, and excellent for studying rare diseases or diseases with a long latency period (long time between exposure and outcome).
• Weakness: Susceptible to Recall Bias (cases with a severe disease might remember their past exposures differently than healthy controls) and we cannot calculate Incidence.