Effectiveness of stress management interventions to change cortisol levels: a systematic review and meta-analysis.

This systematic review and meta-analysis investigated whether various psychological stress management interventions could change cortisol levels, often referred to as the "stress hormone." The researchers synthesized findings from multiple randomized controlled trials (RCTs) to determine the overall effectiveness of these interventions compared to control conditions.

The interventions were categorized into four main types:

**Mindfulness and Meditation:** Practices focusing on present moment awareness, often involving guided attention and breathwork.

**Relaxation:** Techniques aimed at reducing physiological arousal, such as progressive muscle relaxation, diaphragmatic breathing, or guided imagery.

**Mind-Body Therapies:** Broader approaches that integrate mental and physical practices, potentially including yoga, tai chi, or biofeedback.

**Talking Therapies:** Psychological counseling or therapeutic approaches like Cognitive Behavioral Therapy (CBT), designed to help individuals identify and change unhelpful thought patterns and behaviors.

These interventions were compared against various control conditions, which were grouped into:

**Active Control Groups:** Participants received an alternative intervention that was not expected to directly target stress or cortisol, but still involved some level of engagement (e.g., health education, attention control).

**Passive Control Groups:** Participants received no intervention, were on a waitlist, or received usual care, serving as a baseline for comparison.

The primary outcome measured was **cortisol levels**, which were assessed in different ways:

**Cortisol Awakening Response (CAR):** The sharp increase in cortisol levels observed within the first 30-45 minutes after waking. A healthy CAR is typically a robust rise, and dysregulation (either blunted or exaggerated) can indicate stress-related issues.

**Diurnal Cortisol:** Cortisol levels measured at multiple points throughout the day, typically showing a peak in the morning and a gradual decline throughout the day. A "flatter" diurnal slope (less difference between morning and evening levels) is often associated with chronic stress and poorer health.

**General Cortisol Levels:** Single or averaged cortisol measurements, without specific focus on the awakening response or diurnal pattern.

Cortisol was measured from biological samples including blood, saliva, and hair.

The meta-analysis included a combined total of **3,508 participants** across 58 individual studies. The key characteristic of the population was that they were **healthy adults aged 18 years or older**, meaning the studies specifically excluded individuals with pre-existing medical conditions such as cancer, other diseases, obesity, pregnancy, or diagnosed psychiatric conditions. This focus on "non-patient groups" is important as it aims to understand the impact of stress management on general well-being and stress regulation, rather than as a treatment for specific illnesses. The studies were conducted in various settings typical of clinical trials, such as universities, research centers, or community health settings.

Cortisol, the primary outcome, was measured using three main biological sample types:

**Saliva:** This is a common and non-invasive method. Salivary cortisol reflects the unbound, biologically active fraction of cortisol in the blood and is particularly useful for capturing dynamic changes, such as the Cortisol Awakening Response (CAR) and diurnal rhythms, as samples can be collected easily at home by participants.

**Blood:** Blood samples provide a direct measure of circulating cortisol. While accurate, it is more invasive and typically requires collection by trained personnel, making it less practical for frequent, at-home sampling needed for diurnal patterns.

**Hair:** Hair cortisol provides a retrospective measure of cumulative cortisol exposure over a longer period (weeks to months), as cortisol is incorporated into hair strands as they grow. This method is useful for assessing chronic stress levels rather than acute or daily fluctuations.

The specific types of cortisol measurements analyzed were:

**Cortisol Awakening Response (CAR):** This involves collecting saliva samples at specific time points immediately upon waking and then at intervals (e.g., 15, 30, 45, 60 minutes) to capture the characteristic sharp rise in cortisol.

**Diurnal Cortisol Slope:** This involves collecting multiple saliva samples throughout the day (e.g., morning, noon, afternoon, evening) to track the natural decline of cortisol levels from morning to night.

**General Cortisol Levels:** This could refer to single point measurements or averages of cortisol concentrations without specific emphasis on the CAR or diurnal pattern. The meta-analysis prioritized diurnal measures over single measures when both were available.

The individual studies included in the meta-analysis measured cortisol at baseline (before the intervention) and post-intervention to assess changes over time. For studies that reported standard error (SE) or 95% confidence intervals instead of standard deviation (SD), the researchers used standard formulas to calculate SD to ensure consistency for the meta-analysis.

This study was a **systematic review and meta-analysis** of **randomized controlled trials (RCTs)**. This design is considered the highest level of evidence for determining the effectiveness of interventions.

**How they ran the study:**

1. **Protocol and Registration:** The researchers pre-registered their study protocol on PROSPERO (CRD42019120066), which specifies the research questions, search strategy, and analysis plan before conducting the review. This helps to reduce bias and increase transparency.

2. **Eligibility Criteria (PICOS):**

* **Population (P):** Healthy adult subjects (aged > 18 years), who could be stressed or not stressed at baseline. Crucially, studies involving patients with cancer, diseases, obesity, pregnancy, psychiatric, or other health issues were *excluded*.

* **Intervention (I):** Any psychological stress-management interventions, including mindfulness, meditation, relaxation, and talking therapies (like CBT). Pharmacological interventions were excluded.

* **Comparator (C):** Waitlist control or other intervention (active or passive control groups). Studies with no control group were excluded.

* **Outcome (O):** Cortisol level measures in blood, saliva, and hair. Cortisol could be measured with or without an acute stress test. Other physiological measures like heart rate or blood pressure, or only stress test assessments, were excluded.

* **Study Design (S):** Randomized Controlled Trials (RCTs) published in English language journal articles. Non-RCTs, open trials, reviews, posters, presentations, case reports, dissertations, and letters were excluded.

3. **Search Strategy:** Six major electronic databases (Medline, PsychInfo, Embase, CINAHL, Cochrane, and Web of Science) were systematically searched using keywords like "cortisol" and "stress management intervention." Google Scholar was also used to find additional papers citing the included studies. The search was regularly updated, with the last search on April 6, 2023.

4. **Study Selection:** A total of 1171 studies were initially identified, with 3 more from Google Scholar. After removing duplicates, titles and abstracts were screened by one reviewer (OR), with a 20% overlap checked by another (SW) to ensure consistency. Full-text screening was then performed by OR, with 20% overlap by SW. Disagreements were resolved by consensus or by a third researcher (DO'C). The inter-rater reliability was high (K = 0.76, p < .001). This rigorous, two-stage screening process helps ensure that only relevant and high-quality studies are included.

5. **Data Extraction and Coding:** Key information was extracted from the 58 included studies using a standardized data extraction table based on Cochrane guidelines, with additional components specific to cortisol measurement. When data was unclear, study authors were contacted. Data extracted included mean age, standard deviations (calculated from standard errors or confidence intervals where necessary), and details about the intervention, control, and cortisol measurement methods.

6. **Intervention Categorization:** Interventions were coded into four categories: mind-body therapies, mindfulness, relaxation, or talking therapies.

7. **Statistical Approach:** A **random effects meta-analysis** was conducted. This statistical method is appropriate when assuming that the true effect size varies across studies due to differences in populations, interventions, or settings, rather than assuming a single true effect size (as in a fixed-effect model). It provides a more conservative estimate of the overall effect. The meta-analysis calculated Hedges' g, a measure of effect size, which represents the standardized mean difference between the intervention and control groups.

8. **Heterogeneity and Moderation Analysis:** The researchers investigated factors that might explain differences in effect sizes across studies (heterogeneity). They explored the moderating effects of:

* Type of cortisol measurement (blood, saliva, hair; CAR vs. diurnal).

* Type of control group (active vs. passive