A pilot study of an online community resiliency model (CRM) in people with memory impairment Authors: Manju Ramakrishnan, Divinity Amos-Richards, Linda Grabbe, Abha Namjoshi, Ambar Kulshreshtha Published: 2025 ABSTRACT BackgroundHigh stress affects mental and physical health, with African Americans at greater risk due to systemic inequities. The Community Resiliency Model (CRM)®, an evidence-based intervention teaching sensory awareness skills to enhance resiliency, was evaluated among African American and White elderly participants with memory impairment to assess its effectiveness on stress-related outcomes.MethodsFifty-four African American and White participants with mild cognitive impairment (MCI) were recruited in Atlanta, Georgia, for a 3-h online CRM workshop in 2021–2022 delivered by a CRM-trained nurse practitioner. Participants completed surveys pre- and 4 weeks post-intervention, including Perceived Stress Scale (PSS-10), Patient Health Questionnaire (PHQ-9), Somatic Symptoms Scale (SSS-8), and Connor-Davidson Resilience Scale (CD-RISC-10). Paired t-tests assessed the changes in outcome measures.ResultsAmong fifty-four participants, twelve were lost to follow-up. Mean age was 68 (range: 39–83 years); majority were female (81%) and African American (71%). Participants showed improvements in mean differences for depressive symptoms (PHQ-9 difference: 1.4; [95% CI 0.5, 2.4]; p-value = 0.0035), somatic symptoms (SSS-8 difference: 1.9; [95% CI 0.81, 3.0]; p-value = 0.0011), and resilience (CD-RISC-10 difference: − 1.4; [95% CI − 2.8, − 0.0026]; p-value = 0.05). Perceived stress improved but was not statistically significant (PSS-10 difference: 1.0; 95% CI − 0.77, 2.9; p-value = 0.25). No racial differences were observed.ConclusionCRM can enhance mood, alleviate somatic symptoms, and improve resilience in elderly participants with memory concerns. It shows promise as a brief, accessible intervention for mental well-being in diverse populations. Larger studies are needed for generalizability, given the limited sample size.Supplementary InformationThe online version contains supplementary material available at 10.1007/s44192-025-00333-z. Introduction Chronic stress is common among elderly adults and has been proven to affect their quality of life and worsen chronic conditions. High levels of chronic stress can be detrimental to one’s health, contributing to a range of negative effects on physical and mental health, and are associated with significant morbidity and mortality, particularly in the elderly population [1, 2]. Chronic stress has been associated with various health problems, such as hypertension, diabetes, and cardiovascular disease (CVD), as well as anxiety and depression [3–5]. It also leads to worsening of lifestyle factors, including excessive alcohol consumption, smoking, and physical inactivity [6–8]. Furthermore, chronic life stress is also independently associated with cognitive decline, which may be attributed to the activation of the hypothalamic-pituitary-adrenal axis, increased levels of glucocorticoid hormones, and heightened systemic inflammation in the body [9, 10]. African Americans are twice as likely to develop cognitive decline [11]. The prevalence of systemic racism in society further compounds the debilitating effects of stress on African Americans, elevating the risk [9, 10, 12–15]. Studies have found that when African Americans were exposed to segregation and racial discrimination, they experienced a higher risk of developing CVD and cognitive decline when compared to White participants due to these factors causing their stress levels to increase significantly, which led to physiological dysfunctional patterns (i.e., blunted β- adrenergic receptor responsiveness, hypertension, higher levels of a pro-inflammatory cytokine) [9, 10, 12, 13, 15]. The US population has recently experienced a range of worsening stressors due to the COVID-19 pandemic, political polarization, racial prejudices, and economic uncertainty [16]. Stress and its adverse effects on physical and mental health cost the US healthcare system more than $300 billion per year [17]. Thus, it is essential to determine what intervention strategies can mitigate the negative effects of stress to improve the quality of life and health outcomes for all population subgroups. The Community Resiliency Model (CRM) is a self-care program that gives insight into the biological perspective of mental health and stress. Studies have found that CRM is helpful in improving the mental health and well-being of individuals by reducing stress levels [18–21]. CRM accomplishes this by incorporating the concept of interoception, which in turn promotes mindfulness, conscious recognition, and observation of current thoughts, emotions, and other internal or external stimuli to foster tranquility. This has been shown to be a helpful treatment and prevention of mental health disorders, such as depression, anxiety, post-traumatic stress disorder (PTSD), and addiction [19]. However, due to the novelty of this model, few studies capture its effectiveness in improving health outcomes among older adults. Psychological resilience and mental health outcomes are not uniformly experienced across different populations. Previous evidence suggests that African American and White adults have varying access to social support and health care resources and thereby experience higher levels of stress [22, 23]. Building on this rationale, we hypothesized that CRM would reduce depression, somatic symptoms, perceived stress, and improve resilience among older adults with memory complaints. We further hypothesized that the association of depression, anxiety, and stress can vary by race. Examining these outcomes can help inform more equitable interventions. Methods Study design A pilot intervention study was conducted in 2022 to evaluate stress, mood, somatic symptoms, and resilience before and after the intervention. Convenience sampling was used to recruit fifty-four patients from Emory primary care clinics and Emory Goizueta Alzheimer’s Disease Research Center. Patients who were included in the study if they had a diagnosis of mild cognitive impairment per self-report or medical chart review. Participants with dementia or active psychiatric conditions were excluded from the study. Informed consent was obtained from the patients prior to their participation, and the study was approved by the Emory Institutional Review Board (IRB). Study intervention CRM is a skills-based wellness and prevention program that provides a non-stigmatizing, biological perspective on typical human reactions to stress and trauma [24]. CRM allows one to understand their nervous system and learn to track sensations connected to their well-being [24]. CRM was administered in a group setting of 8–10 participants over Zoom by a certified nurse practitioner trained in CRM methods. Survey questionnaires To evaluate the effects of CRM on stress-related health outcomes, participants completed surveys pre- and post-intervention to assess changes in stress levels, resilience, mental health, and physical health. The survey questionnaires were administered via REDCap, a HIPAA-compliant electronic data capture tool at Emory University [25]. Study measures The survey questionnaires captured four measures of interest, which included the Perceived Stress Scale (PSS-10), Patient Health Questionnaire (PHQ-9), Somatic Symptoms Scale (SSS-8), and the Connor-Davidson Resilience Scale (CD-RISC-10) [26–29]. The Perceived Stress Scale is a 10-item scale that assesses perceived stress levels. It has a score range of 0–40, with higher scores indicating higher levels of perceived stress [27]. The Patient Health Questionnaire is a 9-item scale that assesses the severity of depressive symptoms. It has a score range of 0–27, with higher scores indicating severe depression [29]. The Somatic Symptoms Scale is an 8-item scale that assesses physical health by measuring the perceived burden of common somatic symptoms. It has a score range of 0–32, with higher scores indicating a more severe burden of somatic symptoms [28]. The Connor-Davidson Resilience Scale is a 10-item scale that assesses resilience. It has a score range of 0–40, with higher scores indicating higher resilience [26]. Data analysis SAS programming software was utilized to complete the data analysis. Descriptive statistics were performed to analyze the demographic data of the study population. Normality assumptions were assessed for T-tests. Participants with missing values or lost to follow-up were excluded, and analysis was performed using complete data. Two-sided paired t-tests assessed differences in PSS-10, PHQ-9, SSS-8, and CD-RISC-10 scores pre-intervention and post-intervention. Non-parametric tests were performed to confirm the findings obtained using paired t-tests. Results Of the fifty-four participants recruited for the study, 12 were lost to follow-up and excluded from the analysis, leaving 42 participants (n = 42). The average participant age was 68, with ages ranging from 39 to 83. 81% of the participants were female, with the majority being African Americans (71%), non-Hispanic (98%), and having a college education of 4 years or more (74%) (Table 1). Table 1Demographics-descriptive statistics data (n = 42)FrequencyPercentages (%)GenderMale819Female3481RaceWhite1229African American3071EthnicityHispanic12.4Non-Hispanic4198EducationGrade 12 or GED(High school graduate)49.5College 1–3 years (Some college or technical school)716.7College 4 years or more (College graduate)3173.8 Table 1Demographics-descriptive statistics data (n = 42)FrequencyPercentages (%)GenderMale819Female3481RaceWhite1229African American3071EthnicityHispanic12.4Non-Hispanic4198EducationGrade 12 or GED(High school graduate)49.5College 1–3 years (Some college or technical school)716.7College 4 years or more (College graduate)3173.8 Demographics-descriptive statistics data (n = 42) Grade 12 or GED (High school graduate) Prior to the CRM intervention being implemented, participants presented with moderate levels of perceived stress and mild depression (PSS-10 score: 20.3 ± 4.71; PHQ-9 score: 5.02 ± 4.38) (Table 2). Post-intervention, there was significant improvement in mean differences for depressive symptoms (PHQ-9 mean difference: 1.4 [95% CI 0.50, 2.4]; p-value = 0.0035), somatic symptoms (SSS-8 mean difference: 1.9 [95% CI 0.81, 3.0]; p-value = 0.0011), and resiliency (CD-RISC-10 mean difference: − 1.4 [95% CI − 2.8, − 0.0026]; p-value = 0.050; difference was calculated between pre- and post-intervention scores with negative value reflecting higher post-intervention score) (Table 2). Measurements for perceived stress also showed improvement, but the change was not statistically significant (PSS-10 mean difference: 1.0 [95% CI − 0.77, 2.9]; p-value = 0.25) (Table 2). The non-parametric Wilcoxon Signed-Rank test supported the findings of the paired t-test in the overall sample. On further analyses based on race, the baseline outcome scores were similar. The Wilcoxon Signed-Rank tests corroborated the findings from paired t-tests within each race, and the magnitude of improvement was not statistically significant between races on both Welch t-tests and Mann-Whitney tests. (Table 3 and Supplementary Table S1). Table 2Two-sided paired t-test for pre- versus post-interventionMeasureMean (SD)Mean of differencest-value df p-valueWilcoxon p-valueCohen’s DPSS-101Pre20.36 (4.71)1.051.17410.250.640.25Post19.31 (3.73)PHQ-92Pre5.02 (4.38)1.433.1041 < 0.01* < 0.01* 0.34Post3.60 (4.09)CD-RISC-103Pre29.52 (6.13)− 1.40-2.0241 0.05* 0.05* − 0.22Post30.93 (6.40)SSS-84Pre7.55 (5.79)1.903.5141 < 0.01* < 0.01* 0.35Post5.64 (4.96)Significance was defined as p ≤ 0.05Differences were calculated as pre–post scores; negative values indicate higher post-intervention scoresPSS-10 range: 0–40 (higher = more stress); PHQ-9 range: 0–27 (higher = more depression); SSS-8 range: 0–32 (higher = greater somatic burden); CD-RISC-10 range: 0–40 (higher = greater resilience)1Perceived stress scale with a 10-item questionnaire2Patient health questionnaire with each of the 9 DSM-IV criteria3Connor–Davidson resilience scale with a 10-item questionnaire4Somatic symptom scale with an 8-item questionnaire Table 2Two-sided paired t-test for pre- versus post-interventionMeasureMean (SD)Mean of differencest-value df p-valueWilcoxon p-valueCohen’s DPSS-101Pre20.36 (4.71)1.051.17410.250.640.25Post19.31 (3.73)PHQ-92Pre5.02 (4.38)1.433.1041 < 0.01* < 0.01* 0.34Post3.60 (4.09)CD-RISC-103Pre29.52 (6.13)− 1.40-2.0241 0.05* 0.05* − 0.22Post30.93 (6.40)SSS-84Pre7.55 (5.79)1.903.5141 < 0.01* < 0.01* 0.35Post5.64 (4.96)Significance was defined as p ≤ 0.05Differences were calculated as pre–post scores; negative values indicate higher post-intervention scoresPSS-10 range: 0–40 (higher = more stress); PHQ-9 range: 0–27 (higher = more depression); SSS-8 range: 0–32 (higher = greater somatic burden); CD-RISC-10 range: 0–40 (higher = greater resilience)1Perceived stress scale with a 10-item questionnaire2Patient health questionnaire with each of the 9 DSM-IV criteria3Connor–Davidson resilience scale with a 10-item questionnaire4Somatic symptom scale with an 8-item questionnaire Two-sided paired t-test for pre- versus post-intervention Significance was defined as p ≤ 0.05 Differences were calculated as pre–post scores; negative values indicate higher post-intervention scores PSS-10 range: 0–40 (higher = more stress); PHQ-9 range: 0–27 (higher = more depression); SSS-8 range: 0–32 (higher = greater somatic burden); CD-RISC-10 range: 0–40 (higher = greater resilience) 1Perceived stress scale with a 10-item questionnaire 2Patient health questionnaire with each of the 9 DSM-IV criteria 3Connor–Davidson resilience scale with a 10-item questionnaire 4Somatic symptom scale with an 8-item questionnaire Table 3Two-sided paired t-test stratified by raceMeasureRaceMean (SD)-preMean (SD)-postMean diff (pre–post)Paired t testsWilcoxon p-valueCohen’s DPaired t (df)p-value (t)PSS-10White20.17 (3.51)20.67 (1.78)− 0.5− 0.52 (11)0.620.58− 0.17African American20.43 (5.16)18.77 (4.17)1.671.40 (29)0.170.350.36PHQ-9White6.42 (4.72)3.67 (2.31)2.752.09 (11)0.060.070.71African American4.47 (4.18)3.57 (4.65)0.92.55 (29)0.02* 0.02* 0.20CD-RISC-10White27.92 (5.70)30.58 (6.13)− 2.67− 2.70 (11) 0.02* 0.03* − 0.45African American30.17 (6.26)31.07 (6.60)− 0.9− 1.02 (29)0.320.33− 0.14SSS-8White7.42 (6.40)4.75 (4.41)2.671.71 (11)0.120.260.47African American7.60 (5.65)6.00 (5.19)1.63.55 (29)0.001* 0.002* 0.29Significance was defined as p ≤ 0.05Differences were calculated as pre–post scores; negative values indicate higher post-intervention scoresPSS-10 range: 0–40 (higher = more stress); PHQ-9 range: 0–27 (higher = more depression); SSS-8 range: 0–32 (higher = greater somatic burden); CD-RISC-10 range: 0–40 (higher = greater resilience) Table 3Two-sided paired t-test stratified by raceMeasureRaceMean (SD)-preMean (SD)-postMean diff (pre–post)Paired t testsWilcoxon p-valueCohen’s DPaired t (df)p-value (t)PSS-10White20.17 (3.51)20.67 (1.78)− 0.5− 0.52 (11)0.620.58− 0.17African American20.43 (5.16)18.77 (4.17)1.671.40 (29)0.170.350.36PHQ-9White6.42 (4.72)3.67 (2.31)2.752.09 (11)0.060.070.71African American4.47 (4.18)3.57 (4.65)0.92.55 (29)0.02* 0.02* 0.20CD-RISC-10White27.92 (5.70)30.58 (6.13)− 2.67− 2.70 (11) 0.02* 0.03* − 0.45African American30.17 (6.26)31.07 (6.60)− 0.9− 1.02 (29)0.320.33− 0.14SSS-8White7.42 (6.40)4.75 (4.41)2.671.71 (11)0.120.260.47African American7.60 (5.65)6.00 (5.19)1.63.55 (29)0.001* 0.002* 0.29Significance was defined as p ≤ 0.05Differences were calculated as pre–post scores; negative values indicate higher post-intervention scoresPSS-10 range: 0–40 (higher = more stress); PHQ-9 range: 0–27 (higher = more depression); SSS-8 range: 0–32 (higher = greater somatic burden); CD-RISC-10 range: 0–40 (higher = greater resilience) Two-sided paired t-test stratified by race Significance was defined as p ≤ 0.05 Differences were calculated as pre–post scores; negative values indicate higher post-intervention scores PSS-10 range: 0–40 (higher = more stress); PHQ-9 range: 0–27 (higher = more depression); SSS-8 range: 0–32 (higher = greater somatic burden); CD-RISC-10 range: 0–40 (higher = greater resilience) Discussion Our pilot study on African American and White older adults yielded valuable insight into the impact of an evidence-based behavioral intervention in improving mental, somatic, and resiliency outcomes among patients with complaints of cognitive impairment. When stratified by race, there was no statistically significant difference in the mean values for all measures, indicating that the impact of CRM was consistent across both groups. These findings contribute to our understanding of specific target outcomes and focus areas in optimizing the use of CRM in the community. The delivery of interventions like CRM online via phone or computer is particularly relevant during major events, such as COVID-19, when the elderly face increased isolation. The study analysis revealed a noteworthy improvement in depressive and somatic symptoms among participants who completed the online CRM workshop. This is consistent with previous evidence of a decrease in mental health issues when applying the model. In prior research, Habimana et al. explored the effect of CRM on Tutsi genocide survivors. They determined that 6 months of self-regulation training substantially alleviated PTSD related symptoms, including depression and trauma distress [20]. Furthermore, their study also illustrated a notable reduction in perceived stress levels, with participants learning to recognize their symptoms as normal biological responses to trauma and using CRM skills to regulate triggers and negative responses [20]. In contrast, our study did not demonstrate a statistically significant difference in perceived stress, which can be attributed to characteristics of the study population and a short follow-up period. The study’s findings also align with the evidence from previous research conducted by Grabbe et al., where they implemented a CRM pilot intervention for women in addiction treatment, and post-tests revealed a marked decline in somatic complaints, anger, and anxiety symptoms, with a moderate to large effect size [30]. CRM may thus support self-regulation of emotions by affecting interoceptive processes like arousal, cognitive control, and stress response, which is reflected in the improvements in our study [30, 31].Duva et al. studied the effect of CRM on the mental well-being of healthcare workers during the peak of the COVID-19 pandemic and found that the participants demonstrated improvements in well-being measures, stress levels, and helped in fostering emotional regulation in times of crisis [32]. The findings underscore emotional regulation as one of the foundational pillars on which the resiliency model operates. The results of the current study further revealed a significant mean difference between pre-and post-CRM assessments of the participants’ resiliency, indicating meaningful progress. Resilience is the ability to withstand, recover from, and persist in challenging conditions [20]. The changes in molecular pathways, neurotransmitters, and neural pathways are perpetually influenced by genetic, epigenetic, and environmental effects, which primarily affect an individual’s resilience [33]. These factors are molded by chronic toxic exposure to stress and can eventually lead to prolonged illnesses and an increased risk of premature mortality [33]. As an individual’s resiliency is rooted in these physiological mechanisms, the CRM was developed to motivate and train participants to elevate their ability to navigate stress by regulating the autonomic nervous system [30]. The analysis reinforces evidence from prior studies that have shown improvements in anxiety and PTSD symptoms, and resilience post-intervention [34]. This is the first study focused on the effectiveness of CRM, specifically on African American and White older adults with memory complaints. The results of this study demonstrate the potential of a brief CRM intervention in enhancing mental health and resilience among older adults with memory complaints. The small sample size selected for participation, along with the convenience sampling method employed, are the primary limitations of this study. The loss of 12 participants to follow up may have also reduced statistical power and introduced attrition bias. Additionally, there is a potential for historical bias, as data were collected during the COVID-19 pandemic. This is further evidence that CRM can be deployed well even during challenging times, such as the pandemic. Finally, the self-reported nature of outcomes in the study can lead to response bias, potentially contributing to the limitations. Our future goal is to expand this pilot study via increased and more widespread recruitment efforts to generate a larger and more representative population. Conclusion The Community Resiliency Model (CRM) is a concise program designed to alleviate stress-related health outcomes in individuals. It can be easily scaled within primary care and potentially applied in some post-disaster settings for quick dissemination. Our study determined that CRM was effective in improving depressive and somatic symptoms in older adults who report memory disturbances. Strengthening the ability to identify our internal bodily sensations and increasing awareness of self can help develop the emotional fortitude necessary to respond efficiently to stress, ultimately enhancing resilience. Widespread population-level implementation will enhance individuals’ understanding of trauma and its management by regulating the nervous system. As limited research is available on the use of CRM in health-related outcomes, our study lays the groundwork for further investigation to identify key areas of impact. To build on these preliminary results, larger controlled trials are required to confirm these findings. This will help guide and scale the application of this model in diverse populations, improving overall health outcomes in both community settings and clinical practice. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1