Emotional Stress, Loneliness
Conditions
Keywords
Family Caregivers, Dementia
Brief summary
The prevalence of dementia is higher in Hispanics and African Americans than non-Hispanic Whites. Moreover, dementia caregivers often experience loneliness as well decreased health status. The expansion of social media use among Hispanics and African Americans, particularly Twitter - a short message service - offers great promise for improving social support. This study aims to evaluate changes of discussion topics, sentiment and networking styles (i.e., number of followers) among anonymous followers of our two Twitter networks; the African American/Black dementia caregiver group and the Hispanic dementia caregiver group.
Detailed description
The study will utilize Twitter networks to post a daily message for dementia caregivers for a year, and set up a monthly group chat.
Interventions
BEHAVIORALTwitter for Hispanic caregivers
This group will be asked to follow and use (i.e., retweet, reply, like) our Hispanic Twitter network for social support
BEHAVIORALTwitter for African American caregivers
This group will be asked to follow and use (i.e., retweet, reply, like) our African American Twitter network for social support.
Sponsors
Department of Health and Human Services
National Institute on Aging (NIA)
Columbia University
Study design
Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
SUPPORTIVE_CARE
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
Yes
Inclusion criteria
* 18 years of age or older * Black or Hispanic, living in the U.S. including the U.S. territories * a dementia caregiver with any duration, able to speak English or Spanish/bilingual * must agree to terms of conditions of use and privacy policy and rules of one of the two dementia caregiver network (Hispanic @dcnh, Black @dcnaab), the Twitter user agreement of the terms of service, Twitter privacy policy and Twitter rules including intellectual property, violence, misconduct, abuse behavior, private information and spam and security * use a smartphone or a feature phone (i.e., a cell phone with text messaging)
Exclusion criteria
* do not have de-identified Twitter account, children, not a dementia family caregiver
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Number of Participants Who do Not Engage in Twitter Activities | 12 months | Macro level: The number of participants who do not engage in Twitter activities (i.e., retweet, reply, like, post) within the Twitter network for dementia caregivers |
| Emotional Valence Score | 12 months | Emotional valence (macro level) detected from text data (e.g., This is so helpful: emotion score +4; I am sad: emotion score -6). Emotional valence score ranges from -10 to +10; -10 indicates negative valence and +10 indicates positive valence. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Number of Auto-detected Small Groups | 12 months | Meso level: number of small group automatically detected based on Twitter activities (i.e., retweet); for example 0 means no small-group activity detected, 7 small groups mean seven distinct cliques/grouping activities detected where the users are communicating, supporting and responding within our Twitter network. |
| Percentage of Balanced Triads | 12 months | Micro level: % balanced communication triad, 2 or 3 people responding and communicating with equal or similar frequency; 0 reflects non-optimal communication indicating domination, avoidance or ignorance. 8 means there are 8% triads who maintain healthy communication. |
Countries
United States
Participant flow
Recruitment details
Participants were recruited based on flyer distribution via community outreach offline (10 brain and Alzheimer's disease community events, community outreach) and online (Twitter advertisement, emails to 920+ community leaders) between January 2022 and June 2023. The first participant was enrolled on January 12, 2022, and the last was enrolled on June 14, 2023.
Participants by arm
| Arm | Count |
|---|---|
| African American Family Members of Persons With Alzheimer's Disease and Related Dementias African American caregiving Twitter community received a multimedia Tweet message daily for 365 days and was asked to use (i.e., retweet, reply, like) our network community; Group message started on November 1, 2022 until October 31, 2023.
Ingredient: an evidence-based social support Tweet message tailored to African American family members of persons with ADRD \[dosage: 1 group message, dosage form: video and text, frequency: 1 group message daily, administration duration: a short video 20-50 seconds, a group message length 280 characters reading time 3 seconds\] | 506 |
| Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias Hispanic caregiving Twitter community received a multimedia Tweet message daily for 365 days and was asked to use (i.e., retweet, reply, like) our network community; Group message started on November 1, 2022 until October 31, 2023.
Ingredient: an evidence-based social support Tweet message tailored to Hispanic family members of persons with ADRD \[dosage: 1 group message, dosage form: video and text, frequency: 1 group message daily, administration duration: a short video 20-50 seconds, a group message length 280 characters reading time 3 seconds\] | 460 |
| Total | 966 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Lack of Efficacy | 12 | 11 |
| Overall Study | Lost to Follow-up | 125 | 133 |
| Overall Study | Withdrawal by Subject | 2 | 4 |
Baseline characteristics
| Characteristic | Total | Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias | African American Family Members of Persons With Alzheimer's Disease and Related Dementias |
|---|---|---|---|
| Age, Continuous | 47.9 years STANDARD_DEVIATION 15.7 | 49.1 years STANDARD_DEVIATION 15.6 | 46.7 years STANDARD_DEVIATION 15.6 |
| Emotional Valence Score | 0.113 score on a scale STANDARD_DEVIATION 0.908 | 0.073 score on a scale STANDARD_DEVIATION 0.713 | 0.149 score on a scale STANDARD_DEVIATION 1.054 |
| Ethnicity (NIH/OMB) Hispanic or Latino | 707 Participants | 460 Participants | 247 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 255 Participants | 0 Participants | 255 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 4 Participants | 0 Participants | 4 Participants |
| Number of participants who do not engage in Twitter activities | 918 Participants | 447 Participants | 471 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 18 Participants | 18 Participants | 0 Participants |
| Race (NIH/OMB) Asian | 3 Participants | 3 Participants | 0 Participants |
| Race (NIH/OMB) Black or African American | 456 Participants | 57 Participants | 399 Participants |
| Race (NIH/OMB) More than one race | 424 Participants | 317 Participants | 107 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 2 Participants | 2 Participants | 0 Participants |
| Race (NIH/OMB) White | 63 Participants | 63 Participants | 0 Participants |
| Region of Enrollment United States | 966 participants | 460 participants | 506 participants |
| Sex/Gender, Customized Female | 646 Participants | 313 Participants | 333 Participants |
| Sex/Gender, Customized Male | 317 Participants | 145 Participants | 172 Participants |
| Sex/Gender, Customized Prefer not to answer | 3 Participants | 2 Participants | 1 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 506 | 0 / 460 |
| other Total, other adverse events | 0 / 506 | 0 / 460 |
| serious Total, serious adverse events | 0 / 506 | 0 / 460 |
Outcome results
Primary
Emotional Valence Score
Emotional valence (macro level) detected from text data (e.g., This is so helpful: emotion score +4; I am sad: emotion score -6). Emotional valence score ranges from -10 to +10; -10 indicates negative valence and +10 indicates positive valence.
Time frame: 12 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| African American Family Members of Persons With Alzheimer's Disease and Related Dementias | Emotional Valence Score | 1.404 score on a scale | Standard Deviation 2.412 |
| Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias | Emotional Valence Score | 1.483 score on a scale | Standard Deviation 2.229 |
Comparison: Hypothesis: There will be significant pre/post-intervention differences in emotional valence score detected from Tweets for African American dementia caregiver networks.p-value: <0.000195% CI: [0.513, 0.742]a paired t-test
Comparison: Hypothesis: There will be significant pre/post-intervention differences in emotional valence score detected from Tweets for Hispanic dementia caregiver networks.p-value: <0.000195% CI: [1.189, 1.631]a paired t-test
Comparison: There will be no significant differences in emotional valence score detected from Tweets between Hispanic and African American dementia caregiver networks.p-value: 0.695% CI: [-0.186, 0.107]t-test, 2 sided
Primary
Number of Participants Who do Not Engage in Twitter Activities
Macro level: The number of participants who do not engage in Twitter activities (i.e., retweet, reply, like, post) within the Twitter network for dementia caregivers
Time frame: 12 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| African American Family Members of Persons With Alzheimer's Disease and Related Dementias | Number of Participants Who do Not Engage in Twitter Activities | 246 Participants |
| Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias | Number of Participants Who do Not Engage in Twitter Activities | 226 Participants |
Comparison: Hypothesis: There will be significant pre/post-intervention differences in isolated macro-level network structures for African American dementia caregiver networks.p-value: <0.000195% CI: [4.919, 10.323]McNemar
Comparison: Hypothesis: There will be significant pre/post-intervention differences in isolated macro-level network structures for Hispanic dementia caregiver networks.p-value: <0.000195% CI: [9.963, 33.157]McNemar
Comparison: There will be no significant differences in isolated macro-level network structure from Tweets between Hispanic and African American dementia caregiver networks.p-value: 0.92495% CI: [0.787, 1.325]Chi-squared, Corrected
Secondary
Number of Auto-detected Small Groups
Meso level: number of small group automatically detected based on Twitter activities (i.e., retweet); for example 0 means no small-group activity detected, 7 small groups mean seven distinct cliques/grouping activities detected where the users are communicating, supporting and responding within our Twitter network.
Time frame: 12 months
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| African American Family Members of Persons With Alzheimer's Disease and Related Dementias | Number of Auto-detected Small Groups | Pre | 259 Auto-detected Small Groups |
| African American Family Members of Persons With Alzheimer's Disease and Related Dementias | Number of Auto-detected Small Groups | Post | 262 Auto-detected Small Groups |
| Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias | Number of Auto-detected Small Groups | Pre | 231 Auto-detected Small Groups |
| Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias | Number of Auto-detected Small Groups | Post | 239 Auto-detected Small Groups |
Secondary
Percentage of Balanced Triads
Micro level: % balanced communication triad, 2 or 3 people responding and communicating with equal or similar frequency; 0 reflects non-optimal communication indicating domination, avoidance or ignorance. 8 means there are 8% triads who maintain healthy communication.
Time frame: 12 months
Population: Triad census: Counts of each triad type within a network are determined by analyzing all possible combinations of connections of 16 possible communication styles among three nodes (a triad) in each conversation.
| Arm | Measure | Group | Value (NUMBER) |
|---|---|---|---|
| African American Family Members of Persons With Alzheimer's Disease and Related Dementias | Percentage of Balanced Triads | Pre | 0.00 percentage of balanced triads |
| African American Family Members of Persons With Alzheimer's Disease and Related Dementias | Percentage of Balanced Triads | Post | 0.21 percentage of balanced triads |
| Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias | Percentage of Balanced Triads | Pre | 0.00 percentage of balanced triads |
| Hispanic Family Members of Persons With Alzheimer's Disease and Related Dementias | Percentage of Balanced Triads | Post | 0.34 percentage of balanced triads |