The geographic digital divide among dementia caregivers in the U.S.: a cross-sectional study Authors: Jingjing Gao, Eric C. Jones, Louis D. Brown, Benjamin J. Radford, Jiaming Liang, Jack Tsai Published: 2026 ABSTRACT BackgroundDigital health tools, particularly patient portals, can support caregiving, but there is limited understanding of how sociodemographic and geographic factors influence digital health engagement among U.S. caregivers, especially dementia caregivers. Dementia caregiving involves complex coordination, medication management, and frequent healthcare interactions. Therefore, access to online medical records is particularly valuable. Moreover, most previous research focuses on general digital health adoption or patient portal use among patients, not caregivers, particularly among dementia caregivers, and rarely considers geographic differences or the interaction effects between age and caregivers’ self-rated health. Comparing dementia caregivers to those without is essential because dementia caregiving entails more extensive care, longer duration, behavioral challenges, and emotional strain. These factors may increase reliance on digital health tools like online medical records. However, national studies often do not distinguish between dementia and non-dementia caregiving in digital health analyses.ObjectiveThis study explored how sociodemographic and geographic factors are associated with caregivers’ frequency of accessing care recipients’ online medical records, especially for caregivers of individuals living with dementia compared to those caregivers of individuals without dementia. We also evaluated the interaction effect between age and health condition, motivated by the hypothesis that digital engagement is influenced not only by age but also by functional health status.MethodsThis study examined caregivers’ portal access frequency using data from the 2022 National Health Interview Survey, collected between March 7th and November 8th, 2022. Descriptive statistics captured differences in access frequency and sociodemographic characteristics between dementia caregivers and non-dementia caregivers. Three ordered logistic regression models examined predictors of accessing online medical records for dementia caregivers, non-dementia caregivers, and all caregivers, followed by an all-caregiver model with interaction terms to assess moderation effects between caregiver age and self-reported health. ArcGIS Pro was adopted to visually capture the geographic regional divide in portal access frequency across the U.S.ResultsOf the 6,252 total responses, 916 caregivers provided data on portal access frequency. More than half of dementia caregivers (60.25%) and non-dementia caregivers (51.11%) did not access online medical records in the past 12 months. Frequent access (≥ 10 times/year) was low across both groups (dementia caregivers (7.03%) vs. non-dementia caregivers (12.66%)). Survey-weighted ordered logistic regression models identified significant sociodemographic and geographic disparities in portal access frequency. Being female (OR = 1.931, SE: 0.074), having a bachelor’s degree (OR = 1.769, SE: 0.168), and being married (OR = 1.749, SE: 0.068) were associated with higher portal access frequency in the all-caregivers model, and similar results were found in dementia caregivers and non-dementia caregivers. Regarding geographic disparities, dementia caregivers residing in micropolitan areas had substantially lower odds of frequent portal access than those in metropolitan areas (OR = 0.633, SE: 0.071); a similar trend was observed among dementia and non-dementia caregivers. Rural residence was also associated with lower portal access among dementia caregivers (OR = 0.529, SE: 0.152) and non-dementia caregivers (OR = 0.654, SE: 0.097). In the all-caregiver sample, caregivers in small-town areas (OR = 0.522, SE: 0.071) had lower odds of portal access compared to metropolitan caregivers. Health status showed differential patterns across caregiving groups. Among dementia caregivers, reporting very good health (vs. excellent) was associated with lower odds of portal access (OR = 0.268, SE: 0.023), whereas among non-dementia caregivers, very good health was associated with higher portal access (OR = 1.246, SE: 0.062). Good or fair health was associated with lower portal access across all caregiver groups. Interaction analysis revealed that as caregiver age increased, those in poor health had a significantly lower portal access frequency.ConclusionsAccessing online medical records among family caregivers remains limited and varies notably by gender, educational attainment, geographic location, and the interaction between age and health status. These findings highlight the importance of developing tailored digital support strategies that address the intersecting socioeconomic and health-related barriers faced by caregivers, particularly as artificial intelligence-enabled tools become more prevalent in dementia care. Older caregivers in poor health are especially unlikely to adopt digital resources, underscoring the urgent need for targeted outreach and educational initiatives. Additionally, the persistent geographic digital divide among dementia caregivers, especially those in small towns, calls for increased support to promote equitable access to caregiving technologies.Trial registrationThis study conducted a secondary data analysis, so trial registration is NOT required because it does not run a prospective randomized trial. And Institutional Review Board approval is also not needed because the dataset is de-identified and public. Introduction Dementia is an umbrella term encompassing a range of neurocognitive disorders, of which Alzheimer’s disease (AD) is the most common subtype. While AD accounts for the majority of dementia cases, caregiving challenges and health system interactions extend across all dementia types. Alzheimer’s disease (AD) is characterized by progressive cognitive decline, including impairments in memory and spatial cognition [1]. Dementia represents a progressive health condition that places a significant burden on disease management and caregiving [2–15]. Over the past decades, most clinical trials aimed at treating dementia have faced great challenges in creating breakthroughs, which has left caregivers increasingly desperate to find alternative ways to reduce the substantial caring burdens and costs associated with the disease [15–18]. People living with dementia often receive care from family members and friends, and in this article, we refer to them collectively as family caregivers to emphasize their unpaid caregiving services to their loved ones [19]. Dementia affects approximately 55 million people worldwide [20]. In the United States alone, more than 11 million people provide unpaid care for people living with dementia, contributing an estimated $271.6 billion annually in caregiving costs (2021 U.S. dollars) as reported in the Alzheimer’s Association 2022 - Alzheimer’s Disease Facts and Figures report, and these costs are projected to increase substantially in the coming decades due to population aging [21]. Family dementia caregivers are at an elevated risk for stress, burden, anxiety, and depression compared to caregivers of individuals with other conditions [19, 22]. Additionally, these caregivers face greater challenges with work, housing, financial stability, and personal relationships [23]. These challenges to caregivers can be mitigated by the utilization of education, community resources, therapy, and counseling services [24]. However, geographical barriers, transportation barriers, and time limitations often hinder caregivers from accessing such facilities [24]. Additionally, the high demand for care can result in making the commitment to counseling or therapy challenging, especially with in-person health services [25]. In this study, we focus specifically on access to online medical records, typically delivered through patient portals, which allow caregivers to view clinical information, manage appointments, track medications, and communicate with healthcare providers. Access to online medical records is particularly critical for dementia caregivers, as caregiving often involves coordinating care across multiple providers, managing complex medication regimens, and monitoring health changes over time. Without access to patient portals, caregivers may experience delays in obtaining clinical information, increased administrative burden, and reduced ability to effectively support care recipients. These challenges may be amplified for caregivers experiencing geographic or socioeconomic barriers, contributing to disparities in digital health engagement. Patient portals offer advantages such as real-time behavioral data collection to inform care strategies [26]. Patient portals offer numerous benefits to people living with dementia, such as: (1) decreasing the inequalities in access to medical care due to age, geography, and ethnicity [27, 28]; (2) improving life quality and cognitive function [29–31]; (3) improving independence among individuals with early-stage dementia and enhancing caregivers’ confidence [32–35]; and (4) improving the mental health of people living with dementia [36]. Numerous studies reveal that internet-based interventions are effective for both caregivers and patients with dementia [19, 37]. These interventions could encompass remote diagnosis, monitoring, evaluation, and treatment, which can result in lessening financial burden, reducing travel time, and overcoming geographical barriers when physical presence is not feasible [37, 38]. However, relatively little attention has been given to caregivers’ access to online medical records, which represents a distinct and functionally critical form of digital health engagement. Caregivers’ sociodemographic characteristics play a critical role in shaping digital health engagement. Studies have shown that the average age of dementia caregivers ranges from 51 to 67 years old, with a significant proportion being children, followed by spouses, then other relatives [19, 37–40]. Furthermore, the educational background of caregivers varies widely, with 66.9% of caregivers having an education below college level [19, 39–41]. Also, the gender distribution among caregivers shows the prevalence of female caregivers is higher [19, 39, 40]. However, the adoption of patient portals varies based on a variety of sociodemographic characteristics such as age, gender, race, education, and income [42, 43]. For example, racial minority caregivers considered the need for patient portals to help with basic activities of daily living as more important in assisting patients with mild dementia than did non-minorities [42]. Additionally, females were more technology savvy than males, while males were more willing to pay for expensive technologies than females [43]. Meanwhile, caregivers face challenges related to patient portals, unfamiliarity with new technology, dementia-specific symptoms that complicate use, as well as a lack of personalization and support for some tools [44–48]. Furthermore, family caregivers face additional challenges, including a lack of formal training, differences in health literacy, and limited resources, which makes a comprehensive study of the adoption of patient portals and other digital health tools among family dementia caregivers in the U.S. even more urgent. Addressing these disparities requires targeted interventions that consider the socioeconomic context of caregivers, ultimately aiming to enhance the quality of life for individuals affected by dementia. Despite growing attention to digital health disparities, limited research has specifically examined disparities in caregivers’ access to online medical records, particularly among dementia caregivers. Access to patient portals represents a distinct and functionally important form of digital health engagement, enabling caregivers to coordinate care, manage medications, and communicate with providers. However, disparities in access to these systems remain poorly understood, especially across different sociodemographic and geographic groups. To address these gaps, it is important to examine how sociodemographic and geographic factors influence caregivers’ access to online medical records. This study assesses whether sociodemographic characteristics of family caregivers (e.g., age, gender, income, education, and marital status), their health conditions, and geographic location are associated with accessing online medical records. Guided by digital divide theory, we hypothesized that caregivers’ sociodemographic and geographic characteristics would be associated with the frequency of accessing care recipients’ online medical records. Specifically, we expected that younger age, higher educational attainment, higher household income, better self-rated health, and metropolitan residence would be positively associated with higher odds of more frequent portal access. Thus, this study builds on prior work by: Reinforcing known sociodemographic disparities in accessing online medical records;Highlighting the geographic digital divide among dementia caregivers specifically;Introducing evidence on how caregiver health and age interact to create compounding barriers to digital engagement. Reinforcing known sociodemographic disparities in accessing online medical records; Reinforcing known sociodemographic disparities in accessing online medical records; Highlighting the geographic digital divide among dementia caregivers specifically; Highlighting the geographic digital divide among dementia caregivers specifically; Introducing evidence on how caregiver health and age interact to create compounding barriers to digital engagement. Introducing evidence on how caregiver health and age interact to create compounding barriers to digital engagement. Methods Data source This study focuses on the caregivers’ use of digital technology, specifically tracking their care recipients’ online medical records in the past 12 months. The analysis uses data from the 2022 National Health Interview Survey [49], which is a nationally representative survey targeted at civilian, non-institutionalized adults aged 18 or older living in the United States. This survey was conducted from March 7th to November 8th, 2022, and a two-stage design was adopted to recruit respondents who were provided two options (on the web or by paper) to complete the survey. The dataset included key sociodemographic factors and health-related outcomes. The National Health Interview Survey (NHIS) employs a multistage, stratified cluster sampling design to produce nationally representative estimates of the noninstitutionalized U.S. population. Primary sampling units (PSUs) are selected within strata defined by geographic and demographic characteristics and then households within selected PSUs are sampled, and one or more eligible adults are randomly selected for interview. All analyses incorporated NHIS sampling weights to account for unequal probabilities of selection and nonresponse. Variance estimation used balanced repeated replication (BRR) with replicate weights, consistent with NHIS analytic guidelines, to appropriately account for the complex survey design. Ethics statement This study involved secondary data analysis using publicly available, de-identified data from the 2022 National Health Interview Survey (NHIS). The NHIS is administered by the National Center for Health Statistics (NCHS), Centers for Disease Control and Prevention (CDC), and data collection was conducted in compliance with federal regulations governing human subjects research. Because the dataset is fully de-identified and contains no personally identifiable information, this study did not require Institutional Review Board (IRB) approval, in accordance with the U.S. Department of Health and Human Services regulations at 45 CFR 46.104(d)(4) for research involving secondary analysis of existing data. No direct contact with participants occurred, and informed consent was not required. Study population This study included adult participants aged 18 and above. The final analytic sample only included participants with complete data on race/ethnicity, education, income, employment, marital status, geographic location, and self-reported subjective overall health. The total responses were 6,252, while 6,185 of the surveys were considered to be complete. Caregiving status was identified using the NHIS Family Caregiving supplement, which asks respondents whether they provided regular care or assistance to a family member or friend with a health condition or disability in the past 12 months. Respondents who answered affirmatively were classified as caregivers. Among caregivers, those providing care to individuals reported as having Alzheimer’s disease or a related neurocognitive condition were classified as dementia caregivers. However, only 916 respondents who were caregivers answered the question about their frequency of accessing their care recipients’ online medical records. Of the 916 caregivers surveyed, 300 reported that they were currently providing the most care to individuals with conditions such as Alzheimer’s disease, confusion, dementia, forgetfulness, brain injury, stroke, or other neurological issues. Due to the lack of detailed information distinguishing among these neurological conditions, this study broadly classifies these respondents as dementia caregivers. Measures The outcome, “portal access frequency,” was measured using a five-category ordinal variable capturing the number of times caregivers accessed care recipients’ online medical records in the past 12 months (never, once, 2–3 times, 4–9 times, ≥ 10 times). The variable was treated as an ordered outcome in survey-weighted ordered logistic regression models. Information about participants’ sociodemographic characteristics was also collected, including (1) birth gender, which is participants’ self-reported sex assigned at birth. The National Health Interview Survey categorizes this as male or female, based on binary biological sex rather than gender identity; (2) age recorded in years at the time of the interview; (3) race coded as one of the categories: “Non-Hispanic White”, “Non-Hispanic Black or African American”, “Hispanic”, “Non-Hispanic Asian”, “Non-Hispanic Other”; (4) education measured as the highest degree or level of school completed; (5) income reported in income ranges; (6) employment coded as “Yes” and “No” to indicate whether the participants were employed or not; and (7) marriage status coded as “Yes” and “No” to represent whether the participants were married or not. Rural-urban community classification used the Rural-Urban Community Area Code. Participants were classified into one of four geographic categories based on their residential location, using the National Health Interview Survey-defined Rural-Urban Community Area Code system: (1) Metropolitan: Areas with high population density and access to urban infrastructure and services; (2) Micropolitan: Smaller urban areas centered around towns with populations of at least 10,000 but fewer than 50,000 people; (3) Small Town: Communities outside of urbanized areas with moderate population concentrations; (4) Rural: Low-density areas located outside of urban and small-town classifications, often with limited access to healthcare and other resources. This classification captures meaningful differences in community size and access to services, and is commonly used in public health research to assess urban-rural disparities. To further capture the variation in caregiver portal access frequency patterns across different geographic regions across the United States, a spatial map (Fig. 1) was created using ArcGIS Pro to visually show the “Non-Use of Online Medical Records” and “Frequent Use of Online Medical Records”. For Non-Use of Online Medical Records, Caregivers were asked whether they had ever accessed the care recipient’s online medical records; and the percentage of respondents who reported never accessing these records is represented in blue on the map and highlights low technology engagement. For “Frequent Use of Online Medical Records”, it was measured by the percentage of caregivers who reported accessing online medical records more than 10 times in the past year. This metric indicates high-frequency use, reflecting regular engagement with patient health information systems. It is shown in red on Fig. 1. Furthermore, this study collected the subjective overall health of each participant with a self-reported general health status question measured using the standard single-item question: “Would you say your health in general is…” Participants responded using a 5-point Likert scale: (1) Excellent; (2) Very Good; (3) Good; (4) Fair; (5) Poor. This subjective assessment of overall health is a widely accepted indicator of population health and is predictive of morbidity, healthcare utilization, and mortality [50]. Statistical analysis Descriptive statistics are adopted to analyze differences in access frequency and sociodemographic characteristics between the general caregivers and dementia caregivers. This study uses Ordered Logistic Regression as “caregivers’ access to patients’ online medical record” is treated as ordinal categories. This regression was widely used in social science, including public health science [51, 52]. The ordered logistic regression model is preferred if “a dependent variable has more than two categories and the values of each category have a meaningful sequential order where a value is indeed ‘higher’ than the previous one” [53–55]. All statistical analyses were performed using Stata/SE 18.5 (StataCorp, College Station, TX) [56]. ArcGIS Pro was used to visually capture the geographic regional divide in accessing online medical records across the U.S. The outcome variable was caregivers’ frequency of accessing care recipients’ online medical records in the past 12 months, modeled as a five-category ordinal variable (never, once, 2–3 times, 4–9 times, ≥ 10 times). Survey-weighted ordered logistic regression models were estimated to examine associations between caregiver characteristics and portal access frequency. Independent variables included caregiver gender, age (continuous), race/ethnicity (Non-Hispanic White [reference], Non-Hispanic Black, Hispanic, Non-Hispanic Asian, Non-Hispanic Other), educational attainment (< high school [reference], high school graduate, some college, bachelor’s degree, post-baccalaureate degree), household income (continuous), employment status, marital status, rural–urban residence (metropolitan [reference], micropolitan, small town, rural), and self-rated health (excellent [reference], very good, good, fair, poor). Separate models were estimated for (1) dementia caregivers, (2) caregivers of individuals without dementia, and (3) the pooled caregiver sample. All analyses incorporated NHIS sampling weights and balanced repeated replication (BRR) variance estimation to account for the complex survey design. Multicollinearity among independent variables was assessed using variance inflation factors (VIFs), and no evidence of problematic collinearity was observed. Statistical significance was evaluated at the α = 0.05 level. Results Descriptive analysis of accessing online medical records among family dementia caregivers in the U.S. This analysis explored the use of patient portals, specifically online medical records, among family dementia caregivers in the United States. The comparison is made between dementia caregivers (n = 293) and non-dementia caregivers (n = 608). Several demographic variables, including gender, age, race, education, income, employment, marital status, community area, and self-reported health conditions, are presented. The detailed information is presented in Table 1B. Accessing online medical records Family caregivers’ engagement with online medical records varies. Among non-dementia caregivers, a majority (51.11%) reported not accessing the patient’s online medical records in the past 12 months. For dementia caregivers, this number is significantly higher at 60.25%. The proportion of caregivers accessing records 10 or more times in the past year was about 7.03% among dementia caregivers and 12.66% among non-dementia caregivers, suggesting a modest engagement with digital tools for managing medical information. Comparison of demographic factors between dementia caregivers and non-dementia caregivers Gender: Both caregiver groups had a higher proportion of female caregivers. Among non-dementia caregivers, 54.32% are female, and the dementia group shows 67.39% female caregivers. This reflects the broader trend of female-dominated caregiving, which is particularly pronounced in dementia care. Age: The mean age of dementia caregivers is higher, with an average of 57.02 years compared to 48.49 years among non-dementia caregivers. This difference may reflect the fact that dementia typically affects older individuals, requiring care from older caregivers, many of whom may be family members. Race and Ethnicity: The racial composition of dementia caregivers is somewhat different from that of the non-dementia caregiver group. The dementia caregivers have a higher proportion of non-Hispanic White caregivers (63.54%) compared to the non-dementia caregiver group (56.76%). Additionally, non-Hispanic Black or African American caregivers account for 12.07% of dementia caregivers, compared to 10.07% in the non-dementia caregiver population. Education: Education levels are relatively similar between the two groups, with both groups showing the highest proportion of caregivers who have attended some college (44.27% for non-dementia caregivers and 41.48% for dementia caregivers). Income: Income levels among caregivers show that 37.80% of non-dementia caregivers and 39.01% of dementia caregivers earn $75,000 or more. Employment Status: A notable difference is observed in the employment status of caregivers. The proportion of caregivers of individuals with Dementia not employed was 41.65% compared to the non-dementia caregiver group at 46.30%. Marital Status: Married caregivers make up a similar proportion of the dementia caregiver group (69.55%) compared to the non-dementia caregiver group (69.55%).Community Area: Most caregivers across both groups reside in metropolitan areas, with 82.91% of non-dementia caregivers and 85.61% of dementia caregivers reporting a metropolitan living location. Figure 1 presents the regional distribution of caregivers’ access to patients’ online medical records across the nine U.S. Census Regions. Two key patterns were analyzed: the percentage of caregivers who never accessed online medical records (shown in blue bar), and the percentage who accessed them more than 10 times (shown in red bar). Regional Differences in Non-Use (Non-use was operationalized as reporting no access to care recipients’ online medical records during the preceding 12 months, consistent with the NHIS survey question): The highest percentage of caregivers who never accessed online medical records was observed in the Middle Atlantic (79.49%), followed by the East North Central (70.59%), and Mountain region (66.67%). In contrast, the regions of West South Central and New England had the lowest non-use rates at 46.34% and 46.15%, respectively. Frequent Use Patterns: Frequent access (defined as more than 10 times) was highest in the West South Central (17.07%) and Pacific (15.91%) regions. In contrast, several regions, including West North Central, reported 0% of caregivers accessing records more than 10 times. Balanced Engagement: The East North Central and East South-Central regions reported high variation in digital engagement within each region. For instance, while 70.59% of caregivers in the East North Central region never accessed records, 8.82% reported high-frequency access. A similar pattern was seen in the East South-Central region (55.88% non-use vs. 8.82% high use). Gender: Both caregiver groups had a higher proportion of female caregivers. Among non-dementia caregivers, 54.32% are female, and the dementia group shows 67.39% female caregivers. This reflects the broader trend of female-dominated caregiving, which is particularly pronounced in dementia care. Age: The mean age of dementia caregivers is higher, with an average of 57.02 years compared to 48.49 years among non-dementia caregivers. This difference may reflect the fact that dementia typically affects older individuals, requiring care from older caregivers, many of whom may be family members. Race and Ethnicity: The racial composition of dementia caregivers is somewhat different from that of the non-dementia caregiver group. The dementia caregivers have a higher proportion of non-Hispanic White caregivers (63.54%) compared to the non-dementia caregiver group (56.76%). Additionally, non-Hispanic Black or African American caregivers account for 12.07% of dementia caregivers, compared to 10.07% in the non-dementia caregiver population. Education: Education levels are relatively similar between the two groups, with both groups showing the highest proportion of caregivers who have attended some college (44.27% for non-dementia caregivers and 41.48% for dementia caregivers). Income: Income levels among caregivers show that 37.80% of non-dementia caregivers and 39.01% of dementia caregivers earn $75,000 or more. Employment Status: A notable difference is observed in the employment status of caregivers. The proportion of caregivers of individuals with Dementia not employed was 41.65% compared to the non-dementia caregiver group at 46.30%. Marital Status: Married caregivers make up a similar proportion of the dementia caregiver group (69.55%) compared to the non-dementia caregiver group (69.55%). Gender: Both caregiver groups had a higher proportion of female caregivers. Among non-dementia caregivers, 54.32% are female, and the dementia group shows 67.39% female caregivers. This reflects the broader trend of female-dominated caregiving, which is particularly pronounced in dementia care. Age: The mean age of dementia caregivers is higher, with an average of 57.02 years compared to 48.49 years among non-dementia caregivers. This difference may reflect the fact that dementia typically affects older individuals, requiring care from older caregivers, many of whom may be family members. Race and Ethnicity: The racial composition of dementia caregivers is somewhat different from that of the non-dementia caregiver group. The dementia caregivers have a higher proportion of non-Hispanic White caregivers (63.54%) compared to the non-dementia caregiver group (56.76%). Additionally, non-Hispanic Black or African American caregivers account for 12.07% of dementia caregivers, compared to 10.07% in the non-dementia caregiver population. Education: Education levels are relatively similar between the two groups, with both groups showing the highest proportion of caregivers who have attended some college (44.27% for non-dementia caregivers and 41.48% for dementia caregivers). Income: Income levels among caregivers show that 37.80% of non-dementia caregivers and 39.01% of dementia caregivers earn $75,000 or more. Employment Status: A notable difference is observed in the employment status of caregivers. The proportion of caregivers of individuals with Dementia not employed was 41.65% compared to the non-dementia caregiver group at 46.30%. Marital Status: Married caregivers make up a similar proportion of the dementia caregiver group (69.55%) compared to the non-dementia caregiver group (69.55%). Community Area: Most caregivers across both groups reside in metropolitan areas, with 82.91% of non-dementia caregivers and 85.61% of dementia caregivers reporting a metropolitan living location. Figure 1 presents the regional distribution of caregivers’ access to patients’ online medical records across the nine U.S. Census Regions. Two key patterns were analyzed: the percentage of caregivers who never accessed online medical records (shown in blue bar), and the percentage who accessed them more than 10 times (shown in red bar). Regional Differences in Non-Use (Non-use was operationalized as reporting no access to care recipients’ online medical records during the preceding 12 months, consistent with the NHIS survey question): The highest percentage of caregivers who never accessed online medical records was observed in the Middle Atlantic (79.49%), followed by the East North Central (70.59%), and Mountain region (66.67%). In contrast, the regions of West South Central and New England had the lowest non-use rates at 46.34% and 46.15%, respectively. Frequent Use Patterns: Frequent access (defined as more than 10 times) was highest in the West South Central (17.07%) and Pacific (15.91%) regions. In contrast, several regions, including West North Central, reported 0% of caregivers accessing records more than 10 times. Balanced Engagement: The East North Central and East South-Central regions reported high variation in digital engagement within each region. For instance, while 70.59% of caregivers in the East North Central region never accessed records, 8.82% reported high-frequency access. A similar pattern was seen in the East South-Central region (55.88% non-use vs. 8.82% high use). Community Area: Most caregivers across both groups reside in metropolitan areas, with 82.91% of non-dementia caregivers and 85.61% of dementia caregivers reporting a metropolitan living location. Figure 1 presents the regional distribution of caregivers’ access to patients’ online medical records across the nine U.S. Census Regions. Two key patterns were analyzed: the percentage of caregivers who never accessed online medical records (shown in blue bar), and the percentage who accessed them more than 10 times (shown in red bar). Regional Differences in Non-Use (Non-use was operationalized as reporting no access to care recipients’ online medical records during the preceding 12 months, consistent with the NHIS survey question): The highest percentage of caregivers who never accessed online medical records was observed in the Middle Atlantic (79.49%), followed by the East North Central (70.59%), and Mountain region (66.67%). In contrast, the regions of West South Central and New England had the lowest non-use rates at 46.34% and 46.15%, respectively. Frequent Use Patterns: Frequent access (defined as more than 10 times) was highest in the West South Central (17.07%) and Pacific (15.91%) regions. In contrast, several regions, including West North Central, reported 0% of caregivers accessing records more than 10 times. Balanced Engagement: The East North Central and East South-Central regions reported high variation in digital engagement within each region. For instance, while 70.59% of caregivers in the East North Central region never accessed records, 8.82% reported high-frequency access. A similar pattern was seen in the East South-Central region (55.88% non-use vs. 8.82% high use). Fig. 1Spatial distribution of portal access frequency among dementia caregivers across the United States. Note: this map is based on the data collected from the 2022 National Health Interview Survey [49]. For detailed information, see Table 1A below Spatial distribution of portal access frequency among dementia caregivers across the United States. Note: this map is based on the data collected from the 2022 National Health Interview Survey [49]. For detailed information, see Table 1A below Table 1ARegional distribution of caregivers’ access to patients’ online medical records (survey-weighted percentages)U.S. Census RegionStates Included% Never Accessed Online Medical Records% Accessed ≥ 10 Times (Past 12 Months)New EnglandME, NH, VT, MA, RI, CT46.157.60Middle AtlanticNY, NJ, PA79.495.13East North CentralOH, IN, IL, MI, WI70.598.82West North CentralMN, IA, MO, ND, SD, NE, KS57.140.00South AtlanticDE, MD, DC, VA, WV, NC, SC, GA, FL51.359.46East South CentralKY, TN, AL, MS55.888.82West South CentralAR, LA, OK, TX46.3417.07MountainMT, ID, WY, CO, NM, AZ, UT, NV66.678.33PacificWA, OR, CA, AK, HI47.7315.91Notes: (1) “Never accessed” refers to caregivers reporting zero portal access in the past 12 months. (2) “Accessed ≥ 10 times” indicates frequent portal use. (3) Regions follow standard U.S. Census Bureau classifications Table 1ARegional distribution of caregivers’ access to patients’ online medical records (survey-weighted percentages)U.S. Census RegionStates Included% Never Accessed Online Medical Records% Accessed ≥ 10 Times (Past 12 Months)New EnglandME, NH, VT, MA, RI, CT46.157.60Middle AtlanticNY, NJ, PA79.495.13East North CentralOH, IN, IL, MI, WI70.598.82West North CentralMN, IA, MO, ND, SD, NE, KS57.140.00South AtlanticDE, MD, DC, VA, WV, NC, SC, GA, FL51.359.46East South CentralKY, TN, AL, MS55.888.82West South CentralAR, LA, OK, TX46.3417.07MountainMT, ID, WY, CO, NM, AZ, UT, NV66.678.33PacificWA, OR, CA, AK, HI47.7315.91Notes: (1) “Never accessed” refers to caregivers reporting zero portal access in the past 12 months. (2) “Accessed ≥ 10 times” indicates frequent portal use. (3) Regions follow standard U.S. Census Bureau classifications Regional distribution of caregivers’ access to patients’ online medical records (survey-weighted percentages) Notes: (1) “Never accessed” refers to caregivers reporting zero portal access in the past 12 months. (2) “Accessed ≥ 10 times” indicates frequent portal use. (3) Regions follow standard U.S. Census Bureau classifications Health conditions of caregivers Caregiver health status is crucial for understanding their capacity to adopt technology. Among non-dementia caregivers, 42.31% reported their health as “good”, and 39.53% of dementia caregivers fell into this category. A relatively low percentage of both groups reported their health as “poor”. Table 1BPortal access frequency, socioeconomic and geographic characteristics by dementia caregiver status (survey-weighted)VariableNon−Dementia Caregivers (n=608) n (%) / MeanDementia Caregivers (n=293) n (%) / Meanp value† How many times accessed care recipients’ online medical records (past 12 months) <0.001 None333 (51.11)166 (60.25) 1 to 2 times88 (17.91)37 (12.64) 3 to 5 times89 (12.26)32 (11.18) 6 to 9 times37 (6.06)28 (8.89) 10 or more times61 (12.66)30 (7.03) Birth gender <0.001 Male184 (45.68)93 (32.61) Female398 (54.32)202 (67.39) Age (years) 48.4957.02 <0.001 Race/Ethnicity <0.001 Non−Hispanic White274 (56.76)166 (63.54) Non−Hispanic Black or African American110 (10.93)41 (12.07) Hispanic113 (17.56)39 (12.93) Non−Hispanic Asian37 (8.46)17 (5.18) Non−Hispanic Other29 (6.29)14 (6.28) Education <0.001 Less than High School40 (8.90)12 (5.42) High School Graduate89 (16.76)52 (21.15) Some College183 (44.27)93 (41.48) Bachelor’s Degree167 (16.67)79 (17.70) Post−Baccalaureate Degree106 (13.41)59 (14.25) Household income <0.001 063 (8.13)32 (11.20) Less than $20,00099 (15.07)30 (9.26) $20,000 to < $35,00065 (9.59)35 (11.75) $35,000 to < $50,00073 (10.63)39 (10.81) $50,000 to < $75,000103 (18.79)54 (17.97) $75,000 or more213 (37.80)110 (39.01) Employment <0.001 Yes315 (58.35)143 (53.70) No271 (41.65)152 (46.30) Married <0.001 No304 (44.84)116 (30.45) Yes312 (55.16)184 (69.55) Rural–Urban Community Area Code (RUCA) <0.001 Metropolitan531 (82.91)253 (85.61) Micropolitan47 (9.94)22 (5.72) Small town20 (3.65)19 (6.33) Rural18 (3.50)6 (2.35) Self−rated health <0.001 Excellent48 (6.78)31 (9.47) Very good201 (34.19)77 (26.36) Good253 (42.31)127 (39.53) Fair89 (13.80)57 (22.75) Poor16 (2.93)4 (1.89)Note: (1) Values are weighted percentages with unweighted sample sizes shown in column headers. (2) p-values derived from survey-adjusted Wald tests (categorical variables) and survey-adjusted linear regression (continuous age). (3) All estimates account for NHIS complex survey design (BRR replicate weights) Table 1BPortal access frequency, socioeconomic and geographic characteristics by dementia caregiver status (survey-weighted)VariableNon−Dementia Caregivers (n=608) n (%) / MeanDementia Caregivers (n=293) n (%) / Meanp value† How many times accessed care recipients’ online medical records (past 12 months) <0.001 None333 (51.11)166 (60.25) 1 to 2 times88 (17.91)37 (12.64) 3 to 5 times89 (12.26)32 (11.18) 6 to 9 times37 (6.06)28 (8.89) 10 or more times61 (12.66)30 (7.03) Birth gender <0.001 Male184 (45.68)93 (32.61) Female398 (54.32)202 (67.39) Age (years) 48.4957.02 <0.001 Race/Ethnicity <0.001 Non−Hispanic White274 (56.76)166 (63.54) Non−Hispanic Black or African American110 (10.93)41 (12.07) Hispanic113 (17.56)39 (12.93) Non−Hispanic Asian37 (8.46)17 (5.18) Non−Hispanic Other29 (6.29)14 (6.28) Education <0.001 Less than High School40 (8.90)12 (5.42) High School Graduate89 (16.76)52 (21.15) Some College183 (44.27)93 (41.48) Bachelor’s Degree167 (16.67)79 (17.70) Post−Baccalaureate Degree106 (13.41)59 (14.25) Household income <0.001 063 (8.13)32 (11.20) Less than $20,00099 (15.07)30 (9.26) $20,000 to < $35,00065 (9.59)35 (11.75) $35,000 to < $50,00073 (10.63)39 (10.81) $50,000 to < $75,000103 (18.79)54 (17.97) $75,000 or more213 (37.80)110 (39.01) Employment <0.001 Yes315 (58.35)143 (53.70) No271 (41.65)152 (46.30) Married <0.001 No304 (44.84)116 (30.45) Yes312 (55.16)184 (69.55) Rural–Urban Community Area Code (RUCA) <0.001 Metropolitan531 (82.91)253 (85.61) Micropolitan47 (9.94)22 (5.72) Small town20 (3.65)19 (6.33) Rural18 (3.50)6 (2.35) Self−rated health <0.001 Excellent48 (6.78)31 (9.47) Very good201 (34.19)77 (26.36) Good253 (42.31)127 (39.53) Fair89 (13.80)57 (22.75) Poor16 (2.93)4 (1.89)Note: (1) Values are weighted percentages with unweighted sample sizes shown in column headers. (2) p-values derived from survey-adjusted Wald tests (categorical variables) and survey-adjusted linear regression (continuous age). (3) All estimates account for NHIS complex survey design (BRR replicate weights) Portal access frequency, socioeconomic and geographic characteristics by dementia caregiver status (survey-weighted) Note: (1) Values are weighted percentages with unweighted sample sizes shown in column headers. (2) p-values derived from survey-adjusted Wald tests (categorical variables) and survey-adjusted linear regression (continuous age). (3) All estimates account for NHIS complex survey design (BRR replicate weights) Statistical analysis using ordered logistic regression: adopting technologies among family dementia caregivers in the U.S. Table 2 presents regression estimates examining overall subjective health status, socioeconomic status, and geographic predictors of accessing online medical records among caregivers, stratified by caregiver group: dementia caregivers, caregivers of people without dementia, and all caregivers. Among all caregivers (Column 3), female caregivers were significantly associated with a higher likelihood of adopting digital technology compared to male caregivers (OR = 1.931, p < 0.001), while the association was not statistically significant among dementia caregivers. Age was negatively associated with accessing online medical records among caregivers of people without dementia (OR = 0.977, p < 0.001), suggesting younger caregivers are more likely to adopt digital technology. No significant association was observed in the other models. Compared to non-Hispanic White caregivers, non-Hispanic Black or African American caregivers were significantly less likely to adopt digital technology across all caregivers (OR = 0.539, p < 0.05). No significant differences were observed for other racial/ethnic groups in any of the models. Educational attainment was positively associated with accessing online medical records-compared to caregivers with less than a high school education, those with a bachelor’s degree (OR = 1.769, p < 0.001) were significantly more likely to adopt digital technology. Being married was positively associated with accessing online medical records among caregivers of individuals without dementia (OR = 2.028, p < 0.001) and across all caregivers (OR = 1.749, p < 0.001). The geographic region also influenced accessing online medical records. Among all caregivers, caregivers living in micropolitan areas were significantly less likely to adopt the digital technology compared to those in metropolitan areas (OR = 0.633, p < 0.001). Interestingly, among dementia caregivers, those in rural areas were also less likely to access medical records of their loved ones compared to those in metropolitan areas (OR = 0.529, p < 0.05). In the full model (Column 3), the interaction effect between age and health condition on accessing online medical records is statistically significant, which confirms the previous studies that gender and conditions interact with each other on health behaviors [57, 58]. Notably, poor health interacted with age to significantly reduce the likelihood of accessing online medical records (OR = 0.841, p < 0.001), indicating that older caregivers in poor health were particularly less likely to engage with digital technology. A significant negative interaction was also found for those in very good health (OR = 0.948, p < 0.001). Figure 3 visually shows that family caregivers’ overall health condition interacts with their age in terms of their use of technologies for tracking online medical records of people living with dementia. Figure 3 left shows the interaction effect of family caregivers’ age and overall health condition on the predicted probability of family caregivers’ high-frequency access to online medical records of people living with dementia (more than 10 times). For older family caregivers with poor self-rated health, their predicted probability of accessing online medical records frequently is significantly lower than their younger counterparts, which declines sharply with increasing age from 18 to 98. Figure 3 on the left shows that the predicted probability of family caregivers using technologies to track medical records of people living with dementia more than 10 times in the past 12 months is above 80% if they are in overall poor health condition and at the age between 18 and 28. This indicates high technology utilization by young family caregivers who are in an overall poor health condition. On the other hand, if the family caregivers’ overall health condition is excellent, their predicted probability (around 10%) of using technology does not vary dramatically by age. Table 2Survey-weighted ordered logistic regression predicting socioeconomic and geographic characteristics of caregivers on portal access frequency (odds ratios)Variables(1) Dementia Caregivers(2) Non-Dementia Caregivers(3) All Caregivers(4) All Caregivers + Age×Health Female 2.676 (0.191)***2.020 (0.087)***1.931 (0.074)***1.916 (0.073)*** Age 0.987 (0.003)***0.980 (0.002)***0.977 (0.002)***1.011 (0.003)***Compared to Non-Hispanic White: Non-Hispanic Black0.495 (0.042)***0.525 (0.028)***0.539 (0.022)***0.534 (0.023)*** Hispanic0.706 (0.058)***0.877 (0.072)0.740 (0.046)***0.700 (0.044)*** Non-Hispanic Asian0.235 (0.029)***0.754 (0.098)*0.474 (0.050)***0.568 (0.056)*** Non-Hispanic Other0.765 (0.079)*1.957 (0.182)***1.349 (0.081)***1.401 (0.077)***Compared to < High School: High School Graduate1.230 (0.222)1.030 (0.161)1.041 (0.124)1.053 (0.139) Some College1.716 (0.287)**0.887 (0.110)1.122 (0.106)1.084 (0.117) Bachelor’s Degree5.223 (0.890)***1.134 (0.131)1.769 (0.168)***1.647 (0.177)*** Post-Baccalaureate Degree2.407 (0.416)***0.615 (0.072)***0.916 (0.089)0.847 (0.090) Income 1.044 (0.027)1.060 (0.020)**1.076 (0.016)***1.090 (0.017)*** Employed 1.098 (0.086)0.575 (0.033)***0.735 (0.034)***0.744 (0.034)*** Married 1.121 (0.093)2.028 (0.119)***1.749 (0.068)***1.685 (0.069)***Compared to Metropolitan: Micropolitan0.348 (0.045)***0.654 (0.097)**0.633 (0.071)***0.685 (0.092)** Small town0.079 (0.267)0.861 (0.087)0.522 (0.047)***0.508 (0.044)*** Rural0.529 (0.152)*0.804 (0.082)*0.937 (0.075)1.000 (0.084)Compared to Excellent Health: Very good0.268 (0.023)***1.246 (0.062)***0.932 (0.043)14.868 (2.794)*** Good0.634 (0.052)***0.632 (0.041)***0.669 (0.033)***2.244 (0.476)*** Fair0.666 (0.072)***0.458 (0.030)***0.598 (0.033)***3.314 (0.531)*** Poor0.000 (0.000)***1.464 (0.263)*0.947 (0.147)6803.118 (18980.06)** Dementia caregiver (vs. non-dementia) 0.734 (0.025)***0.764 (0.026)*** Age × Health Interaction Terms Age × Very good0.948 (0.003)*** Age × Good0.978 (0.004)*** Age × Fair0.968 (0.003)*** Age × Poor0.841 (0.053)** Observations 268550818818Outcome: portal access frequencyFormat: Odds Ratio (Std. Err.)Significance levels: p < 0.05; ** p < 0.01; *** p < 0.001 Table 2Survey-weighted ordered logistic regression predicting socioeconomic and geographic characteristics of caregivers on portal access frequency (odds ratios)Variables(1) Dementia Caregivers(2) Non-Dementia Caregivers(3) All Caregivers(4) All Caregivers + Age×Health Female 2.676 (0.191)***2.020 (0.087)***1.931 (0.074)***1.916 (0.073)*** Age 0.987 (0.003)***0.980 (0.002)***0.977 (0.002)***1.011 (0.003)***Compared to Non-Hispanic White: Non-Hispanic Black0.495 (0.042)***0.525 (0.028)***0.539 (0.022)***0.534 (0.023)*** Hispanic0.706 (0.058)***0.877 (0.072)0.740 (0.046)***0.700 (0.044)*** Non-Hispanic Asian0.235 (0.029)***0.754 (0.098)*0.474 (0.050)***0.568 (0.056)*** Non-Hispanic Other0.765 (0.079)*1.957 (0.182)***1.349 (0.081)***1.401 (0.077)***Compared to < High School: High School Graduate1.230 (0.222)1.030 (0.161)1.041 (0.124)1.053 (0.139) Some College1.716 (0.287)**0.887 (0.110)1.122 (0.106)1.084 (0.117) Bachelor’s Degree5.223 (0.890)***1.134 (0.131)1.769 (0.168)***1.647 (0.177)*** Post-Baccalaureate Degree2.407 (0.416)***0.615 (0.072)***0.916 (0.089)0.847 (0.090) Income 1.044 (0.027)1.060 (0.020)**1.076 (0.016)***1.090 (0.017)*** Employed 1.098 (0.086)0.575 (0.033)***0.735 (0.034)***0.744 (0.034)*** Married 1.121 (0.093)2.028 (0.119)***1.749 (0.068)***1.685 (0.069)***Compared to Metropolitan: Micropolitan0.348 (0.045)***0.654 (0.097)**0.633 (0.071)***0.685 (0.092)** Small town0.079 (0.267)0.861 (0.087)0.522 (0.047)***0.508 (0.044)*** Rural0.529 (0.152)*0.804 (0.082)*0.937 (0.075)1.000 (0.084)Compared to Excellent Health: Very good0.268 (0.023)***1.246 (0.062)***0.932 (0.043)14.868 (2.794)*** Good0.634 (0.052)***0.632 (0.041)***0.669 (0.033)***2.244 (0.476)*** Fair0.666 (0.072)***0.458 (0.030)***0.598 (0.033)***3.314 (0.531)*** Poor0.000 (0.000)***1.464 (0.263)*0.947 (0.147)6803.118 (18980.06)** Dementia caregiver (vs. non-dementia) 0.734 (0.025)***0.764 (0.026)*** Age × Health Interaction Terms Age × Very good0.948 (0.003)*** Age × Good0.978 (0.004)*** Age × Fair0.968 (0.003)*** Age × Poor0.841 (0.053)** Observations 268550818818Outcome: portal access frequencyFormat: Odds Ratio (Std. Err.)Significance levels: p < 0.05; ** p < 0.01; *** p < 0.001 Survey-weighted ordered logistic regression predicting socioeconomic and geographic characteristics of caregivers on portal access frequency (odds ratios) Outcome: portal access frequency Format: Odds Ratio (Std. Err.) Significance levels: p < 0.05; ** p < 0.01; *** p < 0.001 Fig. 2Impact of socioeconomic and geographic characteristics on portal access frequency among caregivers. Note: Fig. 2 is produced based on the Table 2 model (1–3) Impact of socioeconomic and geographic characteristics on portal access frequency among caregivers. Note: Fig. 2 is produced based on the Table 2 model (1–3) Fig. 3The interaction effect between family caregivers’ health condition and age on their predicted probability of accessing online medical records of people living with dementia in the past 12 months. (a) Predicted probability of accessing online medical records of people living with dementia more than 10 times; (b) Predicted probability of accessing online medical records of people living with dementia zero times. Figure 3 is produced based on the Table 2 model (4). Note: The red line represents that the family caregivers’ health condition is poor, while the blue line represents that the family caregivers’ overall health condition is excellent The interaction effect between family caregivers’ health condition and age on their predicted probability of accessing online medical records of people living with dementia in the past 12 months. (a) Predicted probability of accessing online medical records of people living with dementia more than 10 times; (b) Predicted probability of accessing online medical records of people living with dementia zero times. Figure 3 is produced based on the Table 2 model (4). Note: The red line represents that the family caregivers’ health condition is poor, while the blue line represents that the family caregivers’ overall health condition is excellent Discussion The study compares the demographics and online medical behavior between all caregivers and dementia caregivers. The general trends show that dementia caregivers are slightly older, more likely to be married, and have a higher income compared to the overall caregiver population. Although online medical records access remains infrequent, a portion of dementia caregivers engage with technology more regularly, highlighting opportunities to further support these caregivers with digital tools aimed at improving care management. This descriptive analysis can inform future public health initiatives aimed at increasing access to online medical records and improving support systems for family caregivers of individuals with dementia. Ordered logistic regression results suggest that gender, age, education, marital status, and geographic location are significantly associated with accessing online medical records among family caregivers in general. Specifically, younger, female, highly educated, married caregivers from metropolitan areas are more likely to utilize technology for accessing medical records [64–65]. Additionally, aligning with prior work, we found that non-Hispanic Black caregivers were significantly less likely to adopt digital health technologies compared to non-Hispanic White caregivers. While prior studies have addressed urban-rural digital divides in general health technology access, fewer studies have specifically explored this phenomenon among dementia caregivers. This study shows that dementia caregivers from small towns were significantly less likely to use this digital technology compared to others, which confirms that the geographic digital divide persists across the U.S. for dementia caregivers. These findings highlight the importance of considering demographic and geographic factors when developing policies and interventions aimed at improving portal access frequency among caregivers, particularly in the context of dementia. This indicates that dementia caregivers residing in small towns need more technological support in caregiving, especially as they face the emerging artificial intelligence-assisted technology tools in caregiving. One more critical addition of this study to the existing literature is its focus on geographic regional disparities in caregivers’ access to patients’ online medical records. Regions such as the Mountain and Middle Atlantic showed particularly low levels of engagement, while caregivers in the West South Central and Pacific regions were more likely to access digital records frequently. These patterns suggest that geographic location may influence caregivers’ digital health behaviors, potentially reflecting disparities in infrastructure, health system integration, and digital literacy. Furthermore, the analysis reveals additional complexities in the interplay between caregivers’ age and health status in shaping portal access frequency among dementia caregivers. The interaction effects reveal that caregivers in poor health are less likely to adopt technology as they age, suggesting that health status plays a critical role in the adoption process. Policy and intervention efforts aimed at increasing portal access frequency among caregivers should consider these interaction effects, targeting older caregivers, especially those in poor health or with lower educational attainment. Providing tailored training and support could help reduce barriers and facilitate greater technology engagement among these groups. Our findings demonstrate that caregivers in poor health experience an even steeper decline in accessing online medical records as they age - a pattern that has been underexplored in previous literature. Together, these findings underscore the need for targeted interventions to improve equitable access to online medical records among caregivers, particularly those who are older, in poorer health, or residing in underserved geographic areas. These findings should be interpreted in light of the study’s limitations, particularly the cross-sectional design and reliance on self-reported measures. Limitations This study uses a cross-sectional design and therefore cannot establish causal relationships between socioeconomic characteristics, health status, geographic location, and portal access frequency. In addition, although we conducted subgroup analyses to generate caregiving-context–specific estimates for dementia and non-dementia caregivers, the sample of dementia caregivers (n = 300) is relatively modest. Dementia caregiving differs substantively in care intensity, cognitive demands, and duration of caregiving responsibilities; therefore, estimating separate models allows for clinically meaningful interpretation that may not be fully captured in pooled analyses alone. However, the smaller subgroup size reduces statistical power and may limit the precision of estimates, as reflected in wider confidence intervals and occasional instability in model parameters. Consequently, some non-significant findings among dementia caregivers may reflect limited power rather than the true absence of association. To address this concern, we also estimated pooled models with interaction terms to formally test differences between dementia and non-dementia caregivers while maximizing statistical efficiency. Subgroup findings should therefore be interpreted in conjunction with the pooled interaction results and viewed as exploratory rather than definitive. Conclusion This study provides nationally representative evidence that portal access among family caregivers, particularly dementia caregivers, is shaped by gender, educational attainment, race/ethnicity, geographic location, and the interaction between age and health status. Dementia caregivers had lower overall odds of portal access in pooled models, and caregivers residing in micropolitan and small-town areas were consistently less likely to engage with online medical records. In addition, older caregivers in poorer health exhibited compounded disadvantages in digital engagement. These findings have several implications. For research, future studies should use longitudinal designs to examine causal pathways and incorporate more detailed measures of digital literacy, caregiving intensity, and structural access barriers. For policy, digital health equity initiatives should prioritize caregivers in micropolitan and small-town communities and older caregivers with health limitations. For practice, healthcare systems should implement targeted portal onboarding, training, and usability adaptations tailored to dementia caregiving contexts. By addressing these identified disparities, digital health strategies can more effectively support care coordination, reduce caregiver strain, and improve outcomes for individuals living with dementia and their families.