Abstract
INTRODUCTION
Few studies have explored dementia risk according to sex and gender including for transgender and non‐binary adults. This study evaluated dementia risk factors and risk scores among cisgender, transgender, and non‐binary adults.
METHODS
Observational data were drawn from the 2019 Behavioral Risk Factor Surveillance System. A matched‐cohort approach was used to develop sex (male, female) and gender identity cohorts (cisgender men, cisgender women, transgender men, transgender women, and non‐binary adults) for comparison. Dementia risk scores were calculated using established mid‐life and late‐life risk score algorithms.
RESULTS
Males had higher overall mid‐life dementia risk, and lower late‐life Alzheimer's disease risk compared to females. Transgender men, transgender women, and non‐binary adults had higher overall late‐life risk compared to both cisgender men and women.
DISCUSSION
Future research is needed to build the evidence base for specific risk factors that may be contributing to higher overall risk among understudied and underserved gender groups.
Highlights
Using data from the 2019 Behavioral Risk Factor Surveillance System, this matched‐cohort study found that those assigned female at birth had lower overall mid‐life dementia risk and higher overall late‐life Alzheimer's disease (AD) risk compared to those assigned male at birth.
Transgender men, transgender women, and non‐binary adults all showed higher overall late‐life AD risk compared to cisgender men and cisgender women.
Between‐group differences were found in the incidence of specific risk and protective factors for dementia and AD.
Keywords: Alzheimer's disease risk, dementia risk, gender, sex, sex and gender differences
1. BACKGROUND
Alzheimer's disease (AD) accounts for between 60% to 80% of all cases of dementia. In the United States alone, more than 6.2 million people aged ≥65 years are currently living with AD. By 2025, this figure is expected to grow to 7.2 million. 1 While some risk factors for AD and dementia, such as age, genetics, and family history, cannot be changed, other risk factors can be modified to reduce the risk of cognitive decline and dementia. The 2020 recommendations of The Lancet Commission on dementia prevention, intervention, and care suggests that addressing modifiable risk factors during mid‐life and later life might prevent or delay up to 40% of dementia cases. 2 There is growing interest in the impact of sex and gender on health outcomes. However, no studies to date have examined sex and gender differences in overall mid‐life or late‐life dementia risk inclusive of underserved transgender and gender non‐binary adults.
RESEARCH IN CONTEXT
Systematic review: The authors reviewed the literature using traditional (e.g., PsychInfo, PubMed, Google Scholar) sources and conference abstracts. While there is a growing body of literature exploring sex differences in dementia and Alzheimer's disease (AD) risk, studies have only recently begun to consider both sex and gender differences in risk. No articles were identified that report information on overall dementia and AD risk scores in transgender and non‐binary adults.
Interpretation: In line with previous sex differences research, we found that those assigned female at birth had lower mid‐life dementia risk and higher late‐life AD risk compared to those assigned male at birth. Gender differences were also found. Transgender men, transgender women, and non‐binary adults all showed higher overall late‐life AD risk compared to cisgender men and cisgender women.
Future directions: To overcome the limitations of this study and extend understanding of sex and gender differences in dementia and AD risk, several future directions are noted: (a) large‐scale public health surveys should use standardized sex and gender questions that enable accurate self‐reporting of multi‐faceted sex and gender; (b) the same surveys should measure a broader array of evidence‐based risk and protective factors for dementia and AD, including social, psychological, and cardiometabolic factors that may disproportionately impact risk among underserved communities; and (c) data should be collected to enable exploration of the validity of risk score algorithms developed in predominantly cisgender populations among transgender and gender‐diverse samples.
The terms sex and gender are interrelated and often used interchangeably; however, they are distinct concepts. Sex refers to one's biological status as male, female, or another variation of sex characteristics. Gender is a dynamic social construct that refers to psychological, social, and cultural factors that shape attitudes, behaviors, stereotypes, and knowledge. 3 For individuals who are cisgender, their sex and gender identity match prevailing norms. For individuals who are transgender or gender diverse (TGD), their sex and gender identity do not match prevailing norms.
There is growing evidence for sex differences in dementia and AD risk. Studies have typically found that males have higher rates of vascular risk factors 4 and hypertension 5 than females, but these relationships are age dependent, with the prevalence of both increasing post‐menopause for females. Some studies suggest that females are more susceptible to the effect of mid‐life hypertension on memory, which may increase their risk for developing dementia in later life. 5 Females have shown higher rates of depression than males. 6 Female carriers of the apolipoprotein E ε4 allele, known to be the strongest genetic risk factor for dementia and AD, are at a higher risk of developing dementia compared to males. 7
The role of gender in accounting for differences in dementia and AD risk has been far less thoroughly investigated. Social factors such as low education and occupational attainment are well‐known risk factors in both males and females, with gendered inequalities thought to be partially responsible for lower education and occupational attainment among older females and resultantly, putting them at a high risk for developing dementia. 8 Compared to cisgender adults, TGD adults have been shown to be 65% more likely to report subjective cognitive decline, 9 which has been identified as one of the earliest noticeable signs of AD and other dementias. TGD adults are also at a higher risk of a subset of individual risk factors for AD, including experiencing higher rates of hypertension, hypercholesterolemia, stroke, coronary heart disease, depression, smoking, and heavy drinking. 10
Recently, attempts have been made to outline both sex and gender differences in health broadly, 11 and AD and dementia specifically. 8 , 12 While this work offers insight into the biological and social factors impacting cisgender differences in dementia risk, more work is needed to understand risk among TGD identities. Several issues have limited exploration of sex and gender differences in AD and dementia risk. First, there has been a lack of conceptual clarity regarding sex and gender, which has led to the terms being used interchangeably. Second, there is a lack of consensus on how to measure sex and gender in large‐scale surveys, and third, a lack of available data exploring both sex and gender.
1.1. Study objectives
This study aims to (1) characterize how those assigned male at birth and female at birth differ in mid‐life and late‐life dementia and AD risk scores and risk factors, and (2) characterize how cisgender men, cisgender women, transgender men, transgender women, and gender non‐binary adults differ in mid‐life and late‐life dementia and AD risk factors and risk scores.
2. METHODS
This project was pre‐registered on Open Science Framework (https://osf.io/f5y92).
2.1. Dataset
In this study, we use cross‐sectional data from the 2019 wave of the Behavioral Risk Factor Surveillance Survey (BRFSS)—an annual, nationally representative telephone survey of adults aged > 18 years living in the United States. 13 The sampling frame for the 2019 BRFSS is described elsewhere. 13 The 2019 wave was not combined with any other wave because it is not clear that participants are prevented from being sampled at multiple waves. For sex‐based comparisons, sex at birth was determined based on the question: “What was your sex at birth? Was it male or female?” Seven states collected data on birth sex in 2019.
For gender‐based comparison we used data from 31 states and territories that included an optional module on sexuality and gender identity, which included a differently worded self‐reported sex question (“Are you male or female?”) as well as measuring gender identity. Gender identity was determined based on the question: “Do you consider yourself to be transgender?” If participants responded yes, they were asked a follow‐up question to determine their identity from a list of three identifiers: (1) transgender male‐to‐female (transgender women), (2) transgender female‐to‐male (transgender men), or (3) gender nonconforming (non‐binary adults). If people answered no to the transgender question, then it was assumed that their gender identity aligned to their sex, and they were coded as either cisgender men or cisgender women. The survey also collected information on the following core sociodemographic characteristics (age, education level, race, ethnicity, language background) and dementia and/or AD risk and protective factors (obesity, high cholesterol, stroke, hypertension, use of hypertensive medications, diabetes, heart problems, heart attack, myocardial infarction, depression, kidney disease, smoking status, alcohol consumption, and physical activity).
2.2. Matched‐cohort approach for gender analysis
A total of 955 respondents identified as transgender or gender diverse (by answering “yes” to the transgender gender identity question), and 230,459 participants identified as cisgender (by answering “no” to the transgender gender identity question). We chose to use a matched‐cohort approach to match gender subgroups based on age. Matching on age was crucial as many of the dementia risk factors increase in prevalence in the population with chronical age. We did not follow the propensity matching approach from previous research 10 because in this study, a large amount of missing data on race among non‐binary adults would have led to unacceptably low statistical power. Other factors that have previously been used to match groups in propensity matching—such as education level—are important contributors to dementia risk and were allowed to vary across groups.
As gender non‐binary was the least frequently reported gender identity (n = 226) and younger on average than other gender groups, to control for age confounds, each gender non‐binary adult was matched based on age to two cisgender male and two cisgender female adults. However, as the transgender male and transgender female groups were only slightly larger than the non‐binary group (and comparatively much smaller in size than the cisgender groups), they could only be matched 1:1 with the non‐binary group based on age.
2.3. Risk scores
Overall dementia or AD risk was estimated using three commonly used risk score algorithms that have been validated against clinical diagnoses. 14 , 15 , 16 In order of publication, the Cardiovascular Risk Factors, Aging and Dementia (CAIDE) risk tool estimates mid‐life risk of dementia among adults aged 40 to 65; the Australian National University Alzheimer's Disease Risk Index (ANU‐ADRI) estimates late‐life AD risk and was used with respondents aged 55+; and the Lifestyle for Brain Health (LIBRA) dementia risk tool estimates mid‐ to late‐life dementia risk among adults aged 50+. Age and sex were not included in the calculation of CAIDE and ANU‐ADRI as they were used in sample matching. When respondents had missing data on one or more of the variables required for risk score calculation, they were excluded from analysis. More information on each of the risk scores, including the importance of modifiable risk factors, and transformations to demographic, health, and lifestyle factors to allow for inclusion in the risk score algorithms is included in supporting information.
2.4. Statistical analysis
Linear regression was used to compare risk scores between sex at birth groups. Analysis of variance was used to compare risk scores among the five gender identity subgroups and dummy coded linear regression analysis was used to follow‐up significant between‐gender group differences in risk scores. Chi‐squared analysis was used to compare categorical risk factors. Bonferroni corrected comparisons were used to compare subgroups against each other using α = 0.05. Where the assumption of equality of variance between groups was violated, Welch's correction was used. All statistical analysis was conducted using SPSS v25.
2.5. Role of funding source
The funders of the study had no role in study design, data collection, data analysis, data interpretation, writing of the report, or the decision to submit for publication.
3. RESULTS
3.1. Sex differences
3.1.1. Risk score comparisons
Compared to those assigned male at birth, those assigned female showed lower mid‐life dementia risk measured using the CAIDE (males, n = 7286, M = 2.43, standard deviation [SD] = 2.05, range = 0.00 to 10.00; females, n = 6866, M = 1.85, SD = 1.93, range = 0.00 to 10.00, β = −0.575, [confidence interval (CI): −0.641, −0.510], P < 0.001), higher late‐life AD risk measured by the ANU‐ADRI (males, n = 11,369, M = −1.30, SD = 3.34, range = −6.00 to 15.00; females, n = 10,855, M = −0.75, SD = 3.30, range = −6.00 to 15.00, β = 0.553, [CI: 0.466, 0.641], P < 0.001) and lower mid‐ to late‐life dementia risk measured using the LIBRA tool (males, n = 9947, M = 1.08, SD = 0.076, range = 0.95 to 1.33; females, n = 9088, M = 1.074, SD = 0.079, range = 0.95 to 1.36, β = −0.009, [CI: −0.011, −0.007], P < 0.001).
3.1.2. Risk profiles by sex
Comparisons between those assigned male or female at birth on specific risk and protective factors are summarized in Table 1. Compared to those assigned male, a higher proportion of those assigned female were college educated, used anti‐hypertensive medications, had depression, and ate more than five servings of fruit and vegetables per week. A higher proportion of those assigned female had also never smoked, were abstainers from alcohol, and were inactive. A higher proportion of those assigned male had obesity, high cholesterol, had a history of stroke, heart problems (both heart attack and heart disease), were former or current smokers, heavy drinkers, and reported being highly active.
TABLE 1.
Differences between self‐reported males and females in specific dementia and Alzheimer's disease risk and protective factors.
| Characteristic | Assigned male (n = 25,150) | Assigned female (n = 25,150) | χ 2 | df | P‐value | Cramer's V |
|---|---|---|---|---|---|---|
| Age (M, SD) | 53.71 (17.29) | 53.71 (17.29) | ƒ | |||
| Education level, n (%) | ||||||
| High school only | 7836 (31.3%) | 7028 (28%) | 63.49 | 1 | <0.001 | 0.036 |
| College and above | 17,221 (68.7%) | 18,052 (72%) | ||||
| Race, n (%) | ||||||
| American Indian or Alaskan Native | 247 (0.98%) | 214 (0.85%) | ||||
| Asian | 1342 (5.34%) | 1347 (5.36%) | ||||
| Black or African American | 1166 (4.64%) | 1167 (4.64%) | ||||
| Multiracial | 920 (3.66%) | 970 (3.86%) | ||||
| Native Hawaiian or other Pacific Islander | 421 (1.67%) | 523 (2.08%) | 41.515a | 8 | <0.001 | 0.029 |
| Other | 641 (2.55%) | 492 (1.96%) | ||||
| White | 19,772 (78.62%) | 19,850 (78.93%) | ||||
| Don't know | 288 (1.15%) | 300 (1.19%) | ||||
| Refused | 353 (1.40%) | 287 (1.14%) | ||||
| Ethnicity, n (%) | ||||||
| Hispanic, Latino/a, or Spanish origin | 1600 (6.36%) | 1600 (6.36%) | ||||
| Not of Hispanic, Latino/a, or Spanish origin | 23,288 (92.60%) | 23,349 (92.84%) | 8.117a | 2 | 0.017 | 0.013 |
| Don´t know, refused or missing | 262 (1.04%) | 201 (0.80%) | ||||
| Obesity, n (%) | 7435 (30.8%) | 6654 (29.6%) | 8.052 | 1 | 0.005 | 0.013 |
| High cholesterol, n (%) | 8285 (38.5%) | 7409 (33.8%) | 106.87 | 1 | <0.001 | 0.05 |
| Stroke, n (%) | 976 (3.9%) | 816 (3.3%) | 14.921 | 1 | <0.001 | 0.017 |
| Hypertension, n (%) | 9923 (39.6%) | 7975 (31.8%) | 333.98 | 1 | <0.001 | 0.082 |
| Use of anti‐hypertensive medications, n (%) | 7568 (76.4%) | 6632 (83.3%) | 129.93 | 1 | <0.001 | 0.085 |
| Diabetes, n (%) | 3199 (12.7%) | 2646 (10.5%) | 59.24 | 1 | <0.001 | 0.034 |
| Heart problems | 928 (3.7%) | 359 (1.4%) | 258.18 | 1 | <0.001 | 0.072 |
| Heart attack, n (%) | 1707 (6.8%) | 785 (3.1%) | 361.3 | 1 | <0.001 | 0.085 |
| Angina and coronary heart disease, n (%) | 1555 (6.2%) | 806 (3.2%) | 250.62 | 1 | <0.001 | 0.071 |
| Depressive Disorder, n (%) | 3549 (14.2%) | 6134 (24.5%) | 852.85 | 1 | <0.001 | 0.131 |
| Kidney disorder, n (%) | 851 (3.4%) | 820 (3.3%) | 0.617 | 1 | 0.432 | 0.004 |
| Smoking status, n (%) | ||||||
| Smoking current | 3273 (13.5%) | 2757 (11.3%) | ||||
| Former | 7428 (30.5%) | 5738 (23.6%) | 427.06 | 2 | <0.001 | 0.094 |
| Never | 13,621 (56.0%) | 15,830 (65.1%) | ||||
| Alcohol consumption, n (%) | ||||||
| Heavy drinker | 2320 (9.8%) | 776 (3.2%) | 980.93 | 2 | <0.001 | 0.143 |
| Moderate drinker | 10,847 (45.6%) | 10,389 (43.4%) | ||||
| Abstainer | 10,612 (44.6%) | 12,784 (53.4%) | ||||
| >5 servings of fruits or vegetables | 3322 (15.0%) | 4460 (19.9%) | 188 | 1 | <0.001 | 0.065 |
| Physical activity, n (%) | ||||||
| Highly active | 9453 (41.2%) | 8181 (35.8%) | 143.57 | 3 | <0.001 | 0.056 |
| Active | 3859 (17.3%) | 4450 (19.5%) | ||||
| Insufficiently active | 4045 (17.6%) | 4329 (19.0%) | ||||
| Inactive | 5481 (23.9%) | 5876 (25.7%) | ||||
Note: ƒ = non‐significant (case‐control matched).
Abbreviation: SD, standard deviation.
3.2. Gender differences
3.2.1. Risk score comparisons
Risk scores for each gender identity group are displayed in Table 2. No significant difference in CAIDE (Welch statistic = 2.119, P = 0.085) risk scores were found among the subgroups. There were significant gender differences in ANU‐ADRI risk scores (F = 10.268, P < 0.001). As shown in Table 3, comparisons among gender groups computed using dummy coded regression analysis showed that transgender women, transgender men, and non‐binary adults all had significantly higher late‐life risk scores than both cisgender men and cisgender women. A significant difference between the subgroups was found using the LIBRA risk tool (Welch statistic = 3.215, P = 0.015). However post hoc Games–Howell comparisons could not detect significant differences among the subgroups after correction for multiple comparisons.
TABLE 2.
Risk score comparisons among gender subgroups.
| Gender identity group | CAIDE | LIBRA | ANU‐ADRI | ||||||
|---|---|---|---|---|---|---|---|---|---|
| n | M | SD (range) | n | M | SD (range) | n | M | SD (range) | |
| Transgender women | 34 | 2.56 | 2.15 (0.00–8.00) | 39 | 1.10 | 0.061 (0.95–1.25) | 57 | 1.02 | 3.78 (−6.00–10.00) |
| Transgender men | 34 | 3.44 | 2.54 (0.00–8.00) | 43 | 1.12 | 0.098 (0.95–1.31) | 56 | 1.25 | 4.19 (−6.00–12.00) |
| Non‐binary adults | 30 | 2.80 | 2.41 (0.00–7.00) | 42 | 1.11 | 0.090 (0.95–1.30) | 59 | 0.85 | 3.85 (−6.00–9.00) |
| Cisgender men | 71 | 2.48 | 1.93 (0.00–6.00) | 97 | 1.08 | 0.072 (0.95–1.27) | 120 | −1.41 | 3.29 (−6.00 to 6.00) |
| Cisgender women | 74 | 2.09 | 1.84 (0.00–7.00) | 99 | 1.08 | 0.080 (0.95–1.26) | 123 | −1.10 | 3.44 (−6.00–9.00) |
| Total | 243 | 2.55 | 2.12 (0.00–8.00) | 320 | 1.09 | 0.081 (0.95–1.31) | 415 | −0.30 | 3.77 (−6.00–12.00) |
Abbreviations: ANU‐ADRI, Australian National University Alzheimer's Disease Risk Index; CAIDE, Cardiovascular Risk Factors, Aging and Dementia; LIBRA, Lifestyle for Brain Health; SD, standard deviation.
TABLE 3.
Regression analyses for gender identity differences in late‐life AD risk based on ANU‐ADRI scores.
| Comparison | 95% CI | |||||
|---|---|---|---|---|---|---|
| Identity | Identity | β | Std. error | P | Lower | Upper |
| Transgender women | Transgender men | −0.232 | 0.750 | 0.757 | −1.719 | 1.254 |
| Transgender women | Non‐binary adults | 0.170 | 0.708 | 0.811 | −1.233 | 1.573 |
| Transgender women | Cisgender men | 2.426 | 0.556 | <0.001 | 1.329 | 3.523 |
| Transgender women | Cisgender women | 2.115 | 0.569 | <0.001 | 0.991 | 3.239 |
| Transgender men | Non‐binary adults | 0.403 | 0.749 | 0.592 | −1.082 | 1.887 |
| Transgender men | Cisgender men | 2.658 | 0.582 | <0.001 | 1.509 | 3.808 |
| Transgender men | Cisgender women | 2.348 | 0.595 | <0.001 | 1.173 | 3.522 |
| Non‐binary adults | Cisgender men | 2.256 | 0.554 | <0.001 | 1.163 | 3.349 |
| Non‐binary adults | Cisgender women | 1.945 | 0.567 | 0.001 | 0.827 | 3.063 |
| Cisgender men | Cisgender women | −0.311 | 0.432 | 0.473 | −1.162 | 0.541 |
Abbreviations: ANU‐ADRI, Australian National University Alzheimer's Disease Risk Index; AD, Alzheimer's disease.
3.2.2. Risk profiles by gender
Figure 1 illustrates the percentage of each gender group that reported each risk and protective factor. Risk profile comparisons showed that more cisgender women (67%) were college educated than transgender women (48.4%). Transgender men (5.8%), transgender women (5.6%), and non‐binary adults (5.95%) all showed higher rates of heart attacks compared to cisgender women (1.6%). Transgender men (5.8%) and transgender women (5.6%) also showed higher rates of myocardial infarction and coronary heart disease compared to cisgender women (1.6%). Non‐binary adults showed the highest rates of depression (49.8%), significantly higher compared to cisgender men (13.8%), cisgender women (23.6%), and transgender women (30.7%). Transgender men showed the second highest rates of depression (39.8%), significantly higher than both cisgender men and cisgender women. Non‐binary adults showed higher rates of diabetes (13.8%) compared to cisgender women (5.8%). Non‐binary adults also showed higher rates of kidney problems (8%) compared to cisgender men (0.9%) and cisgender women (1.3%). Cisgender women reported higher rates of having never smoked (66.5%) compared to cisgender men (53.2%). Cisgender women reported higher rates of moderate drinking (51.4%) compared to transgender men (36.8%) and significantly lower rates of heavy drinking (2.4%) compared to cisgender men (11.2%), transgender women (10.1%) and non‐binary adults (9.3%). Last, non‐binary adults showed the highest rates of physical inactivity (35.6%) compared to both cisgender men (23.2%) and women (23.9%). See Table 4 for group comparisons.
FIGURE 1.
Percentage of gender identity groups endorsing dementia and Alzheimer's disease risk and protective factors with significant between‐group differences highlighted.
TABLE 4.
Differences among transgender women, transgender men, non‐binary adults, cisgender men, and cisgender women in specific risk and protective factors for dementia and Alzheimer's disease.
| Transgender women | Transgender men | Non‐binary adults | Cisgender men | Cisgender women | |||||
|---|---|---|---|---|---|---|---|---|---|
| Characteristic | (n = 192) | (n = 199) | (n = 226) | (n = 452) | (n = 452) | χ 2 | df | P‐value | Cramer's V |
| Age (M, SD) | 43.89 (18.975) | 43.75 (19.628) | 42.23 (19.497) | 42.23 (19.475) | 42.23 (19.475) | NS (case‐control matched) | |||
| Education level (n %) | |||||||||
| High school only | 98 (51.6%) | 84 (42.2%) | 88 (39.5%) | 190 (42.2%) | 149 (33.0%) | 20.998 | 4 | <0.001 | 0.118 |
| College and above | 92 (48.4%) | 115 (57.8%) | 135 (60.5%) | 260 (57.8%) | 302 (67.0%) | ||||
| Race, n (%) | |||||||||
| American Indian or Alaskan Native | 12 (6.25%) | 9 (4.52%) | 12 (5.31%) | 57 (12.61%) | 52 (11.50%) | ||||
| Asian | 5 (2.60%) | 13 (6.53%) | 5 (2.21%) | 13 (2.88%) | 13 (2.88%) | ||||
| Black or African American | 15 (7.81%) | 20 (10.05%) | 15 (6.64%) | 14 (3.10%) | 15 (3.32%) | ||||
| Multiracial | 7 (3.65%) | 7 (3.52%) | 14 (6.19%) | 21 (4.65%) | 19 (4.20%) | ||||
| Native Hawaiian or other Pacific Islander | 1 (0.52%) | 11 (5.53%) | 3 (1.33%) | 1 (0.22%) | 2 (0.44%) | 98.996 | 32 | 0.000 | 0.128 |
| Other | 9 (4.69%) | 10 (5.03%) | 12 (5.31%) | 19 (4.20%) | 12 (2.65%) | ||||
| White | 138 (71.88%) | 125 (62.81%) | 160 (70.80%) | 320 (70.80%) | 332 (73.45%) | ||||
| Don't know | 3 (1.56%) | 2 (1.01%) | 2 (0.88%) | 0 (0.00%) | 3 (0.66%) | ||||
| Refused | 2 (1.04%) | 2 (1.01%) | 3 (1.33%) | 7 (1.55%) | 4 (0.88%) | ||||
| Ethnicity, n (%) | |||||||||
| Hispanic, Latino/a, or Spanish origin | 27 (14.06%) | 34 (17.09%) | 29 (12.83%) | 39 (8.63%) | 45 (9.96%) | 20.498 | 8 | 0.009 | 0.082 |
| Not of Hispanic, Latino/a, or Spanish origin | 159 (82.81%) | 163 (81.91%) | 192 (84.96%) | 408 (90.27%) | 404 (89.38%) | ||||
| Don´t know, refused, or missing | 6 (3.13%) | 2 (1.01%) | 5 (2.21%) | 5 (1.11%) | 3 (0.66%) | ||||
| Non‐English‐speaking background, n (%) | 12 (6.3%) | 12 (6.0%) | 11 (4.9%) | 2 (0.4%) | 3 (0.7%) | 38.499 | 4 | <0.001 | 0.159 |
| Obesity (n %) | 50 (28.2%) | 59 (33.5%) | 67 (32.7%) | 132 (30.1%) | 119 (29.6%) | 1.833 | 4 | 0.766 | 0.036 |
| High cholesterol, n (%) | 45 (30.8%) | 49 (31.6%) | 53 (30.6%) | 108 (33.6%) | 94 (27.8%) | 2.686 | 4 | 0.612 | 0.049 |
| Stroke, n (%) | 7 (3.6%) | 6 (3.1%) | 11 (4.9%) | 12 (2.7%) | 13 (2.9%) | 2.814 | 4 | 0.589 | 0.043 |
| Hypertension, n (%) | 58 (30.2%) | 68 (34.5%) | 61 (27.2%) | 147 (32.7%) | 117 (25.9%) | 7.901 | 4 | 0.095 | 0.072 |
| Use of antihypertensive medication, n (%) | 32 (55.2%) | 47 (69.1%) | 39 (63.9%) | 89 (61.0%) | 71 (60.7%) | 2.8443 | 4 | 0.584 | 0.079 |
| Diabetes, n (%) | 21 (11.0%) | 23 (11.6%) | 31 (13.8%) | 39 (8.6%) | 26 (5.8%) | 14.247 | 4 | 0.007 | 0.097 |
| Heart problems | 7 (3.6%) | 5 (2.5%) | 7 (3.1%) | 8 (1.8%) | 4 (0.9%) | 7.212 | 4 | 0.125 | 0.069 |
| Heart attack, n (%) | 11 (5.8%) | 11 (5.6%) | 13 (5.9%) | 18 (4.0%) | 7 (1.6%) | 11.926 | 4 | 0.018 | 0.089 |
| Angina and coronary heart disease, n (%) | 11 (5.8%) | 11 (5.6%) | 11 (5.0%) | 7 (1.6%) | 8 (1.8%) | 16.838 | 4 | 0.002 | 0.106 |
| Depressive Disorder, n (%) | 58 (30.7%) | 78 (39.8%) | 112 (49.8%) | 66 (14.7%) | 109 (24.2%) | 109.647 | 4 | <0.001 | 0.27 |
| Kidney disorder, n (%) | 9 (4.7%) | 6 (3.0%) | 18 (8.0%) | 6 (1.3%) | 5 (1.1%) | 32.100 | 4 | <0.001 | 0.145 |
| Smoking status, n (%) | |||||||||
| Smoking current | 34 (18.0%) | 42 (21.6%) | 35 (16.0%) | 72 (16.3%) | 62 (14.3%) | 18.979 | 8 | 0.015 | 0.08 |
| Former | 34 (18.0%) | 35 (18.0%) | 47 (21.5%) | 119 (27.0%) | 80 (18.5%) | ||||
| Never | 121 (64.0%) | 117 (60.3%) | 137 (62.6%) | 250 (56.7%) | 291 (67.2%) | ||||
| Alcohol consumption, n (%) | |||||||||
| Heavy drinker | 18 (10.1%) | 11 (5.7%) | 20 (9.3%) | 44 (10.3%) | 19 (4.5%) | ||||
| Moderate drinker | 73 (41.0%) | 71 (36.8%) | 8 | 0.015 | 202 (47.8%) | 24.927 | 8 | 0.002 | 0.093 |
| Abstainer | 87 (48.9%) | 111 (57.5%) | 110 (51.4%) | 182 (42.7%) | 202 (47.8%) | ||||
| >5 servings of fruits or vegetables | 25 (14.6%) | 33 (18.4%) | 34 (17.5%) | 53 (13.7%) | 72 (18.8%) | 4.814 | 4 | 0.307 | 0.061 |
| Physical activity, n (%) | |||||||||
| Highly active | 54 (30.3%) | 65 (35.3%) | 8 | 0.002 | 147 (36.2%) | ||||
| Active | 35 (19.7%) | 29 (15.8%) | 34 (16.3%) | 63 (15.2%) | 77 (19.0%) | 23.881 | 12 | 0.021 | 0.076 |
| Insufficiently active | 29 (16.3%) | 37 (20.1%) | 36 (17.3%) | 82 (19.8%) | 85 (20.9%) | ||||
| Inactive | 60 (33.7%) | 53 (28.8%) | 74 (35.6%) | 96 (23.2%) | 97 (23.9%) | ||||
Note: The P‐value refers to the chi‐square test result.
Abbreviations: NS, non‐significant; SD, standard deviation.
4. DISCUSSION
This study is the first to estimate both sex and gender differences in mid‐life and late‐life dementia and AD risk using established risk score algorithms. In line with previous research, sex differences were observed in both mid‐life and late‐life dementia and AD risk, whereby those assigned male at birth had higher mid‐life risk and lower late‐life risk compared to those assigned female at birth. While sample sizes were small for gender comparisons, we offer preliminary evidence for gender identity differences in overall late‐life AD risk estimated using the ANU‐ADRI risk score, but not mid‐life dementia risk estimated using partial CAIDE or LIBRA scores. Of particular importance, we also present preliminary evidence for health disparities in specific modifiable risk factors for dementia and AD that are likely to be disproportionately impacting the dementia and AD risk of underserved gender groups, including transgender men, transgender women, and non‐binary older adults.
4.1. Differences in risk by sex
The first study objective was to characterize how those assigned male and female at birth differ in mid‐life and late‐life dementia and AD risk. In line with previous research, 17 , 18 those assigned male showed higher mid‐life dementia risk, and lower late‐life AD risk compared to those assigned female. The presence of robust sex differences in risk is notable given that in traditional administrations of the AD risk tools, age and sex are weighted in the risk score and are typically strong drivers of risk score outcomes, whereas in this study, sex was a between‐group variable of interest and participants were case‐control matched based on age and race across sex groups.
Overall, females had fewer dementia‐related medical risk factors including heart problems, stroke, and obesity. These results are congruent with the literature in which heart problems including coronary heart disease, myocardial infarction, and heart failure are reported to be higher in males. 19 In this study, higher rates of hypertension were reported by males, and this might be due to low use of antihypertensive medications. Females had a higher prevalence of depression compared to males, which is in line with prior research. 6
Contrary to what has been reported elsewhere, a higher proportion of females in this sample was college‐educated compared to males. We note that education attainment has been higher for females than males in younger generations and is speculated to be a reason the incidence of dementia is decreasing more in females than males in some countries. 20
In line with previous research, a higher proportion of females reported healthy lifestyle habits such as not smoking, not drinking heavily, and consuming higher amounts of fruits and vegetables. 21 , 22 Males were more likely to be physically active than females, which is also in line with global evidence. 23
4.2. Differences in risk by gender identity
Gender differences are reported for late‐life AD risk, but not mid‐life dementia risk. Transgender men, transgender women, and non‐binary adults were each found to have significantly higher overall late‐life risk scores compared to cisgender men and women. This result supports the notion that both sex and gender impact overall AD risk. While the results are preliminary, they may indicate that having a lived experience of incongruence between sex and gender norms may be a social determinant of AD and dementia risk. We speculate that societal stigma and discrimination experienced by TGD adults is likely an important contributor to this effect.
Findings reveal several differences among gender identity groups in individual risk factors for dementia and AD. One of the most concerning differences among groups was found for depression, which was reported as substantially more common among non‐binary adults compared to transgender women, cisgender men, and cisgender women. Transgender men reported the second highest rate of depression, which was also significantly higher than cisgender men and women. Depression in mid‐life has been shown to be associated with up to a 70% increase in risk of AD dementia. 24 Transgender women have previously been shown to experience depression much more than in the general population, that is, estimates of lifetime prevalence of depression in transgender women is as high as 62% 25 while the lifetime depression rate for the general US population is 16.6%. 26 The present study offers compelling evidence for significant mental health disparities experienced by specific gender identities, and it is notable that the most impacted groups (non‐binary adults and transgender men) are also the least studied.
Another concerning finding was the significantly higher rates of kidney problems among non‐binary adults compared to both cisgender men and cisgender women. This finding warrants further research to see if it is a reliable finding.
4.3. Limitations and future directions
There are several important limitations that must be considered when interpreting the results of this study. First, the data are drawn from only those states that collected data on sex and gender required for analysis. Thus, the results cannot be generalized to the broader US population. The average age of our matched samples was relatively young and the education level of females was also marginally higher than what might be expected in the general population. 27 Further, while the BRFSS reports comparable overall response rates to other similar studies 28 it is not clear whether response rates may differ across transgender and cisgender groups, or by other characteristics that may limit the representativeness of the data.
There are also implications stemming from the specific questions that were used to capture sex and gender identity, and the order of their presentation. This study does not address the impact of gender norms or gender relations on dementia risk. Sex and gender questions were not co‐located in the survey, which likely impacted reporting of sex among TGD respondents. Non‐binary adults were first required to report that they were transgender before being presented with a non‐binary option, which may have impacted reporting. These points are further addressed in the supporting information. However, these limitations do not detract from the findings of this paper, which provide the first evidence regarding overall dementia and AD risk that is specific to transgender and non‐binary adults.
In this study, we use a combined race and ethnicity variable (provided by the Centers for Disease Control and Prevention in their publicly available 2019 data set) for sample matching prior to sex‐at‐birth analyses. Future research should attempt to disentangle the social constructs of race and ethnicity, and endeavor to be inclusive of less frequently reported identities wherever possible. 29
We also note that some important modifiable risk factors for dementia and AD that may be disproportionately impacting TGD individuals could not be included in the risk score algorithms because they are not measured by the BRFSS survey. For example, low social engagement is an important risk factor that may be especially prevalent among TGD groups, caused at least in part by social stigma. More than half of transgender adults aged 55+ reported losing close friends and 40% losing contact with children due to their gender identity, 30 and almost half of these older adults live alone compared to less than one fifth of the general population. 31 It is also possible that strong social ties may be particularly protective among TGD adults. Data are needed to explore these hypotheses.
Other factors not accounted for in risk score algorithms may have a significant and disproportionate effect on dementia risk for TGD populations. For instance, increasing evidence suggests that gender‐affirming hormone therapies may increase cardiometabolic risk, 32 , 33 , 34 which could in turn contribute to a higher risk of dementia. Additionally, transgender and gender diverse individuals often face obstacles in accessing health care, which could compound these risks. 35 Improving health‐care access for this population is a pressing issue, and health‐care providers must enhance their knowledge of the unique health‐care needs of TGD individuals. Future research should investigate these factors and their impact on dementia risk in this population.
Finally, while the risk score algorithms applied in this study have been well‐validated among cisgender adults, none to date have been validated among transgender or gender diverse groups. To do so, we need to follow these cohorts into older age to obtain information on incident dementia. We acknowledge that this will be an important future step toward better understanding dementia and AD risk among TGD populations.
Despite the limitations, this study contributes new evidence regarding sex and gender differences in dementia and AD risk that should not be overlooked. The results of this study underscore the need to consider distinct sex and gender pathways in epidemiological research and public health surveillance. Inclusive large‐scale data sets are desperately needed to extend research in the area, and these data sets should follow best practice approaches to collecting data on sex and gender.
5. CONCLUSIONS
Sex and gender have unique relationships to mid‐life and late‐life dementia and AD risk. Future research is needed to build the evidence base for gender differences in specific risk factors, particularly those that may be contributing to higher overall risk among understudied and underserved gender groups.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest. Author disclosures are available in the supporting information.
CONSENT STATEMENT
All participants of the 2019 Behavioral Risk Factor Surveillance System telephone survey provided informed consent.
Supporting information
Supporting Information
Supporting Information
ACKNOWLEDGMENTS
All data used in this study are made available for download by the Centers for Disease Control and Prevention (https://www.cdc.gov/brfss/annual_data/annual_2019.html). B.B., L.Z., S.K., and K.J.A. developed the analysis plan and curated the list of modifiable risk factors and risk score algorithms. L.Z. conducted much of the analysis and B.B. accessed and verified the data. S.K. and K.J.A. reviewed findings and provided feedback on additional analyses. B.B. prepared the manuscript and all authors provided critical revisions. All authors had full access to the data and responsibility for the decision to submit for publication. B.B., L.Z., and K.J.A. are funded by Australian Research Council Grant FL190100011 and S.K. is funded by NHMRC Grant GNT1171279. K.J.A. is also supported by the Australian Research Council Centre for Excellence in Population Ageing Research.
Open access publishing facilitated by University of New South Wales, as part of the Wiley ‐ University of New South Wales agreement via the Council of Australian University Librarians.
Brady B, Zheng L, Kootar S, Anstey KJ. Sex and gender differences in risk scores for dementia and Alzheimer's disease among cisgender, transgender, and non‐binary adults. Alzheimer's Dement. 2024;20:5–15. 10.1002/alz.13317
[Correction added on August 16, 2023, after first online publication: The position of the first and last names of the authors had been in reverse order on the byline and this has been corrected.]
REFERENCES
- 1. Rajan KB, Weuve J, Barnes LL, McAninch EA, Wilson RS, Evans DA. Population estimate of people with clinical Alzheimer's disease and mild cognitive impairment in the United States (2020‐2060). Alzheimers Dement. 2021;17(12):1966‐1975. doi: 10.1002/alz.12362 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413‐446. doi: 10.1016//S0140-6736(20)30367-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Schiebinger L, Klinge I, Paik HY, et al. Gender. Gendered Innovations. Accessed November 9, 2022. http://genderedinnovations.stanford.edu/terms/gender.html [Google Scholar]
- 4. Nianogo RA, Rosenwohl‐Mack A, Yaffe K, Carrasco A, Hoffmann CM, Barnes DE. Risk factors associated with Alzheimer disease and related dementias by sex and race and ethnicity in the US. JAMA Neurol. 2022;79(6):584‐591. doi: 10.1001/jamaneurol.2022.0976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Gilsanz P, Mayeda ER, Glymour MM, et al. Female sex, early‐onset hypertension, and risk of dementia. Neurology. 2017;89(18):1886‐1893. doi: 10.1212/WNL.0000000000004602 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kessler RC, McGonagle KA, Swartz M, Blazer DG, Nelson CB. Sex and depression in the National Comorbidity Survey. I: lifetime prevalence, chronicity and recurrence. J Affect Disord. 1993;29(2‐3):85‐96. doi: 10.1016/0165-0327(93)90026-g [DOI] [PubMed] [Google Scholar]
- 7. Neu SC, Pa J, Kukull W, et al. Apolipoprotein E genotype and sex risk factors for Alzheimer disease: a meta‐analysis. JAMA Neurol. 2017;74(10):1178‐1189. doi: 10.1001/jamaneurol.2017.2188 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Mielke MM. Sex and gender differences in Alzheimer's disease dementia. Psychiatr Times. 2018;35(11):14‐17. [PMC free article] [PubMed] [Google Scholar]
- 9. Flatt JD, Cicero EC, Lambrou NH, et al. Subjective cognitive decline higher among sexual and gender minorities in the United States, 2015‐2018. Alzheimers Dement (NY). 2021;7(1):e12197. doi: 10.1002/trc2.12197. Published July 28, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Pharr JR, Batra K. Propensity score analysis assessing the burden of non‐communicable diseases among the transgender population in the United States using the behavioral risk factor surveillance system (2017‐2019). Healthcare (Basel). 2021;9(6):696. doi: 10.3390/healthcare9060696 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Mauvais‐Jarvis F, Bairey Merz N, Barnes PJ, et al. Sex and gender: modifiers of health, disease, and medicine. Lancet. 2020;10250(396):565‐582. doi: 10.1016/S0140-6736(20)31561-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Rocca WA, Mielke MM, Vemuri P, Miller VM. Sex and gender differences in the causes of dementia: a narrative review. Maturitas. 2014;2(79):196‐201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Centers for Disease Control and Prevention . Behavioral Risk Factor Surveillance System: OVERVIEW: BRFSS 2019. Published July 26, 2019. Accessed November 9, 2022. https://www.cdc.gov/brfss/annual_data/2019/pdf/overview‐20ss19‐508.pdf
- 14. Anstey KJ, Cherbuin N, Herath P, et al. A self report risk index to predict occurence of dementia in three independent cohorts of older adults: the ANU‐ADRI. PLoS ONE. 2014;9:e86141. doi: 10.1371/journal.pone.0086141 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Exalto LG, Quesenberry CP, Barnes D, Kivipelto M, Biessels GJ, Whitmer RA. Midlife risk score for the prediction of dementia four decades later. Alzheimers Dement. 2013;10:562‐570. doi: 10.1016/j.jalz.2013.05.1772 [DOI] [PubMed] [Google Scholar]
- 16. Deckers K, Barbera M, Köhler S, et al. Long‐term dementia risk prediction by the LIBRA score: a 30‐year follow‐up of the CAIDE study. Int J Geriatr Psychiatry. 2020;35:195‐203. doi: 10.1002/gps.5235 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Gong J, Harris K, Peters SAE, Woodward M. Sex differences in the association between major cardiovascular risk factors in midlife and dementia: a cohort study using data from the UK Biobank. BMC Med. 2021;19(1):110. doi: 10.1186/s12916-021-01980-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Prince MJ, Wimo A, Guerchet MM, et al. World Alzheimer Report 2015: The Global Impact of Dementia: An Analysis of Prevalence, Incidence, Cost and Trends . Published August 2015. Accessed November 9, 2022. https://www.alzint.org/u/WorldAlzheimerReport2015.pdf
- 19. Kim MY, Kim K, Hong CH, Lee SY, Jung YS. Sex differences in cardiovascular risk factors for dementia. Biomol Ther (Seoul). 2018;26(6):521‐532. doi: 10.4062/biomolther.2018.159 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Langa KM, Larson EB, Crimmins EM, et al. A comparison of the prevalence of dementia in the United States in 2000 and 2012. JAMA Intern Med. 2017;177(1):51‐58. doi: 10.1001/jamainternmed.2016.6807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. World Health Organization . Global Status Report on Alcohol and Health 2018. Published 2018. Accessed November 9, 2022. https://apps.who.int/iris/bitstream/handle/10665/274603/9789241565639‐eng.pdf
- 22. Shakersain B, Santoni G, Larsson SC, et al. Prudent diet may attenuate the adverse effects of Western diet on cognitive decline. Alzheimers Dement. 2016;12(2):100‐109. doi: 10.1016/j.jalz.2015.08.002 [DOI] [PubMed] [Google Scholar]
- 23. Althoff T, Sosič R, Hicks JL, King AC, Delp SL, Leskovec J. Large‐scale physical activity data reveal worldwide activity inequality. Nature. 2017;547(7663):336‐339. doi: 10.1038/nature23018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. McEwen BS, Gianaros PJ. Central role of the brain in stress and adaptation: links to socioeconomic status, health, and disease. Ann N Y Acad Sci. 2010;1186:190‐222. doi: 10.1111/j.1749-6632.2009.05331.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Clements‐Nolle K, Marx R, Guzman R, Katz M. HIV prevalence, risk behaviors, health care use, and mental health status of transgender persons: implications for public health intervention. Am J Public Health. 2001;91(6):915‐921. doi: 10.2105/ajph.91.6.915 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age‐of‐onset distributions of DSM‐IV disorders in the National Comorbidity Survey Replication [published correction appears in Arch Gen Psychiatry. Arch Gen Psychiatry. 2005;62(6):593‐602. doi: 10.1001/archpsyc.62.6.593 [DOI] [PubMed] [Google Scholar]
- 27. U.S. Census Bureau . Educational Attainment in the United States: 2021. [online] Available at: https://www.census.gov/data/tables/2021/demo/educational‐attainment/cps‐detailed‐tables.html
- 28. Centre for Disease Control . Behavioral Risk Factor Surveillance System. 2020. 201 9 Summary Data Quality Report. Available at: www.cdc.gov/brfss/annual_data/2019/pdf/2019‐sdqr‐508.pdf
- 29. Flanagin A, Frey T, Christiansen SL, AMA Manual of Style Committee . Updated guidance on the reporting of race and ethnicity in medical and science journals. JAMA. 2021;326(7):621‐627. doi: 10.1001/jama.2021.13304 [DOI] [PubMed] [Google Scholar]
- 30. Witten TM. It's not all darkness: robustness, resilience, and successful transgender aging. LGBT Health. 2014;1(1):24‐33. [DOI] [PubMed] [Google Scholar]
- 31. Jaime MG, Motter LA, Tanis J. Injustice at every turn: a report of the national transgender discrimination survey. Washington: National Centre for Transgender Equality and National Gay and Lesbian Task Force. 2011. Available at: www.onlabor.org/wp‐content/uploads/2016/04/ntds_full.pdf [Google Scholar]
- 32. Swe NC, Ahmed S, Eid M, Poretsky L, Gianos E, Cusano NE. The effects of gender‐affirming hormone therapy on cardiovascular and skeletal health: a literature review. Metabolism Open. 2022;13:100173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Dutra E, Lee J, Torbati T, Garcia M, Merz CN, Shufelt C. Cardiovascular implications of gender‐affirming hormone treatment in the transgender population. Maturitas. 2019;129:45‐49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Aranda G, Halperin I, Gomez‐Gil E, et al. Cardiovascular risk associated with gender affirming hormone therapy in transgender population. Front Endocrinol (Lausanne). 2021;12:718200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Safer JD, Coleman E, Feldman J, et al. Barriers to health care for transgender individuals. Curr Opin Endocrinol Diabetes Obes. 2016;23(2):168. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information
Supporting Information