RACIAL AND ETHNIC DIVERSITIES IN COGNITIVE HEALTH APPRAISALS: FINDINGS FROM THE HCAP AGE MODIFIES THE ASSOCIATION SEX DIFFERENCES IN SPEECH AND HIGH-FREQUENCY HEARING LOSS ASSOCIATION WITH COGNITIVE IMPAIRMENT AMONG OLDER ADULTS

and in racial and ethnic subgroups to test direct and interactive effects of cognitive performance, indicated by the Mini-Mental State Examination (MMSE). The regression model for the entire sample showed direct effects of cognitive performance and race/ethnicity on cognitive health appraisals, as well as a significant interaction between cognitive performance and being non-Hispanic Black. Cognitive performance and cognitive health appraisals were positively associated in non-Hispanic Whites but not significantly associated in non-Hispanic Blacks. Our subsequent analysis within each racial/ ethnic group showed that the effect of cognitive performance in non-Hispanic Blacks and Hispanics became either reversed or nonsignificant when background and health variables were considered. Modification by age or chronic medical conditions in each racial and ethnic group was also observed. Overall, these findings suggest that perceptions and appraisals of cognitive health vary by race and ethnicity and hold implications for how these differences should be considered in research and practice with diverse groups of older adults. routine/ manual) depressive status. Effect modification by retirement (early, on-time, late). Results: High education (EF: b=0.37, SE=0.08, p<.001; to the relationship between speech-frequency hearing loss (SFHL), high-frequency hearing loss (HFHL), and cognitive impairment (CI). to determine whether there any This 1,012 adults aged male; average age, 72.61±5.51 years). The participants had their hearing and cognition measured using pure tone audiometry and Mini Mental State Examination (MMSE), respectively. We used the audiometric definition of hearing loss (HL) adopted by the World Health Organization (WHO). Speech-frequencies were measured as 0.5 kHz, 1 kHz, 2 kHz, and 4 kHz; high-frequencies were measured as 4 kHz and 8 kHz. Pure tone average (PTA) was measured as hearing sensitivity. Results: Our studies demon-strated a 37.6% prevalence of HL in males and a 36.0% prevalence of HL in females. Adjusted for confounding variables, the results from a multivariate analysis showed that SFHL was associated with CI in females (OR=2.400, 95% Confidence Interval=1.313–4.385) and males (OR=2.189, 95% Confidence Interval=0.599–2.944). However, HFHL was associated with CI only in females (OR=2.943, 95% Confidence Interval=1.505–5.754). HL was associated with poorer cognitive scores (P<0.05). “Registration” (P<0.05) in MMSE was associated with speech and high-frequency hearing sensitivity. Conclusion: The associations between HL and CI varied according to gender in older community-dwellers, suggesting that different mechanisms are involved in the etiology of HL. Moreover, hearing sensitivity was negatively associated with cognition scores; therefore, early screening for HL and CI among older community-dwelling adults is advised.

we selectively attend to can be driven by the value we ascribe to the information, often based on stimulus factors such as perceptual features that make the information stand out, or conceptual features that make it easy to group information.The current study investigated whether behavioral measures of value-directed strategic processing are differentially affected when value is defined by perceptual versus conceptual features, and how normal cognitive aging impacts processing.Cognitively normal younger (N = 16; mean age: 22.1 ± 2.9 years) and older adults (N = 16; mean age: 66.9 ± 7.3 years) completed two value-directed strategic processing tasks, where value was defined by either perceptual (i.e., uppercase and lowercase letters; Letter Case task) or conceptual (i.e., animals and household items; Categories task) features.Both groups had higher recall on the Categories task compared to the Letter Case task, and higher recall for highthan low-value words.However, older adults recalled fewer total words than younger adults, but the groups did not differ across task types.These findings indicate that manipulating perceptual and/or conceptual features to define value can be used to study value-directed strategic processing in younger and older adults.Furthermore, grouping information based on conceptual features may be more effective for promoting subsequent recall in both younger and older adults.

PHYSICAL ACTIVITY AND EPISODIC MEMORY: AN ANALYSIS OF LONGITUDINAL ASSOCIATIONS OVER 12 YEARS
Rio Tate, 1 and Gizem Hueluer, 2 1.University of South Florida, Tampa, Florida, United States, 2. University of Bonn (Germany) With the increasing prevalence of Alzheimer's and related dementias, it is becoming a growing public health concern to identify modifiable risk factors to prevent cognitive decline.Previous research suggest that physical exercise may promote cognitive function in aging.However, most of this research is based on experimental or cross-sectional studies and fewer studies have studied longitudinal associations over longer time frames.In the present study, we examined how physical activity is related to cognition in older adults.To do so, we applied multilevel models to data from 29,740 participants (age at baseline: M = 63 years, SD = 11 years, 50 to 102 years) from years 2004-2016, measured biennially (waves 7-13) of the Health and Retirement Study.Our findings showed that at the between-person level, those who practiced light, moderate, and vigorous-physical activity more frequently than others showed higher levels of episodic memory than others at the age of 70.At the within-person level, participants performed better than usual on a test of episodic memory on occasions when they reported more frequent physical activity of light, moderate and vigorous intensity.More frequent vigorous physical activity was related to less age-related decline in episodic memory, while the frequency of moderate physical exercise was unrelated to age-related change in memory.Contrary to our expectations, more frequent light physical exercise was related to more age-related decline in episodic memory.We discuss the implications of these findings for future research and practice.Yuri Jang, 1 Eunyoung Choi, 2 Yujin Franco, 3  Nan Sook Park, 4 David Chiriboga, 4 and Miyong T. Kim, 5 1.University of Southern California,Los Angeles,California,United States,2. University of Southern California,LA,California,United States,3. University of Southern California,University of Southern California,California,United States,4. University of South Florida,Tampa,Florida,United States,5

. University of Texas at Austin, Austin, Texas, United States
The aims of the study were: (1) to investigate the relationship between cognitive performance and cognitive health appraisals across non-Hispanic White, non-Hispanic Black, and Hispanic older adults in the United States and (2) to explore within-group variations by examining interactions between cognitive performance and background and health variables.The sample (N = 3,099) included 2,260 non-Hispanic White, 498 non-Hispanic Black, and 341 Hispanic adults aged 65 or older, from the 2016-2017 Harmonized Cognitive Assessment Protocol.Regression models of cognitive health appraisals, indicated by self-rated cognitive health, were examined in the entire sample and in racial and ethnic subgroups to test direct and interactive effects of cognitive performance, indicated by the Mini-Mental State Examination (MMSE).The regression model for the entire sample showed direct effects of cognitive performance and race/ethnicity on cognitive health appraisals, as well as a significant interaction between cognitive performance and being non-Hispanic Black.Cognitive performance and cognitive health appraisals were positively associated in non-Hispanic Whites but not significantly associated in non-Hispanic Blacks.Our subsequent analysis within each racial/ ethnic group showed that the effect of cognitive performance in non-Hispanic Blacks and Hispanics became either reversed or nonsignificant when background and health variables were considered.Modification by age or chronic medical conditions in each racial and ethnic group was also observed.Overall, these findings suggest that perceptions and appraisals of cognitive health vary by race and ethnicity and hold implications for how these differences should be considered in research and practice with diverse groups of older adults.

RETIREMENT AGE MODIFIES THE ASSOCIATION BETWEEN EDUCATION/OCCUPATION AND COGNITION
Monica Nelson, 1 and Ross Andel, 2 1.University of South Florida, Brandon, Florida, United States, 2. University of South Florida, Tampa, Florida, United States Introduction: According to the cognitive reserve and useit-or-lose-it hypotheses, engagement in stimulating activities seems to benefit cognition, with engagement often associated with more education or higher occupational position.However, whether retirement may modify the association between education/occupation and cognition is unclear.We aimed to assess how age at retirement may modify the relationship between education/occupation and cognition.Methods: Older adults (n=360) from the Alzheimer's Disease Neuroimaging Initiative who were cognitively normal and retired at baseline participated.Linear regression was used to assess how educational attainment (high/low) or occupational position (managerial, intermediate/clerical, routine/ manual) related to executive functioning (EF) or memory, controlling for age, sex, depressive symptoms, and health status.Effect modification by retirement (early, on-time, late).Results: High education (EF: b=0.37,SE=0.08, p<.001; memory: b=0.22,SE=0.05, p<.001), intermediate (EF: b=0.26,SE=0.11,p=.019;memory: b=0.18,SE=0.08,p=.018) and managerial (EF: b=0.23,SE=0.12,p=.045;memory: b=0.16,SE=0.08,p=.045)occupations (compared to routine/manual occupations) were associated with better EF and memory performance.High education was significantly associated with better EF and memory for participants who retired early (EF: b=0.43,SE=0.12, p<.001; memory: b=0.29,SE=0.10, p=.004) or on-time (EF: b=0.51,SE=0.15, p=.001; memory: b=0.24,SE=0.10, p=.014), but not for participants who retired late (EF: b=0.19,SE=0.15,p=.200;memory: b=0.09,SE=0.09,p=.334).Intermediate occupations were associated with EF only for participants who retired on-time (b=0.58,SE=0.21, p=.007).Conclusion: Education and occupational position may influence cognition after retirement differently based on retirement timing, with effects most apparent for on-time retirement and substantially reduced for late retirement.

SEX DIFFERENCES IN SPEECH AND HIGH-FREQUENCY HEARING LOSS ASSOCIATION WITH COGNITIVE IMPAIRMENT AMONG OLDER ADULTS
Wang Jingru, Yu Ying, and Guo Qi, Shanghai University of Medicine and Health Sciences, Shanghai, Shanghai, China (People's Republic) Objectives: The purpose of this study was to investigate the relationship between speech-frequency hearing loss (SFHL), high-frequency hearing loss (HFHL), and cognitive impairment (CI).Then to determine whether there are any differences in gender among older community dwellers in China.Methods: This study involved 1,012 adults aged ≥60 years (428 male; average age, 72.61±5.51years).The participants had their hearing and cognition measured using pure tone audiometry and Mini Mental State Examination (MMSE), respectively.We used the audiometric definition of hearing loss (HL) adopted by the World Health Organization (WHO).Speech-frequencies were measured as 0.5 kHz, 1 kHz, 2 kHz, and 4 kHz; high-frequencies were measured as 4 kHz and 8 kHz.Pure tone average (PTA) was measured as hearing sensitivity.Results: Our studies demonstrated a 37.6% prevalence of HL in males and a 36.0%prevalence of HL in females.Adjusted for confounding variables, the results from a multivariate analysis showed that SFHL was associated with CI in females (OR=2.400,95% Confidence Interval=1.313-4.385)and males (OR=2.189,95% Confidence Interval=0.599-2.944).However, HFHL was associated with CI only in females (OR=2.943,95% Confidence Interval=1.505-5.754).HL was associated with poorer cognitive scores (P<0.05)."Registration" (P<0.05) in MMSE was associated with speech and high-frequency hearing sensitivity.Conclusion: The associations between HL and CI varied according to gender in older communitydwellers, suggesting that different mechanisms are involved in the etiology of HL.Moreover, hearing sensitivity was negatively associated with cognition scores; therefore, early screening for HL and CI among older community-dwelling adults is advised.

TELEPHONE-BASED WORD LIST RECALL AND HEARING ABILITY
Taylor Atkinson, and Ross Andel, University of South Florida, Tampa, Florida, United States Certain consonant sounds called fricatives (e.g, "s" and "f") are difficult to hear over the telephone; phones exclude high-frequency sounds that affect their intelligibility.This may be problematic for older adults responding to phonebased memory tests.Many older adults have some degree of hearing loss, and older men have it more in the highfrequency range.Hearing loss, in combination with phone bandwidth restrictions, may reduce older adults' recall of fricative words.Participants (n=3,612, mean age=64.2,60% women) in the 1998 wave of the Health and Retirement study (HRS) completed a word list immediate recall task over the phone.List 4 recall was examined because it was evenly split (5 each) between words with and without fricative consonant sounds.Subjective ratings of hearing and health, age, depression, and education were also measured.A Wilcoxon signed-rank test showed participants recalled fewer fricative (M=2.8)than nonfricative (M=3.0)words, Z=-8.47, p<.001.An ordinal regression for fricative word recall indicated a sex by hearing interaction; males with worse hearing were less likely to recall more fricative words, OR=.94, 95% CI [.88, 1.01], p=.076, after controlling for age, education, health, and depression.An ordinal regression for nonfricative word recall did not show a main effect for hearing or a hearing by sex interaction.For both models, age, education, and health were related to recall.Consonant sounds may influence phone-based word recall, particularly for older men.Attention should be paid to word selection when designing phone-based cognitive tests in order to avoid memory impairment overestimation.