Abstract
Despite agreement among stakeholders that senior centers can promote physical and mental health, research on senior center use in urban populations is limited. Our objective was to describe demographic and health factors associated with senior center use among urban, low-income older adults in order to inform programming and outreach efforts. We used data from a 2009 telephone survey of 1036 adults randomly selected from rosters of New York City public housing residents aged 65 and older. We analyzed senior center use by race/ethnicity, age, gender, health, housing type, and income, and used a forward selection approach to build best-fit models predicting senior center use. Older adults of all ages and of both genders reported substantial use of senior centers, with nearly one third (31.3%) reporting use. Older adults living alone, at risk of depression, or living in specialized senior housing had the greatest use of centers. Senior center use varied by race/ethnicity, and English-speaking Hispanics had a higher prevalence of use than Spanish-speaking Hispanics (adjusted prevalence ratio [PR]=1.69, 95% CI: 1.11-2.59). Spanish-speaking communities and older adults living in non-senior congregate housing are appropriate targets for increased senior center outreach efforts.
Keywords: Aging, Community-Based Services, Mental Health, Older Adults, Public Housing, Racial and Ethnic Diversity, Senior Centers, Service Utilization
Senior centers are places where older adults can go for “support, engagement, and basic needs”.1 Centers are a potential mechanism for linking seniors to resources and services, and for providing attendees with meals and opportunities for socialization, exercise, and education.2 There is broad support for senior centers among health professionals, community leaders, and older adults themselves.1,3–5 From an assessment of New York City senior centers, stakeholders from different arenas identified the following list as “critical and core” functions of senior centers: 1) providing opportunities for social engagement; 2) providing a link to public services and benefits; 3) providing a link to community resources; 4) providing nutritional support; and 5) promoting physical health, mental health, and healthy behaviors.1 Serving poor and minority older adults is a priority for senior centers.1 Ultimately, the impact of senior centers depends both on the benefits for attendees and on the number of seniors who attend.6
Several prior studies have characterized senior center users using data from local and nationwide surveys. Most studies have identified lower income7–9 and older age range (75–84)7,10,11 as factors associated with senior center use. Women have been found to be more likely to attend senior centers than men7,11,12 and to have a more positive perception of senior center benefits.13 Some studies have found that Blacks and other minorities are more likely to use senior centers than Whites, but results have not been consistent, potentially reflecting different study settings and designs.7–9,11 Data are also inconsistent regarding whether having more5,7 or fewer14 social contacts predicts senior center use. Some studies identified that older adults with no or few functional limitations were more likely to attend centers than those with more limitations.7,8,10,15 Other identified predictors of senior center use have included knowledge about and positive attitude towards senior centers,14 friendships with senior center users,16 physical proximity, and transportation access.16,17
In New York City (NYC), there are approximately 300 city-administered senior centers operated by different organizations that vary in size, range of services and programs offered, staffing, and populations served.18,19 In 2009, many (90+) senior centers were located on public housing premises, and some (40) were funded and staffed by the NYC Housing Authority (NYCHA). NYCHA older adult residents are a racially diverse low- and moderate-income population who comprise more than 10 % of all of NYC’s low-income older adults. The proportion of NYCHA residents who are over age 65 is expected to grow over the next few decades due to a low rate of residential turnover.18 NYCHA older adults have poor health compared with NYC older adults overall, including higher rates of chronic conditions and self-reported poor or fair health,20 making senior center services particularly important for this population. The main goals of this research were to characterize senior center use in this population and to determine primary predictors of senior center use, overall and across different racial/ethnic and language groups. Improving our understanding of the predictors of senior center use among older public housing residents can help focus programming and outreach in order for senior centers to better serve poor and minority urban older adults.
Data and Methods
Survey Design and Study Population
The 2009 NYCHA Senior Survey was a telephone survey of NYCHA residents ages 65 and older that contained questions relating to several aspects of older adult health.18 Residents were randomly selected from the NYCHA administrative tenant data system (TDS) which includes all authorized residents and contained valid records and telephone numbers for 58,484 residents aged 65 and older as of June 2009. The survey was conducted over a two week period in June 2009 by trained interviewers at the Baruch Survey Research Unit at the City University of New York. The response rate was 34.7 % and the cooperation rate was 93.4 %, yielding 1036 completed surveys.18 The survey was administered in English, Spanish, Russian, and Chinese. Survey respondents were significantly less likely than non-respondents to be male, Hispanic, and Manhattan residents. Data were weighted to be representative of the NYCHA older adult population on gender, household income, borough, age, and race/ethnicity. After weighting, survey respondents did not differ from non-respondents on the following characteristics captured in the TDS system: household size; type of housing development; rates of mobility; and vision, mental, or hearing disabilities. Survey respondents were slightly less likely to use a wheelchair than non-respondents (3 % vs. 5 %; Pearson χ2 = 4.38; p = 0.036). These findings indicate that the weighted survey data are representative of the NYCHA senior population overall.18 This study analyzed surveys with data on senior center use (n = 1000).
Study Variables
The main outcomes of interest were any use of senior centers and frequent attendance at senior centers. Residents were asked: “In the past three months, how often did you use a senior center?” Response categories included “Every day”, “A few days a week”, “Once a week”, “Once a month or less”, and “Never”. Residents who indicated any senior center use in the past three months were considered senior center users. Among residents who reported using a senior center in the past three months, residents who indicated that they used centers “a few days a week” or “every day” were considered frequent attenders.
Demographic variables included in this analysis, including gender, age, race (recoded as Hispanic, Non-Hispanic Black, Non-Hispanic White, and Non-Hispanic Asian), development type, development name, household size, and household income, were obtained from TDS. Type of housing development was categorized as senior-only, mixed or Naturally Occurring Retirement Community (NORC), and family (non-NORC). NYCHA older adults can choose to live in senior-only housing, where they have case management services and a system of floor captains to prevent isolation. Senior-only developments consist entirely of senior-only housing, while mixed developments have both senior-only and family buildings. NORC programs in NYC are public-private partnerships that provide funding for a range of health and outreach services in communities with high proportions of older residents.21 Mixed developments and NORCs were considered one category for analysis because they both have an elevated proportion of residents ages 62 and older (25-50 %) compared to family developments (<25 %). NYCHA tenure, or length of time residing in NYCHA housing, was categorized as greater than 15 years versus 15 years or less. Place of birth was coded as Puerto Rico, Other US, Other Latin America, and Other International.
Additional data relating to health and demographics, including self-reported health status, Activity of Daily Living (ADL) limitations, current risk for depression, available assistance, and food insecurity were obtained from the NYCHA senior survey. ADL limitations assessed in the survey included difficulty bathing or showering, dressing, eating, getting in and out of bed or chairs, using or getting to the toilet, and getting around inside the home. Risk for depression was assessed using the Patient Health Questionnaire (PHQ-2) scale, a validated scale for depression screening, consisting of the two questions “During the past 2 weeks, how often have you been bothered by little interest or pleasure in doing things?” and “During the past 2 weeks, how often have you been bothered by feeling down, depressed or hopeless?”22 These two items were moderately correlated within this population (Chronbach’s alpha = 0.61). Availability of assistance, one aspect of social support, was determined by asking: “Is there a friend, relative or neighbor who could assist you for a few days if necessary?” Food insecurity was determined by asking three questions about different aspects of potential food insecurity: “In the past 30 days, have you been concerned about having enough food to eat?”, “In the past 30 days, did you ever eat less than you felt you should because there wasn’t enough money to buy food?”, and “In the past 30 days, were you ever hungry but didn’t eat because you weren’t able to get out to buy food?” Respondents who answered yes to any of these three questions were considered food insecure.
In NYC, older adult communities often coalesce among people who are relatively similar to each other both racially and linguistically.23 For this reason, rather than examining race and language as separate variables, a “race/language” variable was created by combining data on race/ethnicity from TDS and survey data on language of survey administration and language spoken at home. The race/language variable divided the Hispanic group into two categories: those interviewed in English and those interviewed in Spanish. The non-Hispanic White group was also divided into two categories: those interviewed in English and those interviewed in Russian. The numbers of non-Hispanic Blacks who were not interviewed in English and did not report speaking English at home (n = 2 interviewed in Spanish) and Asians who were not interviewed in Chinese and did not report speaking Chinese at home (n = 5 interviewed in English) were small, and thus these individuals were excluded from all race/language analysis.
Statistical Procedures
SAS Version 9.2 was used for statistical analysis.24 Multivariable modeling used relative risk regression with generalized estimating equations and robust standard error estimates to account for clustering of respondents within housing developments (Lumley et al. 2006; Orelien 2001).
Forward selection approaches were used to identify significant predictors of (1) any senior center use in the past 3 months and (2) frequent senior center attendance among senior center users. Variables significantly associated with senior center use at alpha(α) = 0.1 by the Rao-Scott χ2 Test were added to the model, and gender was tested as a potential predictor because of consistent findings in the literature that senior center use is more common among women. Variables were added sequentially based on magnitude of association and retained if significant at α = 0.05. Due to minimal missing data (95 %-100 % nonmissing for all variables other than senior center use) we chose pairwise deletion of missing values for all bivariable analysis. The final sample size for any senior center use was n = 970 and for frequent senior center use was n = 313.
After this variable selection process was completed, hypothesized interactions were tested for the model of any senior center use in the past three months. We tested potential interactions between age and ADL limitations and between age and health status. Interactions were retained in the final models if one or more of the interaction terms was significant at α = 0.05 after multiple adjustments. After this process, we selected a model from the two different models (which both had significant interaction terms) based on the QIC goodness-of-fit statistic.
Because of our interest in the relationships between race/language and senior center use and frequency of attendance, we augmented the base models to include any variables found to confound that association. We retained potential confounders (with significant associations at α = 0.1 by Rao-Scott χ2 Test with both senior center outcomes and race/language) in the final model if inclusion changed the beta coefficient for at least one of the race/language categories more than 10 % for ever-use of centers or 15 % for frequent use of centers. Income level was tested as a potential confounder in the frequent senior center attendance model because it is consistently associated with prevalence and frequency of senior center use in the literature.7–9,25
We used 10-fold cross validation to validate our final models. The dataset was randomly partitioned into 10 sections. Each section was used once as a test set against the other 9 sections as a training set. For the cases in each test set, we generated the predicted probability of senior center use by fitting our model to the other 90 % of the data (training set). We then calculated the AUC (c-statistic) for the predicted probabilities that were generated by this cross validation.26,27
Results
Study Population
Older adult residents living in NYCHA in 2009 were predominantly female (71 %) (Table 1). Almost half (49 %) had a household income below the federal poverty level, and about half (54 %) lived alone. Three-quarters of NYCHA older adults lived in family housing (74 %), 14 % lived in senior-only developments, and an additional 12 % lived in mixed developments or NORCs. Most (74 %) had resided in NYCHA for longer than 15 years (median NYCHA tenure, 32 years). The largest race/language groups were English-speaking non-Hispanic Black residents (42 %), Hispanic residents interviewed in English (22 %), and Hispanic residents interviewed in Spanish (21 %).
TABLE 1.
Characteristics of residents ages 65 and older living in New York City Housing Authority developments, June 2009 (n = 1000)a
| Frequency | Percentb | 95 % Confidence Limits | |||
|---|---|---|---|---|---|
| Gender | |||||
| Male | 244 | 29.1 | 25.8 | - | 32.3 |
| Female | 756 | 70.9 | 67.7 | - | 74.2 |
| Age | |||||
| 75 and older | 439 | 44.0 | 40.7 | - | 47.4 |
| 65-74 | 561 | 56.0 | 52.6 | - | 59.3 |
| Race/Language | |||||
| Hispanic, Spanish | 140 | 21.2 | 18.1 | - | 24.3 |
| Hispanic, English | 148 | 22.4 | 19.3 | - | 25.6 |
| Black, English | 568 | 41.8 | 38.6 | - | 45.0 |
| White, English | 58 | 5.9 | 4.3 | - | 7.4 |
| White, Russian | 38 | 3.5 | 2.4 | - | 4.7 |
| Asian, Chinese | 37 | 5.1 | 3.4 | - | 6.7 |
| Housing Type | |||||
| Family | 749 | 73.9 | 70.9 | - | 76.9 |
| Mixed or NORC | 114 | 12.3 | 10.0 | - | 14.6 |
| Senior Only | 137 | 13.8 | 11.5 | - | 16.2 |
| Household Size | |||||
| Lives with Others | 459 | 46.4 | 43.0 | - | 49.8 |
| Lives Alone | 541 | 53.6 | 50.2 | - | 57.0 |
| Household Income | |||||
| <100 % FPL | 464 | 49.0 | 45.6 | - | 52.4 |
| 100 % to < 200 % FPL | 321 | 30.4 | 27.3 | - | 33.5 |
| 200 % + FPL | 214 | 20.6 | 17.9 | - | 23.3 |
| Health Status | |||||
| Excellent/Very Good/Good | 396 | 38.8 | 35.5 | - | 42.1 |
| Fair/Poor | 597 | 61.2 | 57.9 | - | 64.5 |
| Number of ADL (Activity of Daily Living) Limitations Identified | |||||
| At least 1 ADL identified | 275 | 28.5 | 25.4 | - | 31.6 |
| No ADLs identified | 715 | 71.5 | 68.4 | - | 74.6 |
| Place of Birth | |||||
| Puerto Rico | 191 | 30.0 | 26.6 | - | 33.5 |
| Other US | 568 | 46.2 | 42.9 | - | 49.6 |
| Other Latin America | 69 | 10.0 | 7.7 | - | 12.3 |
| Other International | 132 | 13.8 | 11.4 | - | 16.1 |
| Length of Time in NYCHA Housing | |||||
| Fifteen years or less | 253 | 26.3 | 23.3 | - | 29.3 |
| More than 15 years | 736 | 73.7 | 70.7 | - | 76.7 |
a Data were missing from one or more respondents for the following variables: race/language (n = 11), household income (n = 1), health status (n = 7), ADL limitations (n = 10), current risk for depression (n = 10), assistance available (n = 11), length of time in NYCHA housing (n = 11).
b weighted survey data
Any Senior Center Use
Overall, 31 % of NYCHA older adults reported any senior center use in the past three months. In bivariable analysis, a number of factors were associated with senior center use, including living in senior-only developments or in mixed developments/NORCs compared to family developments, having a lower household income, living alone, and having a longer tenure living in NYCHA (Table 2). Race/language was strongly associated with use of senior centers, with the highest prevalence of use among Chinese-speaking Asians (54 %), and the lowest among Hispanics interviewed in Spanish (19 %). Men and women had similar prevalences of senior center use.
TABLE 2.
Senior Center Use among NYCHA older residents, by socio-demographic and health characteristics (n = 1000) a
| Percentb | 95 % Confidence Limits | P-value (GEE) | Prevalence Ratio | 95 % Confidence Limits | |||||
|---|---|---|---|---|---|---|---|---|---|
| Overall | |||||||||
| 31.3 | 28.2 | - | 34.5 | ||||||
| Gender | |||||||||
| Male | 34.5 | 28.2 | - | 40.9 | 0.231 | 1.15 | 0.92 | - | 1.44 |
| Female | 30.0 | 26.5 | - | 33.6 | --- | ref | --- | --- | |
| Age | |||||||||
| 75 and older | 35.0 | 30.1 | - | 39.9 | 0.020 | 1.22 | 1.03 | 1.45 | |
| 65-74 | 28.5 | 24.5 | - | 32.5 | --- | ref | --- | --- | |
| Race/Language | |||||||||
| Black, English | 34.2 | 30.1 | - | 38.3 | 0.001 | 1.79 | 1.25 | - | 2.55 |
| White, English | 31.9 | 19.3 | - | 44.6 | 0.071 | 1.67 | 0.96 | - | 2.90 |
| White, Russian | 27.1 | 11.8 | - | 42.3 | 0.311 | 1.41 | 0.72 | - | 2.77 |
| Asian, Chinese | 54.2 | 37.5 | - | 70.8 | <.0001 | 2.83 | 1.79 | - | 4.47 |
| Hispanic, English | 32.1 | 24.4 | - | 39.8 | 0.019 | 1.67 | 1.09 | - | 2.58 |
| Hispanic, Spanish | 19.2 | 12.3 | - | 26.0 | --- | ref | --- | --- | |
| Development Type | |||||||||
| Senior-only | 47.2 | 38.1 | - | 56.3 | <.0001 | 1.76 | 1.41 | - | 2.20 |
| Mixed or NORC | 40.7 | 30.9 | - | 50.5 | 0.001 | 1.52 | 1.18 | - | 1.96 |
| Family, non-NORC | 26.8 | 23.4 | - | 30.3 | --- | ref | --- | --- | |
| Self-Reported Health | |||||||||
| Excellent/Very Good/Good | 33.5 | 28.4 | - | 38.7 | 0.268 | 1.13 | 0.91 | - | 1.40 |
| Fair/Poor | 29.7 | 25.8 | - | 33.6 | --- | ref | --- | --- | |
| ADL limitations | |||||||||
| At least One ADL limitation | 32.8 | 26.7 | - | 38.9 | 0.520 | 1.07 | 0.87 | - | 1.31 |
| No ADL limitations identified | 30.7 | 27.0 | - | 34.4 | --- | ref | --- | --- | |
| Current Risk for Depression | |||||||||
| Yes | 37.5 | 29.1 | - | 45.9 | 0.106 | 1.24 | 0.95 | - | 1.61 |
| No | 30.2 | 26.8 | - | 33.6 | --- | ref | --- | --- | |
| Live Alone | |||||||||
| Yes | 36.3 | 31.9 | - | 40.7 | 0.002 | 1.42 | 1.14 | - | 1.77 |
| No | 25.6 | 21.3 | - | 29.9 | --- | ref | --- | --- | |
| Assistance Available | |||||||||
| No | 35.8 | 29.7 | - | 42.0 | 0.059 | 1.22 | 0.99 | - | 1.50 |
| Yes | 29.3 | 25.7 | - | 33.0 | --- | ref | --- | --- | |
| Food Insecurity | |||||||||
| Yes | 34.1 | 26.7 | - | 41.4 | 0.424 | 1.11 | 0.86 | - | 1.44 |
| No | 30.7 | 27.2 | - | 34.1 | --- | ref | --- | --- | |
| Household Income Level | |||||||||
| < 100 % FPL | 34.7 | 30.0 | - | 39.3 | 0.011 | 1.53 | 1.10 | - | 2.11 |
| 100 % to < 200 % FPL | 31.9 | 26.4 | - | 37.5 | 0.048 | 1.41 | 1.00 | - | 1.97 |
| 200 % + FPL | 22.7 | 16.6 | - | 28.8 | --- | ref | --- | --- | |
| Income Below Poverty Line | |||||||||
| Yes | 34.7 | 30.0 | - | 39.3 | 0.049 | 1.23 | 1.00 | - | 1.51 |
| No | 28.2 | 24.1 | - | 32.4 | --- | ref | --- | --- | |
| Length of Time in NYCHA Housing | |||||||||
| Fifteen Years or Less | 37.9 | 31.3 | - | 44.4 | 0.021 | 1.29 | 1.04 | - | 1.60 |
| More than 15 Years | 29.4 | 25.8 | - | 33.0 | --- | ref | --- | --- | |
a Data were missing from one or more respondents for the following variables: race/language (n = 11), household income (n = 1), health status (n = 7), ADL limitations (n = 10), current risk for depression (n = 10), assistance available (n = 11), length of time in NYCHA housing (n = 11).
b weighted survey data
Frequent Senior Center Attendance
Among those reporting senior center use, 63 % reported visiting the centers frequently (more than once a week). Frequent attendance was more common among men than among women, and its prevalence also varied significantly by race/language. The race/language group with the lowest prevalence of frequent senior center attendance among users was Chinese-speaking Asians (38 %), compared to 58 % for English-speaking non-Hispanic Blacks and 67-75 % for other race/language groups (Table 3).
TABLE 3.
Prevalence of Frequent Senior Center Attendance (more than once a week) among users (n = 319) a
| Percent b | 95 % Confidence Limits | P-value | Prevalence Ratio | 95 % Confidence Limits | |||||
|---|---|---|---|---|---|---|---|---|---|
| Overall | |||||||||
| 63.0 | 57.3 | - | 68.8 | ---- | ---- | ---- | ---- | ||
| Gender | |||||||||
| Male | 53.7 | 42.5 | - | 64.9 | 0.037 | 0.80 | 0.64 | - | 0.99 |
| Female | 67.4 | 60.9 | - | 73.9 | ---- | ref | ---- | ---- | |
| Age | |||||||||
| 75 and older | 65.4 | 57.3 | 73.5 | 0.413 | 0.93 | 0.78 | - | 1.11 | |
| 65-74 | 60.8 | 52.6 | - | 68.9 | ---- | ref | ---- | ---- | |
| Race/Language | |||||||||
| Black, English | 57.7 | 50.3 | - | 65.0 | 0.045 | 0.76 | 0.59 | - | 0.99 |
| White, English | 67.0 | 44.3 | - | 89.6 | 0.572 | 0.89 | 0.59 | - | 1.34 |
| White, Russian | 68.5 | 37.8 | - | 99.2 | 0.664 | 0.91 | 0.59 | - | 1.40 |
| Asian, Chinese | 38.3 | 17.1 | - | 59.5 | 0.028 | 0.51 | 0.28 | - | 0.93 |
| Hispanic, English | 73.4 | 60.5 | - | 86.2 | 0.853 | 0.97 | 0.72 | - | 1.30 |
| Hispanic, Spanish | 75.4 | 57.9 | - | 93.0 | ---- | ref | ---- | ---- | |
| Development Type | |||||||||
| Senior-only | 68.6 | 56.9 | - | 80.4 | 0.623 | 1.05 | 0.86 | - | 1.28 |
| Mixed or NORC | 46.8 | 30.9 | - | 62.7 | 0.092 | 0.72 | 0.49 | - | 1.06 |
| Family, non-NORC | 65.3 | 58.2 | - | 72.4 | ---- | ref | ---- | ---- | |
| Self-Reported Health | |||||||||
| Fair/Poor | 59.2 | 51.4 | - | 66.9 | 0.149 | 0.86 | 0.70 | - | 1.06 |
| Excellent/Very Good/Good | 68.9 | 60.3 | - | 77.5 | ---- | ref | ---- | ---- | |
| ADL limitations | |||||||||
| At least One ADL limitation | 60.3 | 49.3 | - | 71.2 | 0.526 | 0.94 | 0.78 | - | 1.14 |
| No ADLs identified | 64.1 | 57.3 | - | 70.9 | ---- | ref | ---- | ---- | |
| Current Risk for Depression | |||||||||
| Yes | 66.2 | 53.3 | - | 79.2 | 0.603 | 1.06 | 0.85 | - | 1.32 |
| No | 62.5 | 56.0 | - | 68.9 | ---- | ref | ---- | ---- | |
| Live Alone | |||||||||
| Yes | 67.3 | 60.2 | - | 74.4 | 0.073 | 1.20 | 0.98 | - | 1.47 |
| No | 56.0 | 46.5 | - | 65.6 | ---- | ref | ---- | ---- | |
| Assistance Available | |||||||||
| Yes | 64.5 | 57.5 | - | 71.5 | 0.497 | 1.07 | 0.88 | - | 1.29 |
| No | 60.4 | 50.0 | - | 70.8 | ---- | ref | ---- | ---- | |
| Food Insecurity | |||||||||
| Yes | 69.3 | 57.7 | - | 81.0 | 0.181 | 1.13 | 0.94 | - | 1.35 |
| No | 61.3 | 54.8 | - | 67.9 | ---- | ref | ---- | ---- | |
| Household Income Level | |||||||||
| <100 % FPL | 64.5 | 56.7 | - | 72.3 | 0.450 | 1.12 | 0.83 | - | 1.52 |
| 100 % to < 200 % FPL | 63.2 | 53.0 | - | 73.4 | 0.577 | 1.10 | 0.79 | - | 1.54 |
| 200 % + FPL | 57.4 | 42.2 | - | 72.7 | ---- | ref | ---- | ---- | |
| Length of Time in NYCHA Housing | |||||||||
| Fifteen Years of Less | 60.2 | 49.5 | - | 70.9 | 0.486 | 0.94 | 0.78 | - | 1.12 |
| More than 15 Years | 64.2 | 57.4 | - | 71.0 | --- | ref | --- | --- | |
a Data were missing from one or more respondents for the following variables: race/language (n = 6), health status (n = 4), ADL limitations (n = 4), current risk for depression (n = 4), assistance available (n = 9), length of time in NYCHA housing (n = 1).
b weighted survey data
Regression Results: Any Senior Center Use
Significant predictors of any senior center use in the past three months included race/language, household size, development type, age group, ADL limitations, an interaction between ADL limitations and age group, and current risk of depression (Table 4). Race/language was the strongest predictor of senior center use. Chinese-speaking Asians (prevalence ratio [PR] = 2.44, 95 % confidence interval [CI]: 1.47-4.05), English-speaking non-Hispanic Blacks (PR = 1.72, 95 % CI: 1.20-2.49), and Hispanics interviewed in English (PR = 1.69, 95 % CI: 1.11-2.59) were all significantly more likely to use senior centers than Hispanics interviewed in Spanish. Living in senior-only or mixed/NORC housing type and living alone remained significant positive predictors of senior center use after multiple adjustments, as did being at current risk for depression. There was a significant interaction (p = 0.017) between age group and ADL limitations. Among those without any ADL limitations, seniors aged 75 and older had a higher prevalence of senior center use than those aged 65–74, but no age differences were observed among those with at least one ADL limitation. Neither age nor ADL limitations were significantly related to senior center use in the final model without including the interaction term. Gender was not significantly associated with senior center use when added to the final model. The QIC goodness of fit statistic for this model was 1159.95 and the c-statistic (AUC) from tenfold cross-validation was 0.627.
TABLE 4.
Predictors of senior center use and frequent senior center attendance—best-fit models
| p-value b | Prevalence | 95 % Confidence Limits | |||
|---|---|---|---|---|---|
| Ratio | |||||
| Senior Center Use among All Seniors (n = 970) | |||||
| Race/Language Category | |||||
| Asian, Chinese | 0.001 | 2.44 | 1.47 | - | 4.05 |
| Black, English | 0.004 | 1.72 | 1.20 | - | 2.49 |
| Hispanic, English | 0.015 | 1.69 | 1.11 | - | 2.59 |
| White, English | 0.264 | 1.38 | 0.79 | - | 2.40 |
| White, Russian | 0.751 | 1.11 | 0.58 | - | 2.11 |
| Hispanic, Spanish | --- | ref | --- | --- | |
| Household Size | |||||
| Lives Alone | 0.034 | 1.26 | 1.02 | - | 1.56 |
| Lives with Others | --- | ref | --- | --- | |
| Development Type | |||||
| Senior-Only | <0.001 | 1.56 | 1.24 | - | 1.95 |
| Mixed or NORC | 0.009 | 1.44 | 1.10 | - | 1.89 |
| Family, non-NORC | --- | ref | --- | --- | |
| Age and ADL Limitations (reference is age 65–74, no ADL limitations) | |||||
| ADL Limitations | |||||
| Aged 75 and older a | 0.017 | 1.01 | 0.80 | - | 1.45 |
| Aged 65-74 | 0.207 | 1.23 | 0.89 | - | 1.69 |
| No ADL Limitations | |||||
| Aged 75 and older | 0.001 | 1.45 | 1.18 | - | 1.80 |
| Aged 65-74 | --- | ref | --- | --- | |
| Current Risk of Depression | |||||
| At Risk | 0.024 | 1.29 | 1.04 | - | 1.61 |
| Not At Risk | --- | ref | --- | --- | |
| Frequent Senior Center Use among Seniors using Senior Centers (n = 313) | |||||
| Race/Language Category | |||||
| Black, English | 0.110 | 1.60 | 0.90 | - | 2.85 |
| Hispanic, English | 0.020 | 2.02 | 1.12 | - | 3.66 |
| Hispanic, Spanish | 0.003 | 2.41 | 1.36 | - | 4.25 |
| White, English | 0.071 | 1.84 | 0.95 | - | 3.58 |
| White, Russian | 0.053 | 1.92 | 0.99 | - | 3.73 |
| Asian, Chinese | --- | ref | --- | --- | |
| Race/Language Category (alternate reference) | |||||
| Black, English | <.0001 | 0.67 | 0.56 | - | 0.79 |
| Hispanic, English | 0.014 | 0.84 | 0.73 | - | 0.97 |
| Hispanic, Spanish | --- | ref | --- | --- | |
| Gender | |||||
| Female | 0.012 | 1.33 | 1.064 | - | 1.660 |
| Male | --- | ref | --- | --- | |
| Income Level | |||||
| <100 % Poverty Level | 0.182 | 1.29 | 0.89 | - | 1.87 |
| 100 %-200 % Poverty Level | 0.498 | 1.15 | 0.76 | - | 1.74 |
| >200 % Poverty Level | --- | ref | --- | --- | |
a p-value for interaction term
b weighted survey data
Regression Results: Frequent senior center attendance
A second model was constructed to predict frequent senior center attendance among those who reported using a senior center in the past 3 months. This model included race/language, gender, and income level (Table 4) as significant predictors of frequent attendance. The prevalence of frequent senior center attendance was highest among Hispanics interviewed in Spanish (the group with the lowest overall rate of any senior center use) and was lowest among Chinese-speaking Asians (PR = 2.41; 95 % CI: 1.36-4.25 comparing these groups). Hispanics interviewed in Spanish also had a significantly higher prevalence of frequent attendance than both non-Hispanic Blacks and Hispanics interviewed in English. Women had a higher prevalence of frequent attendance than men (PR = 1.33; 95 % CI: 1.06-1.66). Income was not significantly related to frequent attendance, but adjusting for income increased the apparent association between race/language and senior center use for all groups. The QIC goodness of fit statistic for this model was 400.00 and the c-statistic (AUC) was 0.561.
Discussion
We found that approximately one in three older public housing residents in NYC used senior centers, and one in five used them frequently. This is a much higher level of use than found in a 1984 national survey, where 14 % of older adults overall reported using a senior center in the past year.7 Our analysis showed that older NYCHA residents who lived alone and those at risk for depression as indicated by the PHQ-2 screen were more likely to be senior center users, indicating some success at reaching potentially isolated public housing residents. Unlike prior studies, we also found that men were as likely to use centers as women in this heavily female population, although they tended to visit less frequently. While some researchers and senior center stakeholders have noted lower senior center attendance among “younger” seniors (those aged 65–74),5,17,28 we found this to be the case only among those with no ADL limitations.
In this population, English-speaking non-Hispanic Blacks and Whites had similar prevalences of any senior center use, while Spanish-speaking Hispanics had a lower prevalence of use (differences that persisted even after controlling for potential confounders). Among Hispanics, we found that taking the survey in English strongly predicted use of senior centers even after adjusting for birthplace and length of time in NYCHA housing (data not shown). However, among those attending senior centers, frequent attendance was more common among Hispanics interviewed in Spanish than among other race/ethnic groups. Prior studies on race/ethnicity and senior center use have had inconsistent results.7,8,11,29 This may be because the communities studied differ substantially from each other and from the urban public housing population we surveyed, or because the cultures and orientations of senior centers can vary between different locations. Our findings suggest that the factors driving first use or occasional use of a service may be different than those determining frequent or regular use. In NYC, language proficiency might be a significant barrier to first using senior centers for Hispanics, despite efforts to provide English/Spanish bilingual materials and staff. Spanish-speaking Hispanics may have substantial unmet need for senior center services, since our data show that once they become senior center users they are likely to attend frequently. Race/language groups may also differ in their levels of interest in senior center activities offered, opportunity for socialization outside of senior centers, and knowledge about senior centers.
Senior center use was more common in our study among adults living in senior-only or mixed/NORC public housing developments compared to those in family developments. Residents of senior-only developments and buildings have chosen to live with other older adults, presumably in order to benefit from social opportunities and/or services. Case-management and crisis intervention services are provided to residents, and a system of floor captains ensures that residents are in contact with one another and are not isolated. Available services, self-selection, and/or a sense of community fostered by a high proportion of seniors may be responsible for the elevated prevalence of senior center use at NORCs and senior-only/mixed developments. Residents of developments with higher proportions of seniors may also have more opportunity to learn about the centers through word-of-mouth.
In multivariable analysis, we found that older adults living alone had a higher prevalence of senior center use, consistent with the findings of Krout et al. (1990).7 Seniors living alone may have greater need for senior center services because they might otherwise be at risk of social isolation. We also found a higher prevalence of senior center use among older adults at current risk of depression. In contrast, in a random sample of Missouri seniors, Calsyn and Winter (1999) found that senior center users had better mental health than seniors not using centers.10
By definition, NYCHA residents have low/moderate-incomes. In this population, we did not find a higher prevalence of senior center use among residents in the lowest-income category, as consistently seen in the literature.7,8,12,25,30 However, the NYCHA population has a lower income overall than most other groups that have been studied. We did find that, among senior center users, older adults with incomes below the federal poverty line were more likely to visit centers frequently, potentially indicating greater need for services. Public housing residents with very low incomes may have greater need for senior center services because they lack alternate low-cost activities12 or have a greater need for the specific resources available there,9 such as free meals and assistance with social services and benefits. In a study of senior center attendees in Arizona, Fitzpatrick et al. found that low-income seniors had greater self-perceived benefits from the centers than seniors with higher incomes.13
The strengths of this study lie in the representativeness of the survey and linkage with administrative data to examine the experience of a well-delineated, at-risk population – older adults living in urban public housing – and identify opportunities for improved targeting of senior center services. Results from the cross-validation analysis indicated that our models had moderate accuracy to predict senior center use, suggesting that unmeasured variables also influence the likelihood of senior center attendance. Additional limitations include potential sources of error for several of the variables. Our survey questions were not independently validated in the different languages the survey was administered in, which limits the known validity of the language findings. NYCHA’s TDS only reports on the “authorized” residents that are listed on the apartment lease, which may lead to an under-reporting of the number of people actually living in the household. This under-reporting of household size may have led us to underestimate the effect of living alone on senior center use. All data on health and mental health was self-reported, which may have increased measurement error or led to biases. A limitation to our finding on mental health is that the PHQ-2 screen contains only two questions and may have a limited ability to identify seniors with depression.
We found that older public housing residents in New York City have relatively high rates of senior center use across lines of gender, health status, and age. This is evidence that senior centers may be an appropriate setting for health promotion activities targeting low-income older adults. We also found a lower prevalence of use among Hispanics who were interviewed in Spanish, indicating that language preference can be an important determinant of senior center use. Additional outreach may be needed in this population in New York City. This finding underscores the importance of examining the impacts of both race and language on service utilization in communities with significant immigrant populations. Older adults living in public housing developments with a high proportion of older residents had a high rate of senior center use. Outreach services are provided in these developments, and may help to inform residents about senior center services. We suggest that extending outreach services to a wider range of seniors and increasing publicity efforts for senior centers might help to increase attendance. Our findings demonstrate the importance of community context and outreach to senior center use and have the potential to inform efforts to reduce barriers to service utilization among at-risk urban seniors.
Acknowledgments
The authors would like to thank the survey participants, Baruch Survey Research, Carolyn Greene, Tiffany Harris, Bonnie Kerker, Elizabeth Needham Waddell, Sungwoo Lim, and Hilary Parton at NYCDOHMH; Jackie Berman, Michael Bosnick, and Karen Taylor of DFTA; Richard Greene, Nalini Viswanathan, and Larry Wilensky of NYCHA; Allan Uribe and Katherine Chen of CUNY, and Andrew Rundle of Columbia University.
References
- 1.The New York Academy of Medicine. NYC Senior Centers: Visioning the Future. 2010.
- 2.Dal Santo TS. Senior Center Literature Review: Reflecting and Responding to Community Needs. California Comission on Aging. Sacramento, CA; 2009.
- 3.Hawranik P, Pangman V. Perceptions of a senior citizens’ Wellness Center. The community’s voice. J Gerontol Nurs. 2002;28(11):38–44. doi: 10.3928/0098-9134-20021101-09. [DOI] [PubMed] [Google Scholar]
- 4.Farone DW, Fitzpatrick TR, Tran TV. Use of senior centers as a moderator of stress related distress among Latino elders. J Gerontol Soc Work. 2005;46(1):65–83. doi: 10.1300/J083v46n01_05. [DOI] [PubMed] [Google Scholar]
- 5.Turner KW. Senior Citizens Centers: What They Offer, Who Participates, and What They Gain. J Gerontol Soc Work. 2004;43(1):37–47. doi: 10.1300/J083v43n01_04. [DOI] [Google Scholar]
- 6.Schneider MJ, Chapman DD, Voth DE. Senior center participation: a two-stage approach to impact evaluation. Gerontologist. 1985;25(2):194–200. doi: 10.1093/geront/25.2.194. [DOI] [PubMed] [Google Scholar]
- 7.Krout JA, Cutler SJ, Coward RT. Correlates of senior center participation: a national analysis. Gerontologist. 1990;30(1):72–79. doi: 10.1093/geront/30.1.72. [DOI] [PubMed] [Google Scholar]
- 8.Netzer JK, Coward RT, Peek CW, Henretta JC, Duncan RP, Dougherty MC. Race and Residence Differences in the Use of Formal Services by Older Adults. Res Aging. 1997;19(3):300–332. doi: 10.1177/0164027597193003. [DOI] [Google Scholar]
- 9.Sabin EP. Frequency of Senior Center Use: A Preliminary Test of Two Models of Senior Center Participation. J Gerontol Soc Work. 1993;20(1/2):97–105. doi: 10.1300/J083V20N01_07. [DOI] [Google Scholar]
- 10.Calsyn R, Winter J. Who Attends Senior Centers? J Soc Serv Res. 1999;26(2):53–69. doi: 10.1300/J079v26n02_03. [DOI] [Google Scholar]
- 11.Pardasani M. Senior Centers: Characteristics of Participants and Nonparticipants. Act Adapt Aging. 2010;34:48–70. [Google Scholar]
- 12.Krout JA. Correlates of Senior Center Utilization. Res Aging. 1983;5(3):339–352. doi: 10.1177/0164027583005003004. [DOI] [Google Scholar]
- 13.Fitzpatrick T, McCabe J, Gitelson R, Andereck K. Factors That Influence Perceived Social and Health Benefits of Attendance at Senior Centers. Act Adapt Aging. 2006;30(1):23–45. [Google Scholar]
- 14.Ralston PA. Senior center utilization by black elderly adults: social, attitudinal and knowledge correlates. J Gerontol. 1984;39(2):224–229. doi: 10.1093/geronj/39.2.224. [DOI] [PubMed] [Google Scholar]
- 15.Rosso AL, Taylor JA, Tabb LP, Michael YL. Mobility, disability, and social engagement in older adults. J Aging and Health. 2013;25(4):617–637. doi: 10.1177/0898264313482489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ralston PA. Determinants of Senior Center Attendance and Participation. J Appl Gerontol. 1991;10(3):258–273. doi: 10.1177/073346489101000303. [DOI] [Google Scholar]
- 17.Pardasani M. Senior Centers: Focal Points of Community-Based Services for the Elderly. Act Adapt Aging. 2004;28(4):27–44. [Google Scholar]
- 18.Parton HB, Greene R, Flatley A-M, et al. Health of Older Adults in New York City Public Housing: Findings from the New York City Housing Authority Senior Survey. A joint report by the New York City Housing Authority, the New York City Departments of Health and Mental Hygiene and for the Aging, and the City University of New York School of Public Health at Hunter College. New York, NY; 2011. http://www.nyc.gov/html/doh/downloads/pdf/episrv/nycha_senior_summary.pdf.
- 19.Giunta N, Morano C, Parikh NS, Friedman D, Fahs MC, Gallo WT. Racial and ethnic diversity in senior centers: comparing participant characteristics in more and less multicultural settings. J Gerontol Soc Work. 2012;55(6):467–483. doi: 10.1080/01634372.2011.653519. [DOI] [PubMed] [Google Scholar]
- 20.Parton HB, Greene R, Flatley AM, et al. Health of older adults in New York City public housing: part 1, findings from the New York City Housing Authority Senior Survey. Care Manag J: J Case Manag ; J Long Term Home Health Care. 2012;13(3):134–147. doi: 10.1891/1521-0987.13.3.134. [DOI] [PubMed] [Google Scholar]
- 21.United Hospital Fund. 9 Frequently Asked Questions about NORCs. 2014; http://www.uhfnyc.org/initiatives/aging-in-place/frequently_asked_questions/. Accessed April 29, 2014.
- 22.Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2: validity of a two-item depression screener. Med Care. 2003;41(11):1284–1292. doi: 10.1097/01.MLR.0000093487.78664.3C. [DOI] [PubMed] [Google Scholar]
- 23.Sanjek R. The Future of Us All: Race and Neighborhood Politics in New York City. Ithaca, NY: Cornell University Press; 1998. [Google Scholar]
- 24.Version 9.2 of the SAS System for Windows [computer program]. Cary, NC, USA.
- 25.Calsyn R, Burger GK, Roades LA. Cross-Validation of Differences Between Users and Non-Users of Senior Centers. J Soc Serv Res. 1996;21(3):39–56. doi: 10.1300/J079v21n03_03. [DOI] [Google Scholar]
- 26.Gould MK, Ananth L, Barnett PG. A clinical model to estimate the pretest probability of lung cancer in patients with solitary pulmonary nodules. Chest. 2007;131(2):383–388. doi: 10.1378/chest.06-1261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Subramanian J, Simon R. An evaluation of resampling methods for assessment of survival risk prediction in high-dimensional settings. Stat Med. 2011;30(6):642–653. doi: 10.1002/sim.4106. [DOI] [PubMed] [Google Scholar]
- 28.Lund A, Engelsrud G. ‘I am not that old’: inter-personal experiences of thriving and threats at a senior centre. Ageing Soc. 2008;28:675–692. doi: 10.1017/S0144686X07006988. [DOI] [Google Scholar]
- 29.Laditka SB, Laditka JN, Drake BF. Home- and Community- Based Service Use by Older African American, Hispanic, and Non-Hispanic White Women and Men. Home Health Care Serv Q. 2006;25(3/4):129–153. doi: 10.1300/J027v25n03_08. [DOI] [PubMed] [Google Scholar]
- 30.Friedman D, Parikh NS, Ruiz MA, Fahs MC. Brookdale Demonstration Initiative in Healthy Urban Aging: The NYC Senior Center Survey Final Report. New York, NY; 2010. http://www.brookdale.org/brookdale-assets/research/DFTA-projects/DFTA-NYC-senior-center-survey-final-report-years-1and2.pdf.