Abstract
Introduction
Latinos report lower self-rated health (SRH) than non-Hispanic White persons. However, the association between SRH and medically diagnosed chronic diseases (MDCDs) remains understudied in Latino populations. This study assessed the relationship between a single-item SRH indicator and MDCD status among predominantly Latino adults in Puerto Rico.
Methods
Participants (30–75 years; n=965) of the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends (PROSPECT) reported SRH (excellent/very good, good, or fair/poor) and MDCD (ever vs never). We performed multivariate logistic regressions to evaluate the association between SRH and MDCD, which adjusted for key socioeconomic, demographic, and behavioral confounders.
Results
Twenty-seven percent of participants reported excellent/very good SRH, 39% good, and 34% fair/poor. Participants with fair/poor SRH (vs excellent/very good) were more likely to report MDCD for painful inflammation (odds ratio [OR]=4.95 [95% CI, 3.27–7.48]), kidney disease (4.64 [2.16–9.97]), sleep disorder (4.47 [2.83–7.05]), migraine headaches (4.07 [2.52–6.58]), overweight/obesity (3.84 [2.51–5.88]), depression (3.61 [2.28–5.74]), hypertension (3.59 [2.43–5.32]), high blood sugar (3.43 [2.00–5.89]), cardiovascular disease (3.13 [2.01–4.87]), anxiety (2.87 [1.85–4.44]), arthritis (2.80 [1.83–4.30]), diabetes (2.46 [1.57–3.83]), respiratory problems (2.45 [1.59–3.79]), stomach problems (2.44 [1.57–3.81]), eye disease (2.42 [1.44–4.06]), gallbladder disease (2.34 [1.35–4.05]), liver disease (2.26 [1.38–3.70]), heartburn (2.25 [1.55–3.26]), hyperlipidemia (2.10 [1.44–3.06]), and thyroid conditions (2.04 [1.30–3.21]).
Conclusions
SRH may reflect MDCD burden and serve as a valid screener to efficiently identify Latino individuals in high need of clinical services. This is relevant in Puerto Rico, where chronic disease rates remain high amid limited, disparate access to health care.
Keywords: Self-Rated Health, Medical Diagnosed Chronic Diseases, Underrepresented Population, Puerto Rico, Adults, Morbidity
Introduction
For nearly 70 years, self-rated health (SRH) measures have served a critical role in the rapid assessment of subjective health status.1 SRH is often captured in a single-item question asking: “In general, how would you describe your health?” Respondents then rate “excellent” to “poor” on a 5-point scale, reflecting their perception of their current health status. Published research in the 1970s demonstrated a consistent longitudinal association between SRH and mortality.2 Thereafter, SRH items were deemed valid as a proxy for current health status and incorporated globally across a wide range of surveys for surveillance, risk screening, and clinical trials.1
As a global measure, SRH prompts participants to quickly gather, synthesize, and report on information they deem relevant to their definition of health.2 These definitions vary and not only encompass disease burden but also integrate socioecologic, physical, and psychological domains.3 Additionally, respondents invoke a diverse range of factors such as personal health promotion behaviors,4 healthy neighborhood context,5 and health care access and utilization6 to complete the SRH framework. In this way, SRH represents the intersection of biology and culture,1 as reflected by the differential impact of social identities on the relationship between SRH and health outcomes. These social identities include, but are not limited to, sex and gender identity,7 spoken language,8 and race and ethnicity.9 Therefore, it is valuable for public health and clinical purposes to examine the SRH validity within unique sociocultural contexts and diverse samples.
Though research on its use in Spanish-speaking populations is emerging,10 Hispanic and Latino people (hereafter referred to as Latinos to refer to individuals of all genders and Hispanic or Latino ethnic heritages) remain underrepresented in the SRH literature, and studies accounting for the racial and ethnic heterogeneity of Latino populations are near absent. By ignoring the diversity among Latinos—who make up the largest ethnic minority group in the United States—we risk missing critical differences in how health is defined, how SRH is interpreted, and how this interpretation relates to objective health outcomes.10
Responding to this gap, this study is among the first to examine SRH among adults in the US territory of Puerto Rico, where 99% self-identify as Latino, and the first, to our knowledge, to examine the relationship between SRH and self-reported medically diagnosed chronic disease (MDCD) status. Puerto Rico serves as an important context for this research, as its residents have relatively high rates of chronic disease (compared to mainland US amid limitations in clinical care,11 especially after the recent hurricanes).12 Quick and valid screeners to efficiently identify those in highest need of services13 are immediately necessary; a single-item SRH measure may fill this gap if, as we hypothesize, poor objective health status (as defined by relatively high cumulative lifetime MDCD) is associated with higher odds of reporting fair or poor SRH among adults living in Puerto Rico.
Methods
Study Design
Details on the design and implementation of the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends (PROSPECT) have been published.14 Briefly, PROSPECT is a prospective cohort study designed and implemented through the collaborative efforts of Harvard T.H. Chan School of Public Health, Ponce Health Sciences University, and University of Massachusetts, Lowell. Via multistage sampling, PROSPECT has recruited and enrolled adult residents of Puerto Rico, across over 50 partner clinics to assess psychosocial, lifestyle-related, and cardiometabolic domains. Those eligible were adult residents of Puerto Rico (aged 30–75 years at enrollment), not institutionalized, living in Puerto Rico for at least a year and with no plan to move for the next 3 years, living in a stable dwelling, and able to answer questions without assistance in English or Spanish. The study was registered under ClinicalTrials.gov (NCT03794531) in January 2019; enrollment is ongoing.
Ethics
The institutional review boards of Harvard T.H. Chan School of Public Health, Ponce Health Sciences University, and University of Massachusetts, Lowell approved this study. All procedures were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000. Informed consents were obtained from all participants included in the study.
Data Collection
Trained bilingual (English and Spanish) research assistants conducted baseline in-person interviews with participants (who decided on interview language preference) to ask about sociodemographic, medical, psychosocial lifestyle-related, and cardiometabolic factors and measure biologic samples and anthropometric samples. Participants received a $50 incentive gift card after signing the consent form during the baseline visit. At the time of analysis, 1015 participants have been enrolled, with a target sample of 2000 adults. We analyzed only data from the baseline interview in the current study.
Measures
Sample Characteristics
Participants provided data on demographic factors (eg, age, sex at birth, marital status, and ethnicity) and socioeconomic characteristics (eg, household income, employment history, and duration of formal education). We administered validated psychosocial measures such as the Social Connectedness/Social Assurance Scale15 to evaluate participants’ sense of belonging within a group. We collected detailed data on health behaviors, including alcohol and tobacco use (never, past, current).16
Self-Rated Health
During the interview, we asked participants to rate their health with the question “In general, how would you describe your health?” on a 5-point Likert scale (ie, “excellent,” “very good,” “good,” “fair,” and “poor”). Similar to previous SRH studies,17 we combined the categories at either extreme to create a 3-point scale (ie, “excellent or very good,” “good,” and “fair or poor”).
Medical Diagnosed Chronic Diseases and Medication Use
The participants self-reported MDCDs and medication use. We asked participants if a doctor or health professional had ever told them that they had an illness or medical condition from a list read to them; medication use, year of diagnosis, and current diagnosis were probed for any affirmative response. MDCDs (categorized as ever [past, current] vs never diagnosed) analyzed in the current article included those of diabetes, high blood sugar, thyroid condition, overweight/obesity, stomach problems, gallbladder condition, heartburn, hypertension, hyperlipidemia, cardiovascular disease, kidney disease, liver disease, migraine headaches, sleep disorder, arthritis, osteoporosis, painful inflammation, anxiety, depression, respiratory problems, eye disease, and cancer. As a measure of comorbidity,18 an unweighted sum of lifetime cumulative and current medical diagnoses was calculated for each participant.
Statistical Analysis
For the overall sample and stratified by SRH, we tabulated the frequencies of demographic characteristics, socioeconomic factors, and health behaviors. We tabulated the prevalence of ever MDCD listed above; among those who reported fair or poor SRH, we also listed the frequency of ever versus never MDCDs. We report the frequency of past versus current MDCDs by time point (ie, past, current, never) and related medication use. We conducted χ2 tests for all frequencies described above to examine the statistical significance of these proportions’ differences across groups.
As previously done,19 we used logistic regression modeling to estimate the relative odds of reporting fair or poor (vs excellent or very good) SRH per MDCD (using our summative comorbidity scores for ever and current diagnoses) and for each individual MDCD (as a dichotomous score, ever vs never diagnosed). Three models were fit to examine this relationship. Model 1 was unadjusted; model 2 was adjusted for sex (male, female), employment status (retired, employed, unemployed, other), educational attainment (<high school, high school, some college or associate’s degree, college or more), income ($0–$10,000; $10,001–$20,000; >$20,000), health insurance (no insurance, public insurance, private or other), and social connectedness score; and model 3 additionally adjusted for drinking and smoking behavior (current, ever, never). Family history of diabetes, hypertension, hyperlipidemia, cardiovascular disease, or cancer, and medication use for the tested conditions were considered potential confounders but were ultimately excluded, as the models were unstable given the small sample sizes for these variables.
All analyses were conducted in SAS software version 9.4 (Cary, NC) and defined statistical significance as estimates with a P value less than .05.
Missing Data
Four of the 1015 participants available at the time of analysis were excluded for missing data on SRH, 19 for unreliable interview data, and 27 for incomplete interview data. Of the 965 remaining, 88.1% (n=850) provided data for all MDCDs in this study. However, 74 were missing data for 1 condition, 22 for 2 conditions, and 19 for 3 or more; the most common missing information was arthritis (n=11) and obesity (n=4).
Results
Sample Characteristics
Approximately one-third (34.1% fair or poor, 38.8% good, and 27.1% excellent or very good) of the sample participants assigned themselves to each SRH category (Table 1). Of this sample, a majority identified as female (71.3%), who were less likely to rate their health as excellent/very good than males. Approximately one-third of participants were 41 to 55 years old (34.9%), and another third were 56 to 65 years old (35.5%); participants in the youngest and oldest age categories were more likely to rate their health as excellent/very good than those in the middle-aged categories. Nearly all participants (93.5%) identified as Puerto Rican. Approximately half reported being married (49.8%) and a similar percentage was employed (45.4%); employed participants tended to rate their health as excellent/very good. Almost every participant (95.7%) reported having health insurance; of these, more than half reported private insurance (58.2%). Nearly half reported attending at least some college (46.0%), and 40.8% reported an income of >$20,000; these participants reported higher excellent/very good SRH. Among those surveyed, the mean number of MDCDs across the lifetime was 5.5 (SD=3.8); the mean number of current MDCDs was 3.7 (SD=2.9). The average social connectedness score was 40.5 (SD=9.4), with higher scores for those reporting excellent/very good SRH, and social assurance was 35.2 (SD=9.3). Approximately half reported current alcohol consumption (48.8%), which tended to be higher among those reporting fair/poor SRH (16.1% among fair/poor vs 10.7% excellent/very good), and most have never smoked (66.8%).
Table 1.
Characteristics of adults living in Puerto Rico, by self-rated health (n=965, 2019–2020)
| Overalla | Fair or poorb | Goodb | Excellent or very goodb | ||
|---|---|---|---|---|---|
| (n=965) | (n=329, 34.1%) | (n=374, 38.8%) | (n=262, 27.1%) | ||
| Demographics | % | % | % | % | P valuec |
| Sex | .001*** | ||||
| Male | 28.7 | 24.3 | 26.5 | 37.4 | |
| Female | 71.3 | 75.7 | 73.5 | 62.6 | |
| Age (years) | .0001*** | ||||
| 30–40 | 15.5 | 9.7 | 15.5 | 22.9 | |
| 41–55 | 34.9 | 42.0 | 31.8 | 30.5 | |
| 56–65 | 35.5 | 36.2 | 38.5 | 30.5 | |
| 66–75 | 14.0 | 12.2 | 14.2 | 16.0 | |
| Ethnicity | .7 | ||||
| Other | 6.5 | 7.6 | 6.2 | 5.7 | |
| Puerto Rican | 93.5 | 92.4 | 93.9 | 94.3 | |
| Marital status | .6 | ||||
| Marriedd | 49.8 | 51.4 | 48.1 | 50.4 | |
| Separatede | 25.7 | 27.4 | 26.2 | 22.9 | |
| Single (never married) | 24.3 | 21.3 | 25.7 | 26.0 | |
| Socioeconomic status | |||||
| Area of residence | .1 | ||||
| Rural | 32.9 | 37.4 | 31.0 | 29.8 | |
| Urban | 67.1 | 62.3 | 69.0 | 70.2 | |
| Employment | <.0001*** | ||||
| Retired | 24.0 | 19.2 | 27.5 | 25.2 | |
| Employed | 45.4 | 39.5 | 44.7 | 53.8 | |
| Unemployed | 15.8 | 20.7 | 14.7 | 11.1 | |
| Other | 14.8 | 20.7 | 13.1 | 9.9 | |
| Health insurance | .2 | ||||
| No insurance | 4.3 | 2.7 | 3.5 | 7.3 | |
| Public Insurance | 37.5 | 47.4 | 32.4 | 32.4 | |
| Private or other | 58.2 | 49.9 | 64.2 | 60.3 | |
| Educational attainment | <.0001*** | ||||
| <High school | 19.1 | 25.2 | 15.2 | 16.8 | |
| High school | 8.7 | 11.9 | 6.4 | 8.0 | |
| Some college or associate’s | 25.9 | 31.0 | 23.5 | 22.9 | |
| College or more | 46.0 | 31.6 | 54.6 | 51.9 | |
| Income | <.0001*** | ||||
| $0–$10,000 | 30.0 | 41.6 | 24.3 | 23.3 | |
| $10,001–$20,000 | 26.4 | 25.8 | 28.9 | 23.7 | |
| >$20,000 | 40.8 | 28.6 | 44.7 | 50.8 | |
| Missing | 2.8 | 3.9 | 2.1 | 2.3 | |
| Health status and behavior | |||||
| Number of MDCDsf | |||||
| Ever had | 5.5 (3.8) | 7.5 (3.7) | 5.1 (3.5) | 3.4 (3.0) | <.0001*** |
| Currently have | 3.7 (2.9) | 5.4 (3.02) | 3.4 (2.6) | 2.2 (2.1) | <.0001*** |
| Social support g | |||||
| Social connectedness | 40.5 (9.4) | 38.2 (11.1) | 41.3 (8.6) | 42.1 (7.6) | <.0001*** |
| Social assurance | 35.2 (9.3) | 34.3 (9.9) | 35.6 (8.9) | 35.5 (8.9) | .125 |
| Drinking | .005** | ||||
| Never | 32.0 | 36.2 | 31.6 | 27.5 | |
| Past | 17.5 | 20.4 | 17.7 | 13.7 | |
| Current | 48.8 | 41.3 | 49.2 | 57.6 | |
| Smoking | .1 | ||||
| Never | 66.8 | 64.4 | 68.7 | 67.2 | |
| Past | 19.0 | 17.3 | 19.3 | 20.6 | |
| Current | 12.4 | 16.1 | 10.4 | 10.7 | |
MDCDs, medically diagnosed chronic diseases
Shown as percentage of sample (n=965). Sample sizes for individual variables may not add up to 100% owing to missingness
Shown as percentage of the row sample
Significant differences by self-rated health (“fair or poor,” “good,” “very good or excellent”) were determined from analysis of variance (ANOVA) or χ2 test and confirmed with Fisher exact test. Significance shown *P≤.05; **P≤.01; ***P≤.001
Category includes those who identify as married or living with a partner
Category includes those who identify as divorced, separated, or widowed
Number of conditions includes the total number of reported MDCDs, ever or currently diagnosed
The Social Connectedness/Social Assurance Scale1 is measured as a continuous score (range, 8–48, with 48 representing the highest measured degree of social connectedness/social assurance)
Prevalence of Self-Reported Cumulative Lifetime MDCDs
Conditions with the highest cumulative lifetime prevalence in at least one-quarter of the population, in descending order, included hypertension (45.8%), hyperlipidemia (42.0%), heartburn (40.9%), inflammation of the joints (37.1%), overweight/obesity (30.6%), arthritis (30.6%), anxiety (28.5%), sleep disorders (27.4%), depression (26.6%), and cardiovascular disease (25.0%; Table 2). Conditions with the lowest prevalence, in ascending order, included kidney disease (8.9%), cancer (9.4%), osteoporosis (11.6%), gallbladder disease (14.1%), eye disease (16.7%), and high blood sugar (17.1%), liver disease (18.3%), thyroid condition (22.0%), stomach problems (22.9%), respiratory problems (23.5%), and migraine headaches (23.9%). Most ever MDCDs were reported as current (vs past) diagnoses (Table 3), except for gallbladder disease and cancer (most of which were attributed to the past). Among those reporting fair or poor SRH, a statistically significantly greater proportion reported ever MDCD for every condition recorded, aside from osteoporosis and cancer (for which we did not find the difference in proportions reported to be significant).
Table 2.
Prevalence of self-reported medically diagnosed chronic diseases (ever vs never) among adults in Puerto Rico reporting fair or poor self-rated health by diagnostic status
| Self-rated health: fair or poor |
||||
|---|---|---|---|---|
| Overall prevalencea |
Ever diagnosedb | Never diagnosedb | P valuec | |
| Diabetes & endocrine | ||||
| Diabetes | 23.1 | 48.4 | 29.8 | <.0001*** |
| High blood sugar | 17.1 | 49.7 | 30.8 | <.0001*** |
| Thyroid condition | 22.0 | 43.9 | 30.9 | .0005*** |
| Digestive & gastrointestinal | ||||
| Overweight/obesity | 30.6 | 47.5 | 27.7 | <.0001*** |
| Stomach problems | 22.9 | 46.2 | 30.0 | <.0001*** |
| Gallbladder | 14.1 | 44.9 | 32.3 | .0006*** |
| Heartburn | 40.9 | 43.0 | 27.8 | <.0001*** |
| Heart & vascular | ||||
| Hypertension | 45.8 | 42.5 | 26.6 | <.0001*** |
| Hyperlipidemia | 42.0 | 41.0 | 28.0 | <.0001*** |
| Cardiovascular disease | 25.0 | 51.0 | 28.4 | <.0001*** |
| Kidney & liver | ||||
| Kidney disease | 8.9 | 54.7 | 31.6 | <.0001*** |
| Liver disease | 18.3 | 45.8 | 31.4 | <.0001*** |
| Nervous & musculoskeletal | ||||
| Migraine headaches | 23.9 | 50.2 | 28.9 | <.0001*** |
| Sleep disorder | 27.4 | 52.3 | 27.0 | <.0001*** |
| Arthritis | 30.6 | 47.5 | 26.9 | <.0001*** |
| Osteoporosis | 11.6 | 38.4 | 33.1 | .09 |
| Inflammation of the joints | 37.1 | 51.1 | 23.8 | <.0001*** |
| Mental | ||||
| Anxiety | 28.5 | 50.6 | 27.4 | <.0001*** |
| Depression | 26.6 | 51.8 | 27.4 | <.0001*** |
| Other | ||||
| Respiratory problems | 23.5 | 49.8 | 29.1 | <.0001*** |
| Eye disease | 16.7 | 47.2 | 30.8 | <.0001*** |
| Cancer | 9.4 | 37.4 | 33.7 | .3 |
Shown as percentage of sample (n=965). Sample sizes for individual variables may not add up to 100% owing to missingness
Shown as percentage of those who report fair or poor health among those who have ever been diagnosed with a given condition, versus among those never diagnosed
Significant differences by diagnostic status (ever diagnosed, never diagnosed) were determined from χ2 test and confirmed with Fisher exact test. Significance shown as *P≤.05; **P ≤.01; ***P≤.001
Table 3.
Prevalence of self-reported medically diagnosed chronic diseases (past, current, unknown, or never diagnosed) and related current medication use
| Never diagnosed (% of total)a |
Ever diagnosed (% of total)a |
Past diagnosis (% of ever)b |
Current diagnosis (% of ever)b |
Unknown history (% of total)b |
Taking medication (% of current)c |
|
|---|---|---|---|---|---|---|
| Diabetes & endocrine | ||||||
| Diabetes | 76.3 | 23.1 | 9.4 | 90.6 | 0.6 | 24.3 |
| High blood sugar | 81.5 | 17.1 | 24.8 | 75.2 | 1.5 | 11.3 |
| Thyroid condition | 76.9 | 22.0 | 9.0 | 91.0 | 1.1 | 19.2 |
| Digestive & gastrointestinal | ||||||
| Overweight/obesity | 67.7 | 30.6 | 12.2 | 87.8 | 1.8 | 2.7 |
| Stomach problems | 76.3 | 22.9 | 13.6 | 86.4 | 0.8 | 12.6 |
| Gallbladder | 85.4 | 14.1 | 83.1 | 16.9 | 0.5 | 21.7 |
| Heart & vascular | ||||||
| Hypertension | 53.4 | 45.8 | 7.5 | 92.5 | 0.8 | 25.7 |
| Hyperlipidemia | 56.2 | 42.0 | 20.7 | 79.3 | 1.9 | 12.1 |
| Cardiovascular disease | 74.8 | 25.0 | 33.2 | 66.8 | 0.2 | 13.7 |
| Kidney & liver | ||||||
| Kidney disease | 89.7 | 8.9 | 37.2 | 62.8 | 1.3 | 9.3 |
| Liver disease | 81.2 | 18.3 | 44.1 | 55.9 | 0.4 | 6.1 |
| Nervous & musculoskeletal | ||||||
| Migraine headaches | 75.5 | 23.9 | 19.9 | 80.1 | 0.5 | 16.8 |
| Sleep disorder | 71.8 | 27.4 | 8.0 | 92.0 | 0.8 | 7.4 |
| Arthritis | 67.4 | 30.6 | 4.1 | 95.9 | 2.1 | 8.1 |
| Osteoporosis | 86.8 | 11.6 | 5.4 | 94.6 | 1.6 | 9.4 |
| Mental | ||||||
| Anxiety | 70.5 | 28.5 | 10.5 | 89.5 | 1.0 | 13.0 |
| Depression | 72.1 | 26.6 | 26.1 | 73.9 | 1.2 | 19.5 |
| Other | ||||||
| Respiratory problems | 76.2 | 23.5 | 16.7 | 83.3 | 0.3 | 17.5 |
| Eye disease | 82.1 | 16.7 | 33.5 | 66.5 | 1.2 | 9.3 |
| Cancer | 90.2 | 9.4 | 76.9 | 23.1 | 0.4 | 19.0 |
Note: Heartburn and inflammation of the joints are not included in this table, as participants were only asked about current diagnosis/status (not about past diagnosis). Of those currently diagnosed with heartburn (n=395, 40.9%), 3.3% are taking medication; of those with painful inflammation of the joints (n=358, 37.1%), 2.0% are taking medication for it
Shown as percentage of sample (n=965). Sample sizes for individual variables may not add up to 100% owing to missingness
Shown as percentage of those ever diagnosed with respective condition
Shown as percentage of those currently diagnosed with respective condition
SRH by MDCD
Every additional cumulative lifetime MDCD reported (eg, ever diagnosis) was associated with an unadjusted estimated 44% increase in the odds of reporting fair or poor SRH (vs excellent or very good SRH; model 1: odds ratio [OR]=1.44 [95% CI, 1.36–1.53]; Table 4). Estimates produced through the adjusted model were nearly identical (models 2 and 3: OR=1.44; 95% CI, 1.35–1.54). Alternatively, every additional current condition was linked with an estimated 2% increased odds of reporting fair or poor SRH, both pre and post adjustment for confounding (model 3: OR=1.02; 95% CI, 1.01–1.03).
Table 4.
Odds ratio estimates of reporting poor or fair (vs very good or excellent) self-rated health among adults in Puerto Rico, by self-reported medically diagnosed chronic disease (n=965)
| Modelb | Model 2c | Model 3d | |
|---|---|---|---|
| Number of MDCDsa | |||
| Ever | 1.44 (1.36, 1.53)*** | 1.44 (1.35, 1.54)*** | 1.44 (1.35, 1.54)*** |
| Current | 1.02 (1.01, 1.03)* | 1.02 (1.01, 1.03)* | 1.02 (1.01, 1.03)* |
| Diabetes & endocrine | |||
| Diabetes | 2.58 (1.73, 3.86)*** | 2.45 (1.58, 3.81)*** | 2.46 (1.57, 3.83)*** |
| High blood sugar | 3.82 (2.29, 6.38)*** | 3.44 (2.01, 5.88)*** | 3.43 (2.00, 5.89)*** |
| Thyroid condition | 2.18 (1.44, 3.29)* | 2.01 (1.28, 3.14)* | 2.04 (1.30, 3.21)* |
| Digestive & gastrointestinal | |||
| Overweight/obesity | 3.44 (2.34, 5.06)*** | 3.81 (2.50, 5.81)*** | 3.84 (2.51, 5.88)*** |
| Stomach problems | 2.72 (1.79, 4.13)*** | 2.51 (1.62, 3.91)*** | 2.44 (1.57, 3.81)*** |
| Gallbladder | 2.49 (1.48, 4.18)* | 2.34 (1.35, 4.05)* | 2.34 (1.35, 4.05)* |
| Heartburn | 2.54 (1.80, 3.58)*** | 2.24 (1.55, 3.24)*** | 2.25 (1.55, 3.26)*** |
| Heart & vascular | |||
| Hypertension | 3.22 (2.28, 4.54)*** | 3.64 (2.47, 5.38)*** | 3.59 (2.43, 5.32)*** |
| Hyperlipidemia | 2.28 (1.62, 3.20)*** | 2.04 (1.41, 2.96)* | 2.10 (1.44, 3.06)* |
| Cardiovascular disease | 3.77 (2.48, 5.71)*** | 3.12 (2.00, 4.85)*** | 3.13 (2.01, 4.87)*** |
| Kidney & liver | |||
| Kidney disease | 4.76 (2.29, 9.92)*** | 4.60 (2.14, 9.86)*** | 4.64 (2.16, 9.97)*** |
| Liver disease | 2.65 (1.67, 4.19)*** | 2.31 (1.42, 3.76)* | 2.26 (1.38, 3.70)* |
| Nervous & musculoskeletal | |||
| Migraine headaches | 3.93 (2.55, 6.07)*** | 4.09 (2.54, 6.60)*** | 4.07 (2.52, 6.58)*** |
| Sleep disorder | 5.26 (3.42, 8.09)*** | 4.64 (2.95, 7.29)*** | 4.47 (2.83, 7.05)*** |
| Arthritis | 2.91 (2.01, 4.22)*** | 2.85 (1.86, 4.35)*** | 2.80 (1.83, 4.30)*** |
| Osteoporosis | 1.75 (1.01, 3.04)* | 1.16 (0.63, 2.15) | 1.13 (0.61, 2.09) |
| Painful inflammation | 5.51 (3.77, 8.06)*** | 4.96 (3.29, 7.48)*** | 4.95 (3.27, 7.48)*** |
| Mental & behavioral | |||
| Anxiety | 3.59 (2.43, 5.29)*** | 2.93 (1.90, 4.53)*** | 2.87 (1.85, 4.44)*** |
| Depression | 4.48 (2.94, 6.81)*** | 3.62 (2.29, 5.74)*** | 3.61 (2.28, 5.74)*** |
| Other | |||
| Respiratory problems | 2.77 (1.85, 4.13)*** | 2.51 (1.63, 3.86)*** | 2.45 (1.59, 3.79)*** |
| Eye disease | 2.49 (1.57, 3.96)* | 2.42 (1.44, 4.05)* | 2.42 (1.44, 4.06)* |
| Cancer | 1.48 (0.82, 2.66) | 1.54 (0.83, 2.85) | 1.52 (0.82, 2.84) |
Significance shown as *P≤.05; **P≤.01; ***P≤.001
Shown as odds ratio of reporting poor or fair (vs excellent) self-rated health among adults in Puerto Rico. All odds ratios are derived from logistic regression models, with self-rated health (“fair or poor” vs “very good or excellent” (reference) as the outcome and diagnostic status as the exposure. Diagnostic status is modeled as dichotomous (ever diagnosed vs never)
MDCDs, medically diagnosed chronic diseases
Number of conditions include the total number of reported MDCDs, ever or currently diagnosed
Model 1: Regression model 1 is unadjusted
Model 2: Regression model 2 adjusts for sex (male, female), employment status (retired, employed, unemployed, other), educational attainment (<high school, high school, some college or associate’s, college or more), income ($0–$10,000; $10,001–$20,000; >$20,000), health insurance (no insurance, public insurance, private or other), and social connectedness score
Model 3: Regression model 3 adjusts for everything included in model 2, in addition to history of drinking and smoking (never, past, current)
For all MDCDs measured, aside from cancer and osteoporosis, the estimated odds of reporting fair or poor SRH (vs excellent or very good) were significantly higher among those who reported ever diagnosis than those who reported never being diagnosed. Following stepwise adjustment for key confounders (model 3), statistically significant estimated ORs ranged from 2.04 (95% CI, 1.30–3.21, for thyroid conditions) to 4.95 (95% CI, 3.27–7.48, for painful inflammation). Listed in descending order of magnitude, OR estimates found to be statistically significantly associated with fair or poor SRH response were as follows: painful inflammation (OR=4.95 [95% CI, 3.27–7.48]), kidney disease (4.64 [2.16–9.97]), sleep disorder (4.47 [2.83–7.05]), migraine headaches (4.07 [2.52–6.58]), overweight/obesity (3.84 [2.51–5.88]), depression (3.61 [2.28–5.74]), hypertension (3.59 [2.43–5.32]), high blood sugar (3.43 [2.00–5.89]), cardiovascular disease (3.13 [2.01–4.87]), anxiety (2.87 [1.85–4.44]), arthritis (2.80 [1.83–4.30]), diabetes (2.46 [1.57–3.83]), respiratory problems (2.45 [1.59–3.79]), stomach problems (2.44 [1.57–3.81]), eye disease (2.42 [1.44–4.06]), gallbladder disease (2.34 [1.35–4.05]), liver disease (2.26 [1.38–3.70]), heartburn (2.25 [1.55–3.26]), hyperlipidemia (2.10 [1.44–3.06]), and thyroid condition (2.04 [1.30–3.21]).
Discussion
Our findings support using a single-item SRH measure as a rapid indicator of perceived chronic disease burden among adults residing in Puerto Rico. Nearly 1 in 3 participants in our study reported fair or poor SRH, closely matching the findings of a previous report in Puerto Rico.20 This high percentage observed in this study (34.1%), compared to the estimated 18% of non-Hispanic White peers in the mainland United States,21 may be a consequence of socioeconomic and psychosocial barriers linked with care access and outcomes. It is also possible that this gap reflects differences in the interpretation of the question and responses across different cultural contexts and/or languages.22
It should be noted that the likelihood of reporting poor SRH was not equal among all participants—disparities existed by socioeconomic and demographic characteristics. Specifically, those who identified as female, unemployed, lower income, and of middle age were more likely to report poor/fair SRH. This may be reflective of embedded socially-determined disparities in Puerto Rican and Latino populations living in the United States.23
However, our models adjusted for many of these factors and still noted significant associations between SRH and objective disease burden. The literature underscores protective cultural values (eg, allocentrism, simpatía, familism, religiousness),24 which have demonstrated a link with SRH25 and may feed one’s reserve capacity and enrich population-scale resilience demonstrated by predominantly Latino communities.26 Although we did not examine resiliency, we did incorporate the constructs of social connectedness and assurance into our models, though the minimal change was observed in our stepwise regression. Future longitudinal studies should test if these protective factors act as moderators in these associations.
As hypothesized, for every additional MDCD (ie, ever vs never diagnosed), we observed an adjusted estimated increase of 44% in the odds of reporting poor or fair SRH (vs excellent or very good). Alternatively, for every additional current MDCD reported, the adjusted odds increased only by an estimated 2%. This may suggest that cumulative lifespan experience with disease poses a more significant impact on the perception of current health status than current diagnosis alone, which aligns with the long duration and slow progression27 of most chronic conditions at the center of this study.
Regarding singular diagnoses, we observed higher odds of reporting poor or fair SRH (vs very good or excellent) among those ever (vs never) diagnosed for every condition included in the current study, aside from cancer and osteoporosis. The direction (ever diagnosis associated with lower SRH) and magnitude (ORs ranged from 2.04 to 4.95) of these associations align with a priori hypotheses and the reported findings of prior studies conducted in diverse contexts.28 We observed the strongest association between pain-related conditions (ie, painful inflammation, kidney disease, migraines) and SRH, which matches the results of prior studies on the multidimensional negative impact of chronic pain and poor quality of life and SRH, compared to patients with other long-term conditions.29 Notably, this higher burden of SRH fell mainly in nervous and musculoskeletal conditions. The lack of association observed between SRH versus cancer and osteoporosis could be due to the small sample size of these conditions, the wide variability in disease experience (eg, severity, stage, and management) across and within conditions,30 or that cancer was resolved (listed as a past condition) for most participants. The lack of change attributed to the addition of lifestyle and behavioral factors in the models aligns with the findings of previous researchers who also reported significant but negligible modifications after incorporating these.31
Our study presents some limitations. First, its cross-sectional design limits us from discerning temporality in this relationship; however, others who have examined longitudinal data in other contexts have reported similar associations.17 Further, we only report dichotomous measures of diagnosis history in this article; we do not report on the potential mediators of disease—including quality of life, and disease duration, severity, or management—which may play a role in the association of interest. Finally, we recognize that there is no objective measure of “true health”4; we use history of medical diagnoses as a proxy for this, though it is overly simplistic. To address this, we included vital social indicators and health behaviors as covariates; however, their position and role in the proposed pathway needs consideration in future longitudinal studies. Further, we do not have measures of disability or less common diseases, which likely also play a role in SRH, either through a comorbidity with MDCD or independently.32 Lastly, MDCD history is all self-reported, which may introduce measurement concerns when compared to more objective tools such as chart review.33 Although this is a commonly used data collection method in this setting and has been used by similar studies,33 validity could benefit from a medical records review.
We also recognize notable strengths of this study. To begin, this is one of the first studies on SRH in the context of Puerto Rico, which is a significant setting, as chronic disease rates remain high amid limited, disparate access to health care,11 especially amidst the recent outmigration of doctors to the continental United States following recent disasters.12 Additionally, the multistage sampling approach allowed our research team to enroll and interview a large sample of Puerto Rican adults across the island who remain underrepresented in the literature. That said, our sample predominantly identified as Puerto Rican and female, which limits the generalizability of our findings to other ethnicities and genders, and prevents our examination of the differences in these associations across other identities.7
In conclusion, SRH may reflect diagnosed MDCD burden and serve as a convenient screener, as it has been used for decades across the world,34 for clinicians to efficiently identify individuals in highest need of clinical services.6 Additionally, SRH may prove useful to surveillance researchers35 working towards the equitable distribution of health services and promotion resources across communities that need them most. This rapid identification of health status and related needs is especially important in the context of Puerto Rico’s health care system, strained by recent disasters and consequential outflow of trained clinicians.12
Acknowledgments
This work was supported by the National Heart, Lung, and Blood Institute (grants R01-HL143792 and K01-HL120951 to Josiemer Mattei), the National Institute on Minority Health and Health Disparities (grant R21-MD013650 to Josiemer Mattei), the Robert Wood Johnson Foundation (Culture of Health Leaders Award to Josiemer Mattei), the Harvard Education Program in Cancer Prevention Control, which is funded through a training grant (T32 CA057711 to Cristina Gago), and the CVD Epidemiology Training Program (T32 HL098048 to Martha Tamez). We also want to express appreciation for all the generous contributions made by our interviewers, partner clinics, staff members, and participants.
Footnotes
Conflict of Interest: No conflicts of interest reported by authors
Author Contributions: Research concept and design: Gago, Tamez, López-Cepero, Rodríguez-Orengo, Mattei; Acquisition of data: Rodríguez-Orengo, Mattei; Data analysis and interpretation: Gago, O’Neill, Tamez, López-Cepero, Mattei; Manuscript draft: Gago, O’Neill, Tamez, Rodríguez-Orengo; Statistical expertise: Gago, O’Neill, Tamez, López-Cepero, Mattei; Funding: Rodríguez-Orengo, Mattei; Administrative: Rodríguez-Orengo, Mattei; Technical support: Rodríguez-Orengo, Mattei; Supervision: Rodríguez-Orengo, Mattei
References
- 1.Jylhä M. What is self-rated health and why does it predict mortality: towards a unified conceptual model. Soc Sci Med. 2009; 69(3):307–316. [DOI] [PubMed] [Google Scholar]
- 2.Benyamini Y. Why does self-rated health predict mortality: an update on current knowledge and a research agenda for psychologists. Psychol Health. 2011; 26(11):1407–1413. [DOI] [PubMed] [Google Scholar]
- 3.Fayers PM, Hays R, Hays RD. Assessing Quality of Life in Clinical Trials: Methods and Practice. Oxford University Press; 2005. [Google Scholar]
- 4.Oftedal S, Kolt GS, Holliday EG, et al. Associations of health-behavior patterns, mental health and self-rated health. Prev Med. 2019; 118:295–303. [DOI] [PubMed] [Google Scholar]
- 5.Meyer OL, Castro-Schilo L, Aguilar-Gaxiola S. Determinants of mental health and self-rated health: a model of socioeconomic status, neighborhood safety, and physical activity. Am J Public Health. 2014; 104(9):1734–1741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chamberlain AM, Manemann SM, Dunlay SM, et al. Self‐rated health predicts healthcare utilization in heart failure. J Am Heart Assoc. 2014; 3(3):e000931. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Deeg DJ, Kriegsman DM. Concepts of self-rated health: specifying the gender difference in mortality risk. Gerontologist. 2003; 43(3):376–386. [DOI] [PubMed] [Google Scholar]
- 8.Abdulrahim S, Baker W. Differences in self-rated health by immigrant status and language preference among Arab Americans in the Detroit Metropolitan Area. Soc Sci Med. 2009; 68(12):2097–2103. [DOI] [PubMed] [Google Scholar]
- 9.Lim W-Y, Ma S, Heng D, Bhalla V, Chew SK. Gender, ethnicity, health behaviour & self-rated health in Singapore. BMC Public Health. 2007; 7(1):1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lo CC, Adame JL, Cheng TC. Explaining chronic illness and self-rated health among immigrants of five hispanic ethnicities. J Racial Ethn Health Disparities. 2020; 7(1):177–191. [DOI] [PubMed] [Google Scholar]
- 11.Mattei J, Tamez M, Ríos-Bedoya CF, Xiao RS, Tucker KL, Rodríguez-Orengo JF. Health conditions and lifestyle risk factors of adults living in Puerto Rico: a cross-sectional study. BMC Public Health. 2018; 18(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Rios CC, Ling EJ, Rivera-Gutierrez R, et al. Puerto Rico health system resilience after Hurricane Maria: implications for disaster preparedness in the COVID-19 era. Front Commun. 2021; 5:593939. [Google Scholar]
- 13.Bowling A. Just one question: if one question works, why ask several? Journal. 2005:59:342–345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mattei J, Tucker KL, Falcón LM, et al. Design and implementation of the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic Disease Trends (PROSPECT). Am J Epidemiol. 2021; 190(5):707–717. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lee RM, Robbins SB. Measuring belongingness: the social connectedness and the social assurance scales. J Counsel Psychol. 1995; 42(2):232. [Google Scholar]
- 16.Tucker KL, Mattei J, Noel SE, et al. The Boston Puerto Rican Health Study, a longitudinal cohort study on health disparities in Puerto Rican adults: challenges and opportunities. BMC Public Health. 2010; 10(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wu S, Wang R, Zhao Y, et al. The relationship between self-rated health and objective health status: a population-based study. BMC Public Health. 2013; 13(1):1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN. The Self‐Administered Comorbidity Questionnaire: a new method to assess comorbidity for clinical and health services research. Arthritis Care Res. 2003; 49(2):156–163. [DOI] [PubMed] [Google Scholar]
- 19.Chan YY, Teh CH, Lim KK, et al. Lifestyle, chronic diseases and self-rated health among Malaysian adults: results from the 2011 National Health and Morbidity Survey (NHMS). BMC Public Health. 2015; 15(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kaiser Family Foundation.. Puerto Rico: Fast Facts. 2017. Last accessed June 22, 2022. from https://www.kff.org/racial-equity-and-health-policy/fact-sheet/puerto-rico-fast-facts/#footnote-238005-7
- 21.Shetterly SM, Baxter J, Mason LD, Hamman RF. Self-rated health among Hispanic vs non-Hispanic white adults: the San Luis Valley Health and Aging Study. Am J Public Health. 1996; 86(12):1798–1801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Bzostek S, Goldman N, Pebley A. Why do Hispanics in the USA report poor health? Soc Sci Med. 2007; 65(5):990–1003. [DOI] [PubMed] [Google Scholar]
- 23.Eisenberg-Guyot J, Prins SJ. Relational social class, self-rated health, and mortality in the United States. Int J Health Serv. 2020; 50(1):7–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Marin G, Marin BV. Research with Hispanic Populations. Sage Publications; 1991. [Google Scholar]
- 25.Zullig KJ, Ward RM, Horn T. The association between perceived spirituality, religiosity, and life satisfaction: the mediating role of self-rated health. Soc Indic Res. 2006; 79(2):255–274. [Google Scholar]
- 26.Gallo LC, Penedo FJ, Espinosa de los Monteros K, Arguelles W. Resiliency in the face of disadvantage: do Hispanic cultural characteristics protect health outcomes? J Pers. 2009; 77(6):1707–1746. [DOI] [PubMed] [Google Scholar]
- 27.Wei MY, Kawachi I, Okereke OI, Mukamal KJ. Diverse cumulative impact of chronic diseases on physical health–related quality of life: implications for a measure of multimorbidity. Am J Epidemiol. 2016; 184(5):357–365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mavaddat N, Valderas JM, Van Der Linde R, Khaw KT, Kinmonth AL. Association of self-rated health with multimorbidity, chronic disease and psychosocial factors in a large middle-aged and older cohort from general practice: a cross-sectional study. BMC Fam Pract. 2014; 15(1):1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hadi MA, McHugh GA, Closs SJ. Impact of chronic pain on patients’ quality of life: a comparative mixed-methods study. J Patient Exp. 2019; 6(2):133–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Uzunel E, Lundin H, Wändell P, Salminen H. Association between self-rated health and the risk of hip fracture and mortality in a cohort of older women during a 10-year follow-up. PLoS One. 2021; 16(3):e0247924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cott CA, Gignac MA, Badley EM. Determinants of self rated health for Canadians with chronic disease and disability. J Epidemiol Commun Health. 1999; 53(11):731–736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Galenkamp H, Braam AW, Huisman M, Deeg DJ. Seventeen-year time trend in poor self-rated health in older adults: changing contributions of chronic diseases and disability. Eur J Public Health. 2013; 23(3):511–517. [DOI] [PubMed] [Google Scholar]
- 33.Leikauf J, Federman AD. Comparisons of self‐reported and chart‐identified chronic diseases in inner‐city seniors. J Am Geriatr Soc. 2009; 57(7):1219–1225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.May M, Lawlor DA, Brindle P, Patel R, Ebrahim S. Cardiovascular disease risk assessment in older women: can we improve on Framingham: British Women’s Heart and Health prospective cohort study. Heart. 2006; 92(10):1396–1401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Robine J-M, Jagger C, Group E-R. Creating a coherent set of indicators to monitor health across Europe: the Euro-REVES 2 project. Eur J Public Health. 2003; 13(suppl 3):6–14. [DOI] [PubMed] [Google Scholar]