Abstract
People with HIV (PWH) have an elevated risk for cardiovascular disease (CVD) compared with the general population. This study examined the feasibility, acceptability and preliminary efficacy of a tailored intervention aimed at increasing CVD risk perception and the adoption of heart-healthy behaviors in PWH. Forty adults were randomized to receive personalized feedback on CVD risk and discussion of risk reduction or health education. Participants were issued pedometers and seen for two treatment sessions. Participants were 60% male and had a mean age of 51.5 years. Ninety percent of participants completed all study sessions indicating good feasibility and acceptability. A medium effect size for the difference between treatment and control groups was found on both the Perceived Risk for Heart Disease (d=.38) and the Rapid Eating and Activity for Patients scales (d=.56) at 12 weeks. Atherosclerotic cardiovascular disease (ASCVD) risk score moderated the effect of treatment, such that at high (but not low) ASCVD risk, active intervention, compared to control, was associated with a greater increase in steps between baseline and both 8 (d=.38) and 12 weeks (d=.55). Findings provide preliminary evidence that tailored interventions delivered by nurses may be effective for primary prevention of CVD in PWH.
Keywords: HIV, cardiovascular disease, primary prevention, clinical trial, pilot studies
People with HIV (PWH) are at elevated risk for cardiovascular diseases (CVDs) compared with the general population,(Feinstein et al., 2016) and CVDs are a leading cause of non-HIV-related deaths in PWH(Hatleberg, Lundgren, & Ryom, 2017). PWH are twice as likely to have a myocardial infarction(Paisible et al., 2015), and one-third are at moderate to high 10-year risk of having a cardiac event(Kaplan et al., 2007). Traditional risk factors are highly prevalent among this population, and significantly contribute to the elevated risk(Althoff et al., 2019; Mdodo et al., 2015; Pacek & Cioe, 2015; Rasmussen et al., 2015).
Primary prevention guidelines to achieve cardiovascular health include: 150 minutes or more of weekly physical activity (PA), body mass index (BMI) below 25, blood pressure below 130/85, and abstinence from smoking(Eckel RH et al., 2013; Feinstein et al., 2019). Given the importance of lifestyle modification, it is imperative to test interventions that increase CVD risk perception and the heart-healthy behaviors in PWH. A systematic review of PA interventions for PWH found PA can improve cardiorespiratory fitness, mood, lipids, and quality of life(O’Brien, Tynan, Nixon, & Glazier, 2016). Interestingly, a recent, tailored in-person intervention focused on lifestyle behaviors to reduce CVD risk in PWH was effective at reducing body weight, but not for PA or diet improvement(A. R. Webel et al., 2018).
The purpose of this study was to test feasibility, acceptability, and preliminary efficacy of the Cardiovascular Disease Perceived Risk Awareness Intervention (CVD-PRAI), which utilized personalized feedback and motivational interviewing techniques to improve CVD risk perception and adoption of heart-healthy behaviors in PWH. CVD-PRAI, a tailored intervention based on the Health Belief Model, was developed for PWH using qualitative methods(Cioe, Guthrie, Freiberg, Williams, & Kahler, 2018). We hypothesized that participants in CVD-PRAI would demonstrate equal or greater treatment satisfaction, improved CVD risk perception, and increased PA and diet, compared with control.
METHODS
Study Design.
A pilot randomized clinical trial design was used. Outcomes were assessed at 4, 8, and 12 weeks following baseline. Study procedures were approved by the Brown University IRB.
Participants.
Participants were recruited from two outpatient sites in Providence, RI. Eligibility criteria included: over age 30; HIV-positive; able to complete informed consent in English; and taking antiretroviral therapy with an undetectable HIV viral load. Patients were excluded if they had an established diagnosis of CVD, cerebral vascular accident, or peripheral vascular disease. Sixty-seven subjects were screened; 42 were eligible. Two did not return for treatment after baseline and were not randomized. Nineteen were randomized to the active condition and 21 to control (Figure 1). The 12-week combined retention rate was 97.5%.
Figure 1.
CONSORT diagram showing participant flow. CONSORT = Consolidated Standards of Reporting Trials; CVD-PRAI = Perceived Risk Awareness Intervention
Procedures.
Participants were recruited from August 2016 through February 2017. A research assistant (RA) screened participants by phone. At the baseline visit, participants completed informed consent, were screened to confirm eligibility, and completed baseline measures. A pedometer was issued, with instructions wear it daily, remove at bedtime, and maintain usual level of activity for the first week (to obtain a baseline PA level). Participants were randomized using a computerized urn randomization program(Stout, Wirtz, Carbonari, & Del Boca, 1994), based on atherosclerotic cardiovascular disease (ASCVD) risk score and gender. Participants attended a treatment session with the study nurse one week later and received daily text messages during week 1, and weekly during weeks 2–4 that reminded them to wear their pedometer. A second treatment session took place at week 4. Follow-up assessments with the RA (blinded to study assignment) were completed in-person at weeks 4 and 12, and by phone at week 8. Participants were compensated for completed sessions.
Active Condition.
Treatment sessions (45 minutes each) incorporated personalized feedback and motivational interviewing(Miller & Rollnick, 2002). Participants’ ASCVD risk score and Heart Age(Centers for Disease Control and Prevention, 2015) were calculated and feedback on level of risk and modifiable CVD risk factors were discussed. Emphasis was on personal responsibility for behavior change; advice was given to reduce CVD risk behaviors(Eckel RH et al., 2013). Participants were encouraged to set goals and literature was provided. Smoking cessation resources were provided to smokers, if interested. Session 2 involved a summary of the prior session, review of goals, addressing barriers to change, and discussion of strategies for maintaining long-term behavior change.
Control Condition.
Treatment sessions (20 minutes) included general information about heart disease, encouragement to follow a healthy diet, increased PA, and smoking cessation, as appropriate. Literature was provided. The follow-up session included a review of the prior session and advice to improve heart-healthy behaviors.
MEASURES
Demographic and Clinical Characteristics.
Demographic variables were self-reported. The CD4 T-cell count, HIV viral load, and lipids were collected from the medical record. Height and weight were measured and BMI was calculated(National Heart Lung and Blood Institute, 2017). Baseline depressive symptoms were assessed with the Center for Epidemiologic Studies Depression Scale(Radloff, 1977) (CES-D).
Perceived Cardiovascular Risk.
Perceived risk was measured at baseline, week 4 (just prior to the second treatment session), and week 12. The 20-item Perception of Risk of Heart Disease Scale (PRHDS) measures an individual’s perception of risk of developing heart disease (Ammouri & Neuberger, 2008). It has good test-retest reliability (.61 to .76) and internal consistency (Cronbach alpha = .80). Items are rated on a 4-point Likert scale (1=strongly agree to 4=strongly disagree). Sum scores range from 20–80; higher scores indicate increased risk perception.
Estimated Risk of CVD.
The ASCVD risk score was calculated using the American College of Cardiology/American Heart Association (AHA) Risk Estimator (Goff et al., 2014). This equation estimates the likelihood of having a CVD event (coronary death, nonfatal myocardial infarction, fatal or nonfatal stroke), based on the Pooled Cohort Equations and lifetime risk prediction tools. A score greater than 7.5 is considered higher risk.
PA levels.
Omron Tri Axis Digital Pedometers, which store seven days of activity data and reset each day at midnight, were used. At weeks 1, 4 and 12, the RA reviewed the pedometers to obtain weekly totals for each participant. At week 8, participants self-reported each day’s step count for the past week.
Dietary Behaviors.
Dietary behavior was measured at baseline, week 4, and week 12. The 16-item Rapid Eating and Activity Assessment for Patients (REAP)(Gans et al., 2006; Segal-Isaacson, Wylie-Rosett, & Gans, 2004) is designed for primary care use. It has excellent test-retest reliability (r = 0.86, p < .0001), correlating well with the Healthy Eating Index (r = 0.49).
Treatment Satisfaction.
The 8-item Client Satisfaction Question (CSQ-8) measures satisfaction with various health and human services (Larsen, Attkisson, Hargreaves, & Nguyen, 1979). It has excellent reliability and internal consistency (Cronbach’s alpha = .93) across diverse patient samples. Response options are on a 4-point scale (quite dissatisfied=1 to very satisfied=4). Scores range from 8 to 32; higher scores indicate higher satisfaction.
Total session attendance was measured as the number of sessions completed (two treatment, two assessment). The number and percentage of participants who completed all four sessions was measured.
DATA ANALYSIS
Descriptive analyses were performed using SPSS V.24. Trends in perceived risk scores and level of PA from baseline to week 4, week 8, and week 12 were examined. Hierarchical linear models (HLM) were run in the HLM 7.01 program(Raudenbush, Bryk, & Congdon, 2013), allowing us to retain for analysis participants that contributed at least one follow-up assessment. The person-period dataset for models predicting PRHD and REAP included 120 possible observations (N=40 participants*3 assessments). Data were missing due to failure to complete surveys for only 6 PRHD and 7 REAP observations (5–6%). The person-period dataset for models predicting steps included 160 possible observations (N=40 participants*4 assessments), with 27 observations (17%) missing. All outcomes were normally distributed.
Fully unconditional HLM models were used to determine intraclass correlations (ICCs) for outcomes. For models predicting PRHD and REAP (baseline, 4 weeks, 12 weeks), two dummy-coded time components (4 weeks coded 0, 1, 0; 12 weeks coded 0, 0, 1) were added at Level 1 to represent change from baseline to week 4 week and week 12, respectively. To test main effects, condition was added at Level 2 as a predictor of the intercept (i.e., controlling for any difference in conditions on baseline levels of outcome) and both time effects (i.e., effect of group on change in outcome between baseline and 4- and 12-weeks, respectively). Our approach for models predicting steps was similar, but given 4 assessments, we modeled three dummy-coded time components (4 weeks coded 0, 1, 0, 0; 8 weeks coded 0, 0, 1, 0; 12 weeks coded 0, 0, 0, 1).
All intercept and slope effects were initially specified as random, to allow for individual variation in baseline levels and change over time in the outcomes; however, they were fixed in cases of non-significant variance. Outcomes were standardized, such that model coefficients representing differences between groups could be interpreted as effect size ds (.2=small, .5=medium, .8=large)(Cohen, 1988). For this pilot trial, we focus most on effect sizes, although we also present p-values for each result. With a small sample size, potentially meaningful effect sizes often may not be statistically significant(Sullivan & Feinn, 2012). With limited statistical power, we were interested in whether differences between intervention and control tended to be of a magnitude that might support a future full-scale intervention trial.
In a second set of exploratory models for each outcome, we examined baseline ASCVD as a moderator of the effect of condition on each outcome, by including the ASCVD and interaction term as additional effects on the intercept and time components. Significant and marginally significant interactions were probed at one standard deviation above and below the mean on the moderator(Aiken & West, 1991), in order to calculate effect sizes for simple slopes. In the absence of significant interactions, tabled results show main effect models for parsimony.
RESULTS
Demographic and clinical characteristics of the sample are shown in Tables 1 and 2. Mean age differed by group with those in CVD-PRAI significantly younger; t-tests (M = 48.8 [SD 7.2] vs. 53.9 [SD 6.9], p = .03). ICCs were .53, .35, and .70 for PRHD, REAP, and steps, respectively. In all cases, adequate variability at both the between- and within-person levels suggested that multilevel modeling was appropriate.
Table 1.
Sample demographics at baseline (N = 40).
| Total (N = 40) | CVD-PRAI (n = 19) | Brief Advice (n = 21) | |
|---|---|---|---|
| Mean Age* | 51.5 (7.4) | 48.8 (7.1) | 53.9 (7.0) |
| Mean Education Level | 12.2 (2.2) | 12.4 (1.8) | 12.1 (2.5) |
| Mean CD4 T-cell count | 738 (351.9) | 702 (225.7) | 771 (439.5) |
| N (%) | N (%) | N (%) | |
| Gender | |||
| Male | 24 (60.0) | 12 (63.2) | 12 (57.1) |
| Female | 16 (40.0) | 7 (36.8) | 9 (42.9) |
| Ethnicity | |||
| Hispanic/Latino | 7 (17.5) | 4 (21.1) | 3 (14.3) |
| Not Hispanic/Latino | 33 (82.5) | 15 (78.9) | 18 (85.7) |
| Race | |||
| White | 18 (45.0) | 7 (36.8) | 11 (52.4) |
| Black | 17 (42.5) | 8 (42.1) | 9 (42.9 |
| More than one race | 3 (7.5) | 3 (15.8) | 0 |
| Unknown | 1 (2.5) | 1 (5.3) | 0 |
| Education Level | |||
| High School or less | 28 (70.0) | 12 (63.2) | 16 (76.2) |
| Bachelor’s Degree or higher | 4 (10.0) | 1 (5.3) | 3 (14.3) |
| Employment Status | |||
| Full-time | 7 (17.5) | 4 (21.1) | 3 (14.3) |
| Part-time | 5 (12.5) | 2 (10.5) | 3 (14.3) |
| Unemployed | 3 (7.5) | 2 (10.5) | 1 (4.8) |
| Retired | 2 (5.0) | 2 (10.5) | 0 |
| Disabled | 22 (55.0) | 8 (42.1) | 14 (66.7) |
| Relationship Status | |||
| Married | 9 (22.5) | 5 (26.3) | 4 (19.0) |
| Single/Divorced/Widowed | 31 (77.5) | 14 (73.8) | 17 (80.9) |
| Time Since HIV Diagnosis | |||
| ≤20 years | 26 (65.0) | 13 (68.4) | 8 (38.1) |
| ≥21 years | 14 (35.0) | 6 (31.6) | 13 (61.9) |
| Smoking Status | |||
| Never Smoked | 12 (30.0) | 6 (31.6) | 6 (28.6) |
| Current Smoker | 17 (42.5) | 7 (36.8) | 10 (47.6) |
| Quit in Last 30 Days | 1 (2.5) | 1 (5.3) | 0 |
| Quit >30 Days Ago | 10 (25) | 5 (26.3) | 5 (23.8) |
| HIV Viral load | |||
| ≤20 | 36 (90.0) | 18 (94.7) | 18 (85.7) |
| ≥20 | 4 (10.0) | 1 (5.3) | 3 (14.3) |
| CD4 T-cell count | |||
| Less than 200 | 1 (2.5) | 0 | 1 (4.8) |
| 201–499 | 9 (22.5) | 4 (21.1) | 5 (23.8) |
| Above 500 | 30 (75.0) | 15 (78.9) | 15 (71.4) |
BA: Brief Advice; CVD-PRAI: Cardiovascular Disease Perceived Risk Awareness Intervention
p < .05
Table 2.
Clinical characteristics at baseline (N = 40).
| Total (N = 40) | |
|---|---|
| Years with HIV (M, SD) | 17.4 (8.1) |
| Body mass index (M, SD) | 28.7 (6.7) |
| ASCVD risk score (M, SD) | 9.5 (9.7) |
| Total Chol (M, SD) | 177.0 (36.5) |
| HDL (M, SD) | 42.5 (13.2) |
| LDL (M, SD) | 102.4 (32) |
| Triglycerides (M, SD) | 155.0 (85.4) |
| CES-D (M, SD)) | 16.8 (12.2) |
| Hypertension (n, %) | 9 (22.5) |
| Type 2 diabetes (n, %) | 6 (15) |
| COPD (n, %) | 3 (7.5) |
| Metabolic syndrome (n, %) | 16 (40) |
Treatment satisfaction was high (M = 29.5 (SD 3.5); possible total score of 32. Participants in CVD-PRAI were as satisfied with treatment as those in control. Session attendance was high; mean sessions attended was 3.9 (SD .30). Thirty-six (90%) participants completed all sessions; there was no study dropout and follow-up rates did not differ by condition.
Mean PRHD was 53.2 (SD 5.9) at baseline, 54.1 (SD 5.6) at week 4, and 52.6 (SD 5.0) at week 12, indicating a slight increase in perceived risk at week 4. The mean weekly steps at baseline, weeks 4, 8, and 12, were: 26,600 (SD 18,547); 33,601 (SD 28,251); 32,864 (SD 25,037); and 28,982 (SD 28232), respectively, indicating a peak in PA at week 4 with a slight decrease at weeks 8 and 12, but remaining higher than baseline. The mean REAP score was 25.7 (SD 3.8) at baseline, 27.2 (SD 4.4) at week 4, and 27.1 (SD 4.2) at week 12. REAP scores improved between baseline and week 4 (p = .058), and between baseline and week 12 (p = .06) among both groups. The change approached significance at each timepoint but did not differ significantly by treatment condition at any timepoint.
Primary Outcome Model
Predicting PRHD.
Effects on PRHD are shown in Table 3; as outcomes were standardized, coefficients are interpreted as effect sizes (i.e., between-group differences in standard units for each outcome). The effect of CVD-PRAI on change in PRHD between baseline and 4 weeks was of trivial size (β=.08) and non-significant. The effect of CVD-PRAI on change between baseline and 12 weeks, though non-significant, was of small size (β=.38). Those in the control condition showed a reduction in PRHD (−.30), while those in CVD-PRAI increased a trivial amount on PRHD (−.30 + .38 = .08) more so than those in the control group over this longer follow-up. Moderator analyses revealed no significant interactions between group and baseline ASCVD on change in PRHD between baseline and either 4- or 12- weeks.
Table 3.
HLM Model Predicting PRHD
| B | SE | t | df | p | |
|---|---|---|---|---|---|
| Intercept (BL level) | −0.01 | 0.26 | −0.05 | 38 | 0.96 |
| Effect of group | −0.01 | 0.33 | −0.04 | 38 | 0.97 |
| Change from BL to 4 wks | 0.14 | 0.21 | 0.65 | 38 | 0.52 |
| Effect of group | 0.08 | 0.32 | 0.24 | 38 | 0.81 |
| Change from BL to 12wks | −0.30 | 0.17 | −1.80 | 32 | 0.08 |
| Effect of group | 0.38 | 0.29 | 1.33 | 32 | 0.19 |
Note: BL = baseline, PRHD = Perceived risk of heart disease. Group is coded 0 for control, 1 for treatment; outcomes were standardized for analyses, such that model coefficients representing differences between groups could be interpreted as effect sizes (.2=small, .5=medium, .8=large)
Secondary Outcome Models
Predicting PA level.
Effects on PA level/step counts are shown in Table 4, left side. The size of effects of treatment on change in steps between baseline and 4 weeks was small (β=.24), and although statistically non-significant, suggested a promising increase in steps in the short-term among those in treatment vs control. Effects of treatment on steps between baseline and both 8 weeks (β= −.01) and 12 weeks (β= −.10) were trivial in size and non-significant.
Table 4.
HLM Model Predicting PA/Steps.
| Main Effect Model | Moderator Model | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| β | SE | t | df | p | B | SE | t | df | p | |
| Intercept (BL level) | −0.23 | 0.14 | −1.69 | 38 | 0.09 | −0.22 | 0.14 | −1.61 | 36 | 0.11 |
| Effect of group | 0.09 | 0.25 | 0.39 | 38 | 0.69 | 0.08 | 0.25 | 0.31 | 36 | 0.75 |
| Effect of BL ASCVD | −0.00 | 0.01 | −0.45 | 36 | 0.65 | |||||
| Group x ASCVD | −0.00 | 0.02 | −0.38 | 36 | 0.70 | |||||
| Change from BL to 4wk | 0.21 | 0.15 | 1.37 | 38 | 0.18 | 0.20 | 0.16 | 1.25 | 36 | 0.21 |
| Effect of group | 0.24 | 0.25 | 0.94 | 38 | 0.36 | 0.24 | 0.24 | 1.00 | 36 | 0.32 |
| Effect of BL ASCVD | −0.00 | 0.00 | −0.68 | 36 | 0.49 | |||||
| Group x ASCVD | 0.02 | 0.02 | 0.98 | 36 | 0.33 | |||||
| Change from BL to 8wk | 0.26 | 0.14 | 1.79 | 38 | 0.08 | 0.26 | 0.15 | 1.75 | 36 | 0.08 |
| Effect of group | −0.00 | 0.25 | −0.02 | 38 | 0.98 | 0.02 | 0.25 | 0.10 | 36 | 0.92 |
| Effect of BL ASCVD | −0.01 | 0.01 | −1.18 | 36 | 0.24 | |||||
| Group x ASCVD | 0.03 | 0.02 | 2.27 | 36 | 0.02 | |||||
| Change from BL to 12wk | 0.14 | 0.23 | 0.61 | 38 | 0.55 | 0.16 | 0.22 | 0.69 | 36 | 0.49 |
| Effect of group | −0.10 | 0.33 | −0.31 | 38 | 0.76 | −0.08 | 0.32 | −0.26 | 36 | 0.79 |
| Effect of BL ASCVD | −0.03 | 0.01 | −1.86 | 36 | 0.07 | |||||
| Group x ASCVD | 0.07 | 0.02 | 2.82 | 36 | 0.00 | |||||
Note: BL = baseline, ASCVD = Atherosclerotic Cardiovascular Disease; outcomes were standardized for analyses, such that model coefficients representing differences between groups could be interpreted as effect sizes (.2=small, .5=medium, .8=large).
As shown in Table 4, right side, there were significant interactions between treatment and baseline ASCVD on change in steps between baseline and both 8 weeks (β= 0.03) and 12 weeks (β=0.07). While probing the 8-week interaction (not shown in tables) did not reveal statistically significant effects at either high or low levels of ASCVD, at high levels of ASCVD, there was a positive, small effect of treatment (β=.36, p=.349) suggesting more of an increase in steps between baseline and 8 weeks in treatment vs control. At low levels of ASCVD, there was a negative, small effect of treatment (β =−.31, p=.395), suggesting less of an increase in steps between baseline and 8 weeks in treatment vs control. Probing the 12-week interaction revealed that at high ASCVD, there was a moderately sized but non-significant increase (β=.55, p=.243) in steps between baseline and 12 weeks in CVD-PRAI compared to control. At low ASCVD, there was significantly less of an increase in steps (β= - .72, p=.122) in CVD-PRAI compared to control.
Predicting REAP.
Effects on REAP are shown in Table 5. The effect of treatment on change on REAP between baseline and 4 weeks was trivial (β=.08) and non-significant, but there was a moderately sized treatment effect (β=.56) on change between baseline and 12 weeks that approached significance. Moderator analyses revealed no significant interactions between group and baseline ASCVD on change in REAP between baseline and either 4 or 12 weeks.
Table 5.
HLM Model Predicting REAP
| β | SE | t | df | p | |
|---|---|---|---|---|---|
| Intercept (BL level) | −0.06 | 0.22 | −0.27 | 38 | 0.79 |
| Effect of group | −0.36 | 0.28 | −1.29 | 38 | 0.21 |
| Change from BL to 4 wks | 0.30 | 0.27 | 1.09 | 38 | 0.28 |
| Effect of group | 0.1 | 0.35 | 0.40 | 38 | 0.69 |
| Change from BL to 12wks | 0.07 | 0.25 | 0.26 | 30 | 0.79 |
| Effect of group | 0.56 | 0.32 | 1.76 | 30 | 0.09 |
Note: BL = baseline, Group is coded 0 for control, 1 for treatment
DISCUSSION
To our knowledge, this is the first prospective study to incorporate personalized feedback and motivational interviewing to improve CVD risk perception and heart-healthy behaviors in PWH. While a previous systematic review noted dropout rates approaching 30%(Vancampfort, Mugisha, Richards, De Hert, Lazzarotto, et al., 2017), we report high study retention and satisfaction. Recently, investigators presented a qualitatively developed “human-centered design” approach for a nurse-led CVD prevention clinic in which recommendations for a tailored approach for PWH were made(Aifah et al., 2020). They recommended several key components, namely, integration of a prevention nurse in the HIV clinic, use of the ASCVD risk score, provision of educational materials, and use of a “three key take-home points” summary for patients following the prevention consultation. Interestingly, our CVD-PRAI tailored intervention incorporated all these features and this likely contributed to our success.
Our study provides initial evidence that compared to brief advice, two study sessions with personalized feedback can result in some improvement in both perceived risk and PA levels. Other studies have shown that supervision throughout the study period by trained professionals were important predictors of treatment effect(Vancampfort, Mugisha, Richards, De Hert, Lazzarotto, et al., 2017), however continuous supervision may be costly and impractical for widespread dissemination of effective interventions. Identification of cost-effective factors, such as personalized feedback and utilization of readily available resources in the HIV clinic (such as nurses), are important. Our study documented significant increases in step counts in the first month, however, step counts trended downward following the 8-week phone session until the 12-week follow-up. In PWH, it has been noted that pain and depressive symptoms are associated with lower levels of PA(Vancampfort, Mugisha, Richards, De Hert, Probst, et al., 2017). Future studies should examine the relationship of these variables and PA levels over time. Similar downward trends after 6 weeks were noted in an intervention for women with HIV(Mabweazara, Leach, Ley, & Smith, 2018). A systematic review in the general population suggested that intervention delivery through increased personal contact increased effectiveness(Bull et al., 2018). Perhaps, additional, frequent follow-up sessions with the nurse are needed to sustain activity.
A recent study found modest associations between ASCVD risk perception and CVD prevention behaviors(A. Webel et al., 2020). While we did not see a significant relationship between perceived risk and PA, we noted a marked effect for those with higher ASCVD scores; participants with higher risk (vs. lower CV risk) experienced greater increases in step counts. The effect for those at higher ASCVD risk (scores greater than 7.5%) is especially promising. Future research is needed to elucidate how to structure health promotion advice so that it is salient for patients at all risk levels. Despite these observations, CVD-PRAI appears to be feasible and effective, promoting increased CVD risk perception and increased PA levels. It could be readily adapted for all PWH and delivered by clinic nurses where they receive their care. Further testing with larger samples is warranted.
Consistent with the literature, the prevalence of metabolic syndrome in this sample was high (40%), and studies have found that this is associated with lower PA(Forde et al., 2018). A multiple behavior approach, addressing PA and healthy diet, has been shown to be most effective in prompting behavior change(Vancampfort, Mugisha, Richards, De Hert, Probst, et al., 2017). CVD-PRAI may have increased effectiveness if sustained for a longer period of time.
This study had multiple strengths. It had swift recruitment, strong engagement, and excellent retention. The intervention, although more in-depth than other interventions tested in PWH(Vancampfort, Mugisha, Richards, De Hert, Probst, et al., 2017), required minimal training and only about two hours of total participant contact time. An objective measure of PA was used, as opposed to self-report which has often been relied upon in other studies. The study however had some limitations. Our ability to see the full intervention effect may have been limited by the small sample. The two groups differed in mean age, and we don’t know how this affected the findings. The study was conducted in the Northeast U.S. with patients that were engaged in care and may not be reflective of all PWH.
In summary, a nurse-directed program focused on CVD risk reduction using personalized feedback may result in improvement in CVD risk perception, PA levels, and dietary behaviors among PWH without heart disease. Further research is needed with larger samples and increased focus on smoking cessation to test the extent to which a nurse-led program can improve ASCVD risk and long-term clinical outcomes, such as reduced CVD events. Further research should also examine whether regular assessment of PA and diet in clinical settings leads to reduction in CVD risk, the intervention intensity needed to produce maximal change in behavior, and cost-effectiveness/sustainability. This approach, if found to be effective, could be integrated into clinical care and be readily disseminated given the availability of nurses in most HIV clinics in the U.S. and globally.
Acknowledgments
FUNDING: This project was supported by grant number K23NR014951 from the National Institute for Nursing Research. This work was facilitated by the Providence/Boston Center for AIDS Research (P30AI042853).
REFERENCES
- Aifah A, Okeke NL, Rentrope CR, Schexnayder J, Bloomfield GS, Bosworth H, . . . Vedanthan R. (2020). Use of a human-centered design approach to adapt a nurse-led cardiovascular disease prevention intervention in HIV clinics. Prog Cardiovasc Dis, 63(2), 92–100. doi: 10.1016/j.pcad.2020.02.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Aiken LS, & West SG (1991). Multiple regression: Testing and interpreting interactions. . Thousand Oaks, CA: Sage Publications, Inc. [Google Scholar]
- Althoff KN, Gebo KA, Moore RD, Boyd CM, Justice AC, Wong C, . . . Gange SJ (2019). Contributions of traditional and HIV-related risk factors on non-AIDS-defining cancer, myocardial infarction, and end-stage liver and renal diseases in adults with HIV in the USA and Canada: a collaboration of cohort studies. The Lancet HIV, 6(2), e93–e104. doi: 10.1016/s2352-3018(18)30295-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ammouri AA, & Neuberger G. (2008). The perception of risk of heart disease scale: Development and psychometric analysis. J Nurs Meas, 16(2), 83–97. [DOI] [PubMed] [Google Scholar]
- Bull ER, McCleary N, Li X, Dombrowski SU, Dusseldorp E, & Johnston M. (2018). Interventions to Promote Healthy Eating, Physical Activity and Smoking in Low-Income Groups: a Systematic Review with Meta-Analysis of Behavior Change Techniques and Delivery/Context. Int J Behav Med, 25(6), 605–616. doi: 10.1007/s12529-018-9734-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Centers for Disease Control and Prevention. (2015). Framingham Study Heart Age Calculator. Retrieved January 2016 [Google Scholar]
- Cioe PA, Guthrie KM, Freiberg MS, Williams DM, & Kahler CW (2018). Cardiovascular Risk Reduction in Persons Living With HIV: Treatment Development, Feasibility, and Preliminary Results. J Assoc Nurses AIDS Care, 29(2), 163–177. doi: 10.1016/j.jana.2017.11.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cohen J. (1988). Statistical power analysis for the behavioral sciences. (Second Edition. ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. . [Google Scholar]
- Eckel RH, Jakicic JM, Ard J, Hubbard VS, de Jesus JM, Lee IM, . . . Y SZ. (2013). AHA/ACC guideline on lifestyle management to reduce cardiovascular risk: a report of the American College of Cardiology American/Heart Association Task Force on Practice Guidelines. Circulation, 00, 000–000. doi: 10.1161/01.cir.0000437740.48606.d1 [DOI] [Google Scholar]
- Feinstein MJ, Bahiru E, Achenbach C, Longenecker CT, Hsue P, So-Armah K, . . . Lloyd-Jones DM. (2016). Patterns of Cardiovascular Mortality for HIV-Infected Adults in the United States: 1999 to 2013. Am J Cardiol, 117(2), 214–220. doi: 10.1016/j.amjcard.2015.10.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Feinstein MJ, Hsue PY, Benjamin LA, Bloomfield GS, Currier JS, Freiberg MS, . . . Post WS (2019). Characteristics, Prevention, and Management of Cardiovascular Disease in People Living With HIV: A Scientific Statement From the American Heart Association. Circulation, 140(2), e98–e124. doi: 10.1161/CIR.0000000000000695 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Forde C, Loy A, O’Dea S, Mulcahy F, Gormley J, & Daly C. (2018). Physical Activity is Associated with Metabolic Health in Men Living with HIV. AIDS Behav, 22(6), 1965–1971. doi: 10.1007/s10461-017-1715-8 [DOI] [PubMed] [Google Scholar]
- Gans KM, Risica PM, Wylie-Rosett J, Ross EM, Strolla LO, McMurray J, & Eaton CB (2006). Development and evaluation of the nutrition component of the Rapid Eating and Activity Assessment for Patients (REAP): a new tool for primary care providers. J Nutr Educ Behav, 38(5), 286–292. doi: 10.1016/j.jneb.2005.12.002 [DOI] [PubMed] [Google Scholar]
- Goff D Jr., Lloyd-Jones D, Bennett G, Coady S, D’Agostino R Sr., & al., e. (2014). 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol, 63(25 Pt B), 2935–2959. doi: 10.1016/j.jacc.2013.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hatleberg CI, Lundgren JD, & Ryom L. (2017). Are we successfully managing cardiovascular disease in people living with HIV? Curr Opin HIV AIDS. doi: 10.1097/COH.0000000000000417 [DOI] [PubMed] [Google Scholar]
- Kaplan RC, Kingsley LA, Sharrett AR, Li X, Lazar J, Tien PC, . . . Gange SJ (2007). Ten-year predicted coronary heart disease risk in HIV-infected men and women. Clinical Infectious Diseases, 45(8), 1074–1081. doi: 10.1086/521935 [DOI] [PubMed] [Google Scholar]
- Larsen DL, Attkisson CC, Hargreaves WA, & Nguyen TD (1979). Assessment of client/patient satisfaction: development of a general scale. Eval Program Plann, 2(3), 197–207. [DOI] [PubMed] [Google Scholar]
- Mabweazara SZ, Leach LL, Ley C, & Smith M. (2018). A six week contextualised physical activity intervention for women living with HIV and AIDS of low socioeconomic status: a pilot study. AIDS Care, 1–5. doi: 10.1080/09540121.2018.1470309 [DOI] [PubMed] [Google Scholar]
- Mdodo R, Frazier EL, Dube SR, Mattson CL, Sutton MY, Brooks JT, & Skarbinski J. (2015). Cigarette smoking prevalence among adults with HIV compared with the general adult population in the United States: cross-sectional surveys. Annals of Internal Medicine, 162(5), 335–344. doi: 10.7326/M14-0954 [DOI] [PubMed] [Google Scholar]
- Miller W, & Rollnick S. (2002). Motivational interviewing: Preparing people for change. (2nd edition ed.). New York, NY: Guilford Press. [Google Scholar]
- National Heart Lung and Blood Institute. (2017). BMI Calculator. Retrieved from nhlbi.nih.gov
- O’Brien KK, Tynan AM, Nixon SA, & Glazier RH (2016). Effectiveness of aerobic exercise for adults living with HIV: systematic review and meta-analysis using the Cochrane Collaboration protocol. BMC Infect Dis, 16, 182. doi: 10.1186/s12879-016-1478-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pacek LR, & Cioe PA (2015). Tobacco Use, Use Disorders, and Smoking Cessation Interventions in Persons Living With HIV. Curr HIV/AIDS Rep, 12(4), 413–420. doi: 10.1007/s11904-015-0281-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paisible AL, Chang CC, So-Armah KA, Butt AA, Leaf DA, Budoff M, . . . Freiberg MS. (2015). HIV infection, cardiovascular disease risk factor profile, and risk for acute myocardial infarction. Journal of Acquired Immune Deficiency Syndromes, 68(2), 209–216. doi: 10.1097/qai.0000000000000419 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Radloff LS (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. [Google Scholar]
- Rasmussen LD, Helleberg M, May MT, Afzal S, Kronborg G, Larsen CS, . . . Obel N. (2015). Myocardial infarction among Danish HIV-infected individuals: population-attributable fractions associated with smoking. Clin Infect Dis, 60(9), 1415–1423. doi: 10.1093/cid/civ013 [DOI] [PubMed] [Google Scholar]
- Raudenbush SW, Bryk AS, &, & Congdon R. (2013). HLM 7.01 for Windows [Computer software]. : Skokie, IL: Scientific Software International, Inc. [Google Scholar]
- Segal-Isaacson CJ, Wylie-Rosett J, & Gans KM (2004). Validation of a short dietary assessment questionnaire: the Rapid Eating and Activity Assessment for Participants short version (REAP-S). Diabetes Educ, 30(5), 774, 776, 778 passim. doi: 10.1177/014572170403000512 [DOI] [PubMed] [Google Scholar]
- Stout RL, Wirtz PW, Carbonari JP, & Del Boca FK (1994). Ensuring balanced distribution of prognostic factors in treatment outcome research. J Stud Alcohol Suppl, 12, 70–75. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/7723001 [DOI] [PubMed] [Google Scholar]
- Sullivan GM, & Feinn R. (2012). Using Effect Size-or Why the P Value Is Not Enough. J Grad Med Educ, 4(3), 279–282. doi: 10.4300/JGME-D-12-00156.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vancampfort D, Mugisha J, Richards J, De Hert M, Lazzarotto AR, Schuch FB, . . . Stubbs B. (2017). Dropout from physical activity interventions in people living with HIV: a systematic review and meta-analysis. AIDS Care, 29(5), 636–643. doi: 10.1080/09540121.2016.1248347 [DOI] [PubMed] [Google Scholar]
- Vancampfort D, Mugisha J, Richards J, De Hert M, Probst M, & Stubbs B. (2017). Physical activity correlates in people living with HIV/AIDS: a systematic review of 45 studies. Disabil Rehabil, 1–12. [DOI] [PubMed] [Google Scholar]
- Webel A, Horvat Davey C, Schexnayder J, Currie J, Al Battashi H, Chang J, & Longenecker CT (2020). Impact of Perceived Cardiovascular Risk on Cardiovascular Disease Prevention Behaviors in People With and Without HIV Infection. J Acquir Immune Defic Syndr, 83(5), 513–521. doi: 10.1097/qai.0000000000002290 [DOI] [PubMed] [Google Scholar]
- Webel AR, Moore SM, Longenecker CT, Currie J, Horvat Davey C, Perazzo J, . . . Josephson RA. (2018). Randomized Controlled Trial of the SystemCHANGE Intervention on Behaviors Related to Cardiovascular Risk in HIV+ Adults. J Acquir Immune Defic Syndr, 78(1), 23–33. doi: 10.1097/qai.0000000000001635 [DOI] [PMC free article] [PubMed] [Google Scholar]