Association of Life’s Simple 7 with reduced clinically manifest abdominal aortic aneurysm: the ARIC Study

Abstract

To optimize cardiovascular health the American Heart Association (AHA) has recommended “Life’s Simple 7 (LS7)”. We tested the hypothesis that greater adherence to the LS7 cardiovascular risk metric is associated with reduced risk of developing abdominal aortic aneurysm (AAA). 14,375 black and white participants aged 45 to 64 years at the baseline visit of the Atherosclerosis Risk in Communities (ARIC) Study cohort were included in this analysis. A 14-point summary score for LS7 was calculated, and participants were classified as having poor (0–4), average (5–9), or ideal (10–14) cardiovascular health. We also counted the number of ideal components. Poisson regression was used to calculate incidence rates for AAA, and Cox regression to calculate hazard ratios adjusted for age, race, sex, and socioeconomic status. Over 25 years of follow-up, we identified 545 clinically manifest AAA events. Incident rates per 1000 person-years declined markedly across LS7 categories: 3.4 for the “poor” category, 2.2 for “average”, and 0.9 for “ideal”. Compared to individuals in the “poor” LS7 category, individuals in the “average” category had 52% lower AAA risk (95% CI: 37% to 63%) and those in the “ideal” category had 80% lower risk (95% CI: 72% to 86%). For every additional ideal component, there was a 28% lower risk of AAA (95% CI: 23% to 33%). Greater adherence to AHA’s LS7 cardiovascular risk metric is associated with a reduced risk of clinically manifest AAA. These findings support the recommendation to follow LS7 for primary prevention of AAA.

Introduction

Abdominal aortic aneurysm (AAA) is a permanent, progressive, potentially fatal, localized dilation of the abdominal aorta. It is usually asymptomatic and is discovered incidentally or following rupture.¹ As many as 1 in 3 AAAs eventually rupture if left untreated.² The mortality following emergency surgery for individuals presenting with ruptured AAAs ranges from 30 – 60%.^3–6

Screening provides a means for early identification of AAA cases and thus allows for institution of treatment aimed at slowing progression of AAA and prevention of possible rupture. The US Preventive Services Task Force (USPSTF) recommends one-time screening for AAA in men aged 65 to 75 who have ever smoked.⁷ While there are significant benefits to screening for AAA, there are also limitations and harms associated with screening.^1,7 Firstly, screening has no impact on reducing AAA incidence. Secondly, it often triggers surgical repair, which can be beneficial but carries some attendant morbidity and mortality.⁷ Lastly, monitoring progression of AAA may cause psychological harm and reduce quality of life.^1,8

The only strategy to reduce incidence of AAA is primary prevention through risk factor identification and control. Since 2010, the American Heart Association (AHA) has advocated following “Life’s Simple 7(LS7)” for general cardiovascular health. The LS7 cardiovascular risk metric provides ideal, intermediate, and poor criteria for seven risk factors or behaviors, namely smoking, body mass index, physical activity, diet, total cholesterol, blood pressure, and fasting serum glucose. Better adherence to ideal LS7 is associated with reduced incidence of cardiovascular disease and cancer.^9,10 However, the association between LS7 and AAA, specifically, is currently unknown.

Research has established several risk factors for AAA.^11–16 For example, in the Atherosclerosis Risk in Communities (ARIC) Study, Tang et al. documented that greater age, male sex, white race, current and former smoking, pack-years of smoking, hypertension, peripheral artery disease (PAD), greater height, greater total or low-density lipoprotein (LDL) cholesterol, and lower high-density lipoprotein cholesterol (HDL-C) were risk factors for AAA.¹⁷ Yet to our knowledge, no study has specifically examined the relation between adherence to LS7 and incident AAA. Understanding the relation between LS7 and clinically manifest AAA might help in simplifying the steps taken to prevent AAA and provide an additional means for individuals to monitor their risk of AAA. We therefore tested the hypothesis that greater adherence to LS7 is associated inversely with AAA incidence in a prospective population-based cohort.

Methods

Study population

The ARIC Study is a prospective community-based cohort study to identify risk factors for atherosclerosis and cardiovascular diseases. A total of 15,792 men and women aged 45 to 64 years were recruited between 1987 and 1989 from 4 communities in the United States (Washington County, MD; the northwest suburbs of Minneapolis, MN; Jackson, MS; and Forsyth County, NC).¹⁸ Participants subsequently were examined in 1990–92, 1993–1995, 1996–98, 2011–13, and 2016–17. Participants are also followed up by annual or semiannual telephone interviews. ARIC was approved by the institutional review board of each participating center, and all participants provided written informed consent for each study visit.

Of the 15,792 participants recruited at baseline, for this report, we applied the following exclusions: race other than white or black (n = 48), prior AAA surgery or angioplasty (n = 11), uncertain AAA status during follow-up (n = 30), missing LS7 data (n = 1,328). 14,375 participants were available for longitudinal analysis after these exclusions.

Measurement of Life’s Simple 7

As described in a previous publication¹⁹ we used the ARIC baseline visit, when participants were middle-aged, as the timepoint for assessment and classification of LS7 characteristics. Measurements included a food frequency questionnaire, physical activity, body mass index, smoking, plasma total cholesterol, seated blood pressure after a 5-minute rest, and fasting blood glucose.¹⁸ We classified each factor according to AHA’s LS7 categories of Ideal, Intermediate, or Poor.^9,20

AAA ascertainment

ARIC identified incident AAAs by several strategies.^17,21 In the annual telephone calls with ARIC participants, interviewers asked about any interim hospitalizations or identified deaths, and records for these outcomes were sought. ARIC also conducted surveillance of local hospitals to identify additional possible AAA hospitalizations or deaths. In addition, ARIC linked participant identifiers with fee-for-service Medicare data from the Centers for Medicare and Medicaid Services for the time-period from 1991 to 2011 to find any missing hospital or outpatient AAA events for those >65 years of age. We identified clinical AAAs as those with a hospital discharge diagnosis from any source or 2 Medicare outpatient claims that occurred at least 1 week apart, with International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 441.3 (ruptured AAA), 441.4 (AAA without mention of rupture), procedure code 38.44 (AAA resection and replacement) or 39.71 (AAA endovascular repair); or a listed cause of death coded as ICD-9 441.3 or 441.4 or International Classification of Diseases, 10th Revision code I71.3 (ruptured AAA), or I71.4 (AAA without mention of rupture). Although clinically detected, these AAAs include both symptomatic and asymptomatic AAAs that were medically documented. Because of our focus on AAA, we treated thoracic, thoracoabdominal, or unspecified aortic aneurysms as nonevents.

Statistical analysis

Baseline LS7 distributions were described for those who did or did not develop incident AAA. Categorical variables were presented as percentages. Continuous variables were presented as means and standard deviations (SD).

Using the LS7 components we created 2 variables for analysis: 1) The number of ideal components of LS7 met by each participant and 2) a score in which each component was given points of 0, 1, or 2 which represents poor, intermediate, or ideal health categories, respectively.²² The sum of these scores in 2) across the 7 components was categorized as poor (0–4), average (5–9), and ideal (10–14) for cardiovascular health. We performed multivariable Cox regression to examine the relation of the two LS7 variables with time to incident AAA. Our main models were adjusted for age (years), sex (male, female), race (black, white), and socioeconomic status using baseline data on maximum educational level (<high school, high school graduate, >high school, and unknown) and baseline annual income (<$12000, $12000–24999, $25000–49999, $50000+, and unknown). Poisson regression was used to calculate the crude incident rates of incident AAA using these same LS7 variables. We also generated Kaplan-Meier failure plots, using STATA, version 12 (Stata, College Station, TX). We repeated these analyses after excluding smoking from our LS7 summary score. Statistical significance was based on a 2-sided type 1 error p-value of 0.05. The main statistical analyses were performed using SAS 9.4 (SAS institute Inc., Cary, NC).

Results

Among the 14,375 ARIC participants free of AAA at baseline, the mean (SD) age was 54.2 (5.8) years, 24.7% were black, and 55.3% were women. Over a median follow-up time of 22.6 years, from 1987–89 through 2011, there were a total of 545 incident AAA events. As shown in Table 1, participants who developed AAA were at baseline more likely to be older, male, and white, compared to participants who did not develop AAA. Those who developed AAA had a higher prevalence at baseline of “poor” blood pressure, diet, smoking, and total cholesterol, but a lower prevalence of poor BMI and glucose.

Table 1.

Demographic characteristics and LS7 components at baseline, 1987–89, between participants who did and did not develop incident AAA, 1987–89 through 2011, ARIC.

Demographics and LS7 Components in 1987–89a	No AAA through 2011	Developed AAA through 2011
N	13,830	545
Age (years, mean and SD)	54.1 (5.8)	56.8 (5.1)
Sex (% female)	56.4	28.1
Race (% black)	25.2	13.6
Body mass index (%)
Poor	27.1	22.9
Intermediate	39.4	45.7
Ideal	33.5	31.4
Blood pressure (%)
Poor	27.2	32.7
Intermediate	30.6	33.9
Ideal	42.2	33.4
Blood glucose (%)
Poor	9.3	6.1
Intermediate	38.6	46.6
Ideal	52.1	47.3
Healthy diet (%)
Poor	30.7	35.4
Intermediate	63.9	60.9
Ideal	5.4	3.7
Physical activity (%)
Poor	37.0	37.3
Intermediate	24.8	20.9
Ideal	38.1	41.8
Smoking status (%)
Poor	24.6	49.5
Intermediate	2.8	5.7
Ideal	72.5	44.8
Total cholesterol (%)
Poor	25.7	33.6
Intermediate	37.2	38.2
Ideal	37.1	28.3
Education
<High school	22.5	27.0
High school graduate	32.7	32.8
>High school	44.7	40.0
Unknown	0.1	0.2
Annual Income
<$12,000	13.7	9.36
$12000–24999	21.1	19.3
$25000–49999	35.4	43.5
$50000+	24.0	23.5
Unknown	5.8	4.4

^aDefinitions per Lloyd-Jones DM, Hong Y, Labarthe D, et al.^9,20

AAA, abdominal aortic aneurysm; ARIC, Atherosclerosis Risk in Communities; LS7, Life’s Simple 7; N, number; SD, standard deviation.

As shown in Table 2, the crude incident rate for incident AAA per 1000 person-years was 0.9 for the ideal LS7 summary category, 2.2 for the average category, and 3.4 for the poor category. Kaplan-Meier failure curves (Figure 1) also showed that the LS7 categories were inversely and strongly associated with the probability of incident AAA during follow-up (log-rank P-value <.0001). HRs (95% CIs) adjusted for age, race and sex (Table 2) were 0.20 (0.14, 0.28) for the ideal LS7 category and 0.48 (0.37, 0.63) for the average category, compared with the poor LS7 category as the reference (P-trend <0.0001).

Table 2.

Crude incidence rates and adjusted HRs with 95% CIs of clinically manifest AAA in relation to baseline LS7 categories, ARIC, 1987–89 through 2011.

AAA	LS7 Categories in 1987–89			P-trend
	Poor (0–4)	Average (5–9)	Ideal (10–14)
Events, N	63	409	73
Person-years	18 484	182 956	85 164
Incidence ratea	3.4	2.2	0.9
HRb (95% CI)	Ref.	0.48 (0.37, 0.63)	0.20 (0.14, 0.28)	<0.0001

^aCrude incidence per 1000 person-years.

^bAdjusted for race (white, black), sex (female, male), age (years), baseline educational level (<high school, high school graduate, >high school, and unknown), and baseline annual income (<$12000, $12000–24999,$25000–49999,$50000+, and unknown).

AAA, abdominal aortic aneurysm; ARIC, Atherosclerosis Risk in Communities; CI, confidence interval; HR, hazard ratio;LS7, Life’s Simple 7; N, number.

Figure 1.

Kaplan-Meier failure curves for clinically manifest AAA, by baseline LS7 categories, ARIC, 1987–89 through 2011. AAA, abdominal aortic aneurysm; ARIC, Atherosclerosis Risk in Communities; LS7, Life’s Simple 7.

As Figure 2 shows, the inverse association between LS7 categories and incident AAA differed modestly by sex (P-interaction = 0.03) but not by race (P-interaction = 0.47). Compared to women in the poor category, women in the ideal LS7 category had 90% lower risk of AAA (95% CI: 80% to 95%). In men, risk was 76% (63% to 84%) lower in the ideal versus poor category.

Figure 2.

Sex and race-specific HRs for clinically manifest AAA in relation to LS7 categories at baseline, ARIC, 1987–89 through 2011. AAA, abdominal aortic aneurysm; ARIC, Atherosclerosis Risk in Communities; HR, hazard ratio; LS7, Life’s Simple 7.

As shown in Table 3, the number of ideal LS7 components was also inversely associated with incidence of AAA, ranging from 4.6 events per 1000 person-years in individuals with 0 ideal components to 0.5 events per 1000 person-years in individuals with ≥6 ideal components. Only 11.7% of the cohort had >5 ideal LS7 components, and they had approximately one-tenth the risk of AAA of those having no ideal components. Modeled as a continuous variable, for every additional ideal LS7 component attained, there was a 28% lower risk of incident AAA (95% CI: 23% to 33%). The association between our LS7 variables and AAA after excluding smoking from the summary scores (Supplemental Tables 1 and 2) were similar to the results presented in Tables 2 and 3. We also report the adjusted HRs for AAA for the individual components of LS7 (Supplemental Tables 3 and 4). Although, as expected, not smoking seemed the biggest LS7 contributor, better blood pressure and total cholesterol, and to a lesser degree healthy diet, and physical activity, also contributed to lower AAA rates.

Table 3.

Crude incidence rate and adjusted HRs with 95% CIs of clinically manifest AAA in relation to number of ideal LS7 components at baseline, ARIC, 1987–89 through 2011.

	Participants at Risk
Number of Ideal LS7 components	N	%	Person-years	AAA events, N	Incidence ratea	HR (95% CI)b
0	390	2.7	6,499	30	4.6	Ref.
1	2236	15.6	41,162	125	3.0	0.59 (0.40, 0.88)
2	3707	25.8	71,888	162	2.3	0.39 (0.27,0.58)
3	3796	26.4	76,700	131	1.7	0.30 (0.20, 0.44)
4	2565	17.8	53,705	74	1.4	0.26 (0.17, 0.41)
5	1279	8.9	27,828	19	0.7	0.14 (0.08,0.26)
6–7	402	2.8	8,823	4	0.5	0.10 (0.04, 0.30)

^aCrude incidence per 1000 person-years.

^bAdjusted for race (white, black), sex (female, male), age (years), baseline educational level (<high school, high school graduate, >high school, and unknown), and baseline annual income (<$12000, $12000–24999, $25000–49999, $50000+, and unknown).

AAA, abdominal aortic aneurysm; ARIC, Atherosclerosis Risk in Communities; CI, confidence interval; HR, hazard ratio;LS7, Life’s Simple 7; N, number.

Discussion

In this large population-based longitudinal study, over a median follow-up time of 22.6 years, we found that greater adherence to AHA’s LS7 metrics was associated with lower risk of incident AAA in middle-aged black and white adults. The inverse association of LS7 components with incident AAA was strong and was independent of age, race, and sex, though stronger in women than in men. These findings suggest that adherence to LS7 recommendations in midlife can help in the primary prevention of incident AAA.

While screening high-risk individuals is an effective way for early identification of AAA and allows for early intervention, it is associated with a high financial cost, psychological harm, and it misses potentially fatal cases in individuals not labelled as high-risk by the current screening guidelines. Over the years some pharmaceutical options have been proposed as means to delay the progression of AAA and reduce the incidence of AAA rupture. Studies done to assess the benefits of these medications have shown: no benefit for the use doxycycline in reducing the rate of AAA progression;²³ inconsistent results for the use of angiotensin converting enzyme inhibitors to delay AAA growth;^24,25 and a reduction in the risk of rupture for participants on statins compared to participants who have never used statins.^26,27 Lifestyle advice for the management of AAA remains largely unchanged and has focused majorly on smoking cessation, blood pressure control, and the treatment of dyslipidemia.³ Screening, statin use, and lifestyle interventions are targeted towards preventing progression and rupture of AAA and not preventing incident AAA in the first place.³

AHA’s LS7 guidelines aim to reduce the onset of elevated cardiovascular risk factors population-wide and thereby reduce the occurrence of major cardiovascular events and attendant morbidity and mortality. This population approach to primary prevention provides a potentially efficient and effective means to reduce the incidence of AAA and subsequent morbidity and mortality. Our study adds to the evidence showing that adherence to LS7 reduces coronary disease, stroke, heart failure,¹⁹ venous thromboembolism,^22,28 cancer,²⁹ and other non-cardiovascular diseases²⁹ by demonstrating that greater adherence to LS7 in middle-age is associated with reduced incidence of AAA. Our representation of LS7 categories is similar to previous studies on the association of LS7 and various cardiovascular diseases^19,22,28 and allows for comparability of findings.

Of course, many studies have shown associations between components of the LS7 metric and incident AAA. These include positive associations of AAA with smoking,^11,17 obesity,^12,13 total cholesterol,¹⁷ blood pressure,³⁰ and diets low in fruits,³¹ low in both fruits and vegetables,³² and low in energy.³³ In contrast, associations between physical activity and AAA have been inconclusive, and an inverse relationship between fasting serum glucose and AAA has been consistently reported.^14–16 We were unable to find any study that examined the relationship of all the LS7 components, taken together, and incident AAA.

Limitations

One limitation of this study is that we used a single measure of LS7 components, precluding evaluation of AAA risk in relation to changes in these components or the overall score. Secondly, although the ARIC sample was community-based, our results generalize only to whites and blacks. Thirdly, we did not have information on other important risk factors such as family history of AAA.^34,35 Fourthly, although we excluded participants with prior AAA surgery at ARIC baseline, we did not screen the whole ARIC cohort to identify and exclude the relatively few prevalent AAAs we might have missed in the 45–64 year old volunteers. Lastly, we used ICD codes to ascertain the diagnosis of AAA from hospital records and death certificates. The validity of ICD codes is likely imperfect, and we certainly would have missed undiagnosed AAAs and AAAs that did not present to the hospital, although it seems unlikely that this should greatly affect the conclusions of our study.

Conclusion

Greater adherence to LS7 is associated with a reduced incidence of AAA. These findings provide evidence that adherence to LS7 may be useful for the primary prevention of AAA.