Hepatocyte growth factor (HGF) is a mesenchyme-derived cytokine released in response to endothelial injury and may be elevated due to insufficient compensatory mechanisms (1–4). However, HGF has cardioprotective effects in cardiac tissue through the activation of several biological pathways such as angiogenesis, anti-apoptosis, anti-inflammation and anti-fibrosis (1–4). Several studies have provided evidence that HGF is an independent predictor of coronary heart disease, heart failure, stroke and progression of atherosclerosis (5–8). Likewise, higher serum levels of HGF have been found in individuals with cardiovascular disease (CVD) risk factors e.g., smoking, obesity, hypertension and diabetes (9–13).

The American Heart Association (AHA) introduced the concept of “ideal cardiovascular health (CVH)” defined as meeting specific criteria for seven metrics. These modifiable health behaviors and factors include smoking, physical activity, body mass index (BMI), diet, blood pressure, total cholesterol and fasting plasma glucose (14, 15). The goal of the concept is to use the CVH metrics to measure and monitor CVH along with the incidence, mortality and outcomes of CVD globally (14–17). With prior studies showing that favorable CVH is associated with better endothelial function (18), lower risk of incident CVD (19) and lower levels of CVD biomarkers such as GlycA, homocysteine and cardiac troponin (20, 21), it is expected that individuals with favorable CVH would also be at lower risk of endothelial injury.

However, as little is known about the association between favorable CVH and HGF, our study aimed to analyze data from the Multi-Ethnic Study of Atherosclerosis (MESA) to examine whether participants with favorable CVH have lower levels of HGF and if the association is modified by age, sex or race/ethnicity.

Most of the baseline characteristics of study participants (N = 6,490) varied by the categories of the CVH score and HGF tertiles as reported in Table 1 and in the Supplementary Table S2. The mean age (SD) of participants was 62 (10) years and 53% were female. As illustrated in Figure 1, participants with inadequate CVH scores had higher HGF levels compared to those with optimal CVH scores. Additionally, the mean (SD) of CVH scores declined across tertiles of HGF [9.4 (2.0), 8.6 (2.1) and 7.8 (2.2) for tertiles 1–3, respectively] as shown in the Supplementary Table S2.

Table 2 shows the associations between CVH and HGF. A one-unit increment in the continuous measure of the CVH score was significantly associated with 3% lower HGF levels after adjusting for sociodemographic factors. For the categorical measure of the CVH score, the average and optimal scores were significantly associated with 8% and 12% lower HGF levels, respectively, compared to inadequate scores. In addition, a greater number of ideal metrics was associated with lower HGF levels. For example, study participants with 5 and 6–7 ideal metrics had 18% and 23% lower HGF levels, respectively. These associations were slightly attenuated when compared to the unadjusted model. The tests for effect modification by age, sex and race/ethnicity were not statistically significant (p > 0.05).

Table 3 shows the associations between the individual CVH metrics and HGF. Participants in the ideal category for smoking had 11% lower HGF levels in the adjusted model compared to those in the poor category of smoking which was a stronger association in comparison to the unadjusted model. Although slightly attenuated, ideal BMI and physical activity were associated with an adjusted 12% and 4% lower HGF levels, respectively. Participants who reported eating an ideal diet had 7% lower HGF levels with no appreciable change from the unadjusted model. However, those who met the ideal criteria for total cholesterol had an adjusted 3% higher HGF levels. Ideal blood pressure and blood glucose were associated with 7% and 11% lower HGF levels, respectively, although the estimates were slightly attenuated in comparison to the estimates from the unadjusted model.

In this cross-sectional analysis of 6,490 adults free of clinical CVD at baseline, we found that participants with higher CVH scores or greater number of ideal metrics had lower levels of HGF. Similarly, we found that participants who met the ideal criteria for the individual CVH metrics had lower levels of HGF except for total cholesterol where ideal levels were associated with higher levels of HGF.

Prior studies in the MESA cohort have examined the association between HGF and CVD. One of these studies found that the risk of incident stroke was 17% greater with every standard deviation increase in circulating HGF independent of sociodemographic and CVD risk factors (7). Another study from the same cohort found that every standard deviation increase in circulating HGF was associated with 12% greater odds of having any coronary artery calcium and 20% increased risk of coronary heart disease after adjusting for traditional CVD risk factors (30). Furthermore, in another analysis from the MESA cohort, the risk of heart failure and heart failure with preserved ejection fraction was 59% and 90% higher, respectively, for participants with HGF levels in the 3rd tertile compared to those in the 1st tertile, independent of sociodemographic and CVD risk factors (31). These studies support our findings of an association between favorable CVH and lower levels of HGF since HGF is elevated in CVD as a response to endothelial damage (7). Factors that lead to HGF release may mediate the association between poorer CVH and incident CVD events.

In addition, our results showing that ideal CVH metrics (excluding total cholesterol) were associated with lower levels of HGF are supported by the findings from prior research. For example, Chen et al. (32) demonstrated that a history of current or previous smoking increases HGF expression in non-tumor lung tissue. The authors postulated that smoking increases HGF levels by activating HGF gene expression, however, the mechanism by which this process occurs is yet to be determined (32). Rehman et al. (10) showed that obesity was associated with more than three times increase in HGF levels and this finding was correlated linearly with BMI. The presence of excess adipose tissue may explain the association between poor BMI and elevated HGF (10). The explanation for the association between ideal physical activity and lower HGF levels is still under investigation.

In contrast to the inverse association for the other CVH metrics, we found that ideal cholesterol was paradoxically associated with higher HGF. This finding has also been observed in another study by Hiratsuka et al. (13) where participants with higher HGF levels in the fourth quartile had lower total cholesterol levels compared to those in the first quartile. However, the biological pathway responsible for the association between higher HGF levels and lower cholesterol levels is still unknown. Nakamura et al. (11) reported HGF levels in participants with hypertension was significantly higher than in participants without hypertension. In addition, Bancks et al. (12) observed that higher levels of circulating HGF was significantly associated with incident type 2 diabetes in the MESA cohort. These findings are not surprising given that hypertension and diabetes are associated with endothelial damage and dysfunction which may lead to elevated HGF levels (11, 12).

Our study revealed that poorer CVH was associated with higher HGF levels, therefore measuring HGF levels may be of utility in identifying people in poorer CVH because prior research has provided insight into the clinical significance of HGF as a risk marker for subclinical and clinical CVD (7, 8, 30). With the increasing burden and cost of managing CVD, effective interventions that promote favorable CVH are needed (33). These public health intervention programs should encourage lifestyle modification which will reduce the risk of developing CVD through several potential mechanisms including the reduction of endothelial injury as indicated by lower HGF levels (33).

In this multicenter cross-sectional study of adults free of CVD at baseline, we found that favorable CVH was associated with lower levels of HGF. Interventions aimed at promoting and preserving favorable CVH may reduce endothelial injury and decrease CVD risk as indicated by lower serum HGF levels. However, there is need for prospective studies to examine the relationship between CVH and HGF to determine whether HGF levels can be modified by improvements in the CVH metrics. More research is needed to provide further understanding of the underlying mechanisms driving the associations.

The datasets presented in this article are not readily available because there are restrictions on sharing the actual MESA database because it has patient identifiers and protected health information and thus is subject to IRB approval. We have signed a data use agreement (DUA) with the MESA Coordinating Center that we cannot directly share data and that anyone with access to the data has to follow MESA policies and procedures for data use. We are not allowed to personally share the datasets and have to keep them stored securely in a Cloud using encrypted software for protection. However, other researchers may obtain the minimal dataset required to replicate our study's findings by submitting a proposal through the NIH BioLINCC (https://biolincc.nhlbi.nih.gov/studies/mesa/). Additionally, one can apply to the MESA Coordinating Center to become a new investigator after signing a DUA. The findings of this study should be easily reproducible through the methods described in this paper. Requests to access the datasets should be directed to Craig Johnson (wcraigj@uw.edu) at the Coordinating Center or at https://www.mesa-nhlbi.org/.

The MESA study protocol was approved by the Institutional Review Boards (IRB) at each recruitment center and informed consent was given by all participants. At Johns Hopkins University, this study was approved by the Johns Hopkins School of Medicine IRB applicant number NA_00030361. The participants provided their written informed consent to participate in this study.

OOs, OOg, and EM: study conception and design and drafting of manuscript. OOs and OOg: acquisition of data and analysis. OOs, OOg, RF, CN, GB, NL, SB, and EM: interpretation of data and critical revision and approval of final version submitted. All authors contributed to the article and approved the submitted version.

The MESA study was supported by contracts HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, and N01-HC-95169 from the National Heart, Lung and Blood Institute (NHLBI) and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences. HGF measurement was funded by R01HL98077. EM was additionally funded by the Amato Fund for Women's Cardiovascular Health at Johns Hopkins University. The funding sources had no role in the study design; collection, analysis and interpretation of data; writing the report; and the decision to submit the report for publication. The specific roles of the authors are articulated in the author contributions section.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.