Athletic training is known to induce morphological and functional cardiovascular adaptations of cardiac chambers, collectively known as athlete's heart (1). Apart from the widely established benefits of exercise, recent evidence suggests there may be some overlap between physiology and pathophysiology, and occasionally, with cardiac diseases. The duration of exposure to demanding training and the type of sports may play a role in these processes (2). Cardiac magnetic resonance (CMR) has been increasingly used in establishing an accurate diagnosis given that exercise may lead to cardiac remodeling that in certain situations can be clinically challenging to differentiate from various cardiomyopathies (3, 4).

CMR is a non-radiating imaging modality with high spatial resolution, which not only is the reference standard for functional and morphological assessment but also has the benefit of tissue characterization by exploiting gadolinium-based contrast late enhancement (LGE) as a marker of myocardial fibrosis (MF) (5, 6). Gadolinium-based contrast compounds freely enter the extracellular space but not intact myocardial cells. Under pathologic conditions, cellular death and fibrotic processes lead to expansion of extracellular space, while cell damage means that myocardial cell membranes become permeable to contrast. These phenomena significantly alter the kinetics of gadolinium, leading to higher peak uptake from the myocardium and delayed washout (7). Interestingly specific patterns of LGE have been sporadically detected in athletic individuals, although data so far are not consistent, coming from small-sample studies that frequently lack comparisons with sedentary controls or “lifelong,” veteran athletic individuals which could provide further insight in the underlying pathophysiological mechanisms (8).

Other important techniques, such as native T1 and extracellular volume (ECV) may prove clinically relevant in athletic individuals providing additional information regarding cellular and extracellular pathology, even though they have not yet been widely studied in these populations (9). Native T1 times quantify the time required for the net magnetization vector of a myocardial area to return to pre-excitation levels. Native T1 depends highly on the tissue composition (10). Elevated T1 times have been reported in several myocardial pathologic conditions. Pertinent to the topic of this review, T1 values increase in situations of increased free water within expanded interstitial space and could be used to detect and quantify interstitial myocardial fibrosis (11). The measurement of pre- and post-contrast T1 times enable the CMR-based calculation of ECV with good correlation to the actual histology-derived collagen volume fraction (12). The ratio of intra- to extracellular space shifts significantly in myocardial fibrosis, and ECV is uniquely capable of detecting such changes. It thus constitutes an ideal modality for the clinical applications described in the present review (13).

All available evidence indicates that the prevalence of MF, as documented by LGE in CMR scans, is significantly higher in athletes compared to sedentary controls. The pooled frequency of 16.6% is lower than the 21.1% reported by Zhang et al. (16), with the difference owed to the updated set of studies, which include new data, updated numbers from some research teams and exclude some previous, non-peer-reviewed datasets.

In a close examination of the included studies, LGE was indeed considerably more frequent in athletes compared to controls, even when excluding RV insertion point fibrosis. Studies that included younger athletes also had a significantly higher prevalence of LGE compared to studies including veteran athletes. This is an interesting finding that is open to interpretation. It is not inconceivable that the differences in recruitment strategy between studies that led to an age disparity also caused differences in the prevalence of LGE. Importantly, Domenech-Ximenos et al., one of the included studies with the lowest mean athlete age, only administered gadolinium-based contrast to a sub-set of participating patients, potentially introducing bias in LGE results (21). Overall, the long-term effects of endurance exercise on the heart are widely unknown. Endurance exercise is associated with a transient increase of biomarkers of cardiac damage and there is growing evidence that lifelong male athletes aged above 40 years show a higher prevalence of a higher coronary plaque burden, and a different MF pattern compatible with subclinical infarction compared with relatively sedentary healthy controls (32). As no adequate data were available for females, no sex-specific conclusions could be safely drawn from our analysis. Further, ideally prospective studies with sizeable athlete populations are required to determine the effect of duration exercise on the incidence, pattern, and extent of MF as well as its prognostic relevance.

Quite interestingly, native T1 values were consistently shown to be significantly decreased in athletes, both in 1.5T and 3T magnetic fields. It has been suggested that in athletic left ventricular hypertrophy (LVH), native T1 seems to be decreased suggesting that physiological athletic LVH represents enhanced cellular hypertrophy, unlike any other sort of LVH mechanism (6). The ECV as a marker of myocardial tissue remodeling and excessive collagen deposition is also a robust measure of diffuse MF and fairly interesting from a pathophysiological perspective. It would not be unreasonable to assume that, in the setting of exercise-induced cardiomyocyte hypertrophy (6), the ECV is expected to be reduced. That said, the ECV was not found significantly increased in athletes compared to controls, and indeed some studies yielded the opposite result (23, 27). Whether this is the result of study selection bias, relatively small numbers of included so far studies, or whole-heart sub-clinical expansion of ECV through increased collagen deposition in maladapted athletes' hearts (33) is an interesting research question that deserves further investigation.

In conclusion, we report interesting data based on the latest evidence in MF assessment from both contrast based and non-contrast CMR techniques. The present systematic review and meta-analysis contains updated data about LGE prevalence along with a comprehensive risk of bias assessment, in combination with a novel meta-synthesis of data regarding T1 and ECV values in the endurance athletes. Non-specific MF in athletic individuals is a somewhat frequent finding in highly trained athletes and there seem to be variation attributable to age (and therefore potentially the duration of exposure to exercise) and sex. Contrast based non-specific LGE and native T1 are found to be particularly useful in discrimination of athletic vs. sedentary individuals. Further data are certainly required to elucidate the underlying physiological and pathophysiological mechanisms. More importantly, whether these non-specific LGE patterns in athletes are actually associated with adverse events-particularly for underrepresented athlete groups such as women and veteran athletes, as well as the effect of deconditioning on the fibrotic process, are all certainly topics for further research.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

EA and CV conceived and designed the present systematic review. AA and PP also contributed meaningfully to the design of the systematic review and meta-analysis. DM, AT, and CB performed the systematic literature search. DM, AT, CB, and AA performed the analysis of data. EA, DM, AT, and CB wrote the initial draft versions of the manuscript. PP, AA, and CV performed significant revisions. All authors provided critical feedback and helped shape the research, analysis, and manuscript.

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.