Interactions Between Enhanced Polygenic Risk Scores and Lifestyle for Cardiovascular Disease, Diabetes, and Lipid Levels
Circulation: Genomic and Precision Medicine
Abstract
Background:
Both lifestyle and genetic factors confer risk for cardiovascular diseases, type 2 diabetes, and dyslipidemia. However, the interactions between these 2 groups of risk factors were not comprehensively understood due to previous poor estimation of genetic risk. Here we set out to develop enhanced polygenic risk scores (PRS) and systematically investigate multiplicative and additive interactions between PRS and lifestyle for coronary artery disease, atrial fibrillation, type 2 diabetes, total cholesterol, triglyceride, and LDL-cholesterol.
Methods:
Our study included 276 096 unrelated White British participants from the UK Biobank. We investigated several PRS methods (P+T, LDpred, PRS continuous shrinkage, and AnnoPred) and showed that AnnoPred achieved consistently improved prediction accuracy for all 6 diseases/traits. With enhanced PRS and combined lifestyle status categorized by smoking, body mass index, physical activity, and diet, we investigated both multiplicative and additive interactions between PRS and lifestyle using regression models.
Results:
We observed that healthy lifestyle reduced disease incidence by similar multiplicative magnitude across different PRS groups. The absolute risk reduction from lifestyle adherence was, however, significantly greater in individuals with higher PRS. Specifically, for type 2 diabetes, the absolute risk reduction from lifestyle adherence was 12.4% (95% CI, 10.0%–14.9%) in the top 1% PRS versus 2.8% (95% CI, 2.3%–3.3%) in the bottom PRS decile, leading to a ratio of >4.4. We also observed a significant interaction effect between PRS and lifestyle on triglyceride level.
Conclusions:
By leveraging functional annotations, AnnoPred outperforms state-of-the-art methods on quantifying genetic risk through PRS. Our analyses based on enhanced PRS suggest that individuals with high genetic risk may derive similar relative but greater absolute benefit from lifestyle adherence.
Introduction
Poor lifestyle has long been known to confer risk for cardiovascular diseases (CVD),1,2 which are the leading cause of morbidity and mortality in the world.3 To improve cardiovascular health, the American College of Cardiology and the American Heart Association has recommended a healthy lifestyle with guidelines for diet, physical activity, obesity, and tobacco use.4
Previous studies have also demonstrated that genetics plays an important role in CVD risk.5–7 In recent years, large-scale genome-wide association studies (GWAS) have rapidly expanded our knowledge on the genetic variants that are associated with CVD and its key drivers including type 2 diabetes (T2D) and dyslipidemia.8–11 Many methods have been subsequently proposed for aggregating information from GWAS results to quantify the genetic risk for an individual through polygenic risk scores (PRS).12–17
As CVD risk is conferred by both lifestyle and genetic factors, several studies have sought to examine the interactions between these 2 groups of risk factors for CVD traits18–20 and reached the conclusion that genetic factors (summarized through PRS) and lifestyle factors independently contribute to cardiovascular disorders and related diseases, including coronary artery disease (CAD), atrial fibrillation (AF), stroke, hypertension, and T2D.19,21 However, the PRS used in previous studies were constructed either based on GWAS results from a limited number of samples or using less optimal risk prediction methods, providing a less accurate representation of an individual’s genetic risk.12–15,18–21 In addition, previous studies also coarsely stratified the study population by PRS into a limited number of risk groups without considering that the empirical risk of common diseases increased sharply in the extreme tails of the PRS distribution.22–25 As summarized recently by Khera et al,25 larger GWASs and improved statistical methods could derive PRS with better prediction accuracy. Here, we set out to construct more predictive PRS for CAD, AF, and T2D from the largest-to-date GWAS summary statistics and then comprehensively investigate the extent to which the genetic predisposition of CAD, AF, and T2D can alter the effect of lifestyle adherence on disease outcomes, especially for those with extremely high genetic risk. In addition, we also extend the investigation of potential interactions between PRS and lifestyle on blood lipid levels, which are crucial intermediate traits for CVD.6,26
Methods
A full-length description of the methods is available as part of the Data Supplement (Methods, Tables I and II, Figures I and II in the Data Supplement). Because of the sensitive nature of the individual-level data collected for this study, requests to access the data set from qualified researchers trained in human subject confidentiality protocols may be sent to UK Biobank (UKBB) at https://www.ukbiobank.ac.uk/register-apply. The summary-level data (eg, PRS weights) are available from the corresponding author upon request. All analytical methods are available and reported.12–15 The study was approved by the UKBB27 and by the ethic committee of Yale University. All individuals have provided written informed consent.
Results
Population Characteristics
Of the 502 618 participants aged 40 to 69 years in the UKBB, 226 522 were excluded according to the exclusion criteria stated in Methods in the Data Supplement. The remaining 276 096 individuals were divided into a training set of 92 928 individuals and a testing set of 183 168 individuals. In the training set, there were 4746 CAD cases, 3606 AF cases and 4639 T2D cases. A total of 16 719 individuals were excluded from lipid level analysis due to missingness or their taking cholesterol-lowering medication. The testing set for each disease was constructed by further excluding prevalent cases. This yielded 176 238 participants with 3467 incident cases for CAD, 178 651 participants with 4025 incident cases for AF, 178 138 participants with 4659 incident cases for T2D, and 144 939 participants for the lipid testing set (Figure I in the Data Supplement). Baseline characteristics for the population are provided in Table.
| Characteristics | Testing set (N=183 168) |
|---|---|
| Age at recruitment, mean (SD), y | 56.81 (8.00) |
| Number of females (%) | 98 611 (53.84) |
| Years in education, mean (SD), y | 15.08 (5.05) |
| Townsend deprivation index at recruitment, mean (SD) | −1.63 (2.90) |
| Annual household income, No. (%), £ | |
| <18 000 | 32 481 (17.73) |
| 18 000–30 999 | 40 213 (21.95) |
| 31 000–51 999 | 41 952 (22.90) |
| 52 000–100 000 | 34 158 (18.65) |
| >100 000 | 9223 (5.04) |
| Unknown | 25 141 (13.73) |
| Smoking, No. (%) | |
| Ideal (never) | 101 012 (55.14) |
| Intermediate (former) | 65 762 (35.90) |
| Poor (current) | 16 394 (8.95) |
| Body mass index, No. (%) | |
| Mean (SD) | 27.30 (4.71) |
| Ideal (<25 and ≥18.5 kg/m2) | 60 913 (33.26) |
| Intermediate (<30 and ≥25 kg/m2) | 77 966 (42.57) |
| Poor (≥30 kg/m2) | 42 802 (23.37) |
| Exclusion (missing or <18.5 kg/m2) | 1487 (0.81) |
| Physical activity, No. (%) | |
| Ideal (regular physical activity) | 89 309 (48.76) |
| Intermediate (some physical activity) | 53 613 (29.27) |
| Poor (limited physical activity) | 40 246 (21.97) |
| Diet, No. (%) | |
| Ideal (adequate intake of >5 dietary components) | 27 643 (15.09) |
| Poor (inadequate intake of >5 dietary components) | 155 525 (84.91) |
| Combined lifestyles, No. (%) | |
| Ideal (≥3 lifestyle factors in ideal status) | 29 080 (15.88) |
| Intermediate (2 lifestyle factors in ideal status) | 138 410 (75.56) |
| Poor (≤1 lifestyle factors in ideal status) | 15 678 (8.56) |
Prediction Performance of PRS
Four PRS methods were considered in our study: P+T,12 LDPred,13 PRS continuous shrinkage (CS),14 and AnnoPred.15 P+T is also the standard PRS method where the marginal effect sizes from GWAS summary statistics were directly used as weights, and the SNPs were selected after LD-clumping and P value thresholding.12 LDPred and PRS-CS are both Bayesian approaches that model the LD information extracted from a reference panel, where LDPred assumes an independent point-normal prior while PRS-CS assumes a CS prior on the SNP effect sizes.13,14 AnnoPred is a Bayesian framework that further leverages functional annotations in quantifying genetic risk.15
Our empirical results suggest that AnnoPred achieved the best predictive performance for all 6 traits in both the training and testing sets, with the AUC in the testing set being 0.643 (95% CI, 0.637–0.648) for CAD, 0.632 (95% CI, 0.625–0.639) for AF, and 0.645 (95% CI, 0.639–0.651) for T2D, respectively; and R2 being 0.0751 (95% CI, 0.0728–0.0775) for total cholesterol (TC), 0.0744 (95% CI, 0.0720–0.0767) for triglyceride, and 0.0705 (95% CI, 0.0682–0.0729) for LDL-C (low-density lipoprotein cholesterol), respectively (Figure 1, Table III in the Data Supplement). In particular, the optimal PRS involved 2 994 054 variants for CAD, 2 996 792 variants for AF, 2 996 760 variants for T2D, 1 198 743 variants for TC, 1 197 954 variants for triglyceride, and 1 197 834 variants for LDL-C, respectively.
Associations of Disease Risk With PRS and Combined and Individual Lifestyle Factors
As shown in Figure 2, a risk gradient was clearly observed across the 10 genetic risk groups for each overall lifestyle status where individuals with high PRS were at higher risk of CAD, AF, and T2D events than those with low PRS. This trend was especially visible for participants in the right tail of the PRS distribution, where the risk of developing disease increased sharply as PRS increased. For participants with an intermediate combined lifestyle status, the adjusted hazard ratios (HRs) were 4.23 (95% CI, 3.39–5.28) for CAD, 3.69 (95%, 2.96–4.60) for AF, and 3.67 (95%, 2.99–4.52) for T2D, when comparing the group with >99% PRS to the group with 40% to 60% PRS (Table IV and V in the Data Supplement).
Within each of the 10 PRS-defined genetic risk groups, poor lifestyle was consistently associated with increased disease risk when compared with intermediate lifestyle, whereas healthy lifestyle was associated with decreased disease risk. Compared with the group of individuals with mid-range genetic risk (40%–60% PRS) and intermediate combined lifestyle, the adjusted HRs for the group of individuals having the top 1% PRS while leading poor lifestyles increased to 5.23 (95% CI, 3.01–9.09) for CAD, 5.43 (95%, 3.34–8,82) for AF, and 6.67 (95%, 4.55–10.1) for T2D. On the contrary, for individuals with similarly high genetic risk (top 1% PRS) but healthy lifestyle, the corresponding risk decreased to 3.19 (95% CI, 1.64–6.19) for CAD, 2.15 (95%, 1.15–4.02) for AF, and 0.98 (95%, 0.41–2.36) for T2D, demonstrating the benefit of leading a healthy lifestyle even for this extremely high genetic risk group.
Relative Disease Risk Reduction When Leading a Healthy Lifestyle
As the 3 HR curves across the 10 genetic risk bins for the 3 lifestyle groups are almost parallel to each other (Figure 2), the UKBB data suggest that the effects of lifestyles may be independent of PRS for CAD, AF, and T2D at the log HR scale. We then performed formal statistical tests to investigate the multiplicative interactions between PRS and combined lifestyle on disease outcomes under the Cox proportional hazard regression model. Consistent with our impressions from the figures, no significant multiplicative interactions were identified after Bonferroni correction (Tables VI and VII in the Data Supplement).
We further studied the joint effects of genetic risk and individual lifestyle factors as summarized in Figures III through V in the Data Supplement. At the individual lifestyle factor level, each lifestyle factor seems to still exert an effect independent of genetic risk on disease outcomes, albeit the magnitude of relative risk across each lifestyle factor varies among 3 diseases. Smoking and body mass index (BMI) were both important risk factors for CAD and AF, while diet and physical activity did not clearly influence risk. For T2D, BMI still played a prominent role, and the effect from physical activity could also be observed clearly, while no significant effects from smoking and diet were observed.
Absolute Disease Risk Reduction When Leading a Healthy Lifestyle
In contrast to the independent relationship between reduced HR and PRS, absolute risk reduction (ARR) from healthy lifestyle was greater in the group at higher PRS for CAD, AF, and T2D (Figure 3, Table VIII in the Data Supplement). More specifically, after lifestyle modification, in the extremely high PRS group (PRS >99%), AR of CAD reduced from 7.9% to 3.4% (ARR=4.5%); while in the intermediate PRS group (40%< PRS <60%), AR of CAD reduced from 2.8% to 0.9% (ARR=1.9%); and in the low PRS group (PRS <10%), AR of CAD reduced from 1.6% to 0.6% (ARR=1.0%). For AF, after lifestyle modification, the AR reduced from 10.7% to 3.5% (ARR=7.2%) in the extremely high PRS group (PRS >99%), from 3.3% to 1.5% (ARR=1.8%) in the intermediate PRS group (40%< PRS <60%) and from 1.1% to 0.6% (ARR=0.5%) in the low PRS group (PRS <10%). Strikingly, there was an apparent sharp increase of ARR in the extreme right tail of AF PRS distribution, where the ARR was <3% even in group with PRS >90% and PRS <99%, but rose up to 7.2% in the group with PRS >99%. Overall, the AR of T2D was larger than CAD and AF. The ARR from lifestyle adherence for T2D was also larger; after lifestyle modification, the AR reduced from 14.3% to 1.9% (ARR=12.4%) in the extremely high PRS group (PRS >99%), from 5.6% to 0.6% (ARR=5.0%) in the intermediate PRS group (40% <PRS <60%), and from 3.1% to 0.3% (ARR=2.8%) in the low PRS group (PRS <10%).
We also studied the ARR from individual lifestyle adherence within each PRS groups (Figures VI through VIII in the Data Supplement). For CAD, we could observe the same trend that ARR was greater in the group at higher PRS from the modification of smoking behavior, BMI, and physical activity. In the extremely high PRS group (PRS >99%), changing the smoking status from poor to ideal alone could lead to an ARR of 5.2%. For AF, ARR from lifestyle adherence mainly came from the modification of BMI. Especially in the extremely high PRS group (PRS >99%), ARR from BMI modification was as high as 6.7%. For T2D, the ARRs from all of the 4 individual lifestyle components increased with PRS. Specifically, BMI modification alone could lead to an ARR of 11.1% and smoking behavior modification alone could lead to an ARR of 6.3%.
Associations of Lipid Levels With PRS and Combined and Individual Lifestyle Factors
For all 3 blood lipid types, the mean levels increased across PRS quantiles for each combined lifestyle category, while healthy lifestyle was associated with decreased mean levels within each PRS group (Figure 3, Table IX in the Data Supplement).
According to the recommended guidelines set by the National Cholesterol Education Program,28,29 only the group with the lowest genetic risk (<5% PRS) in each lifestyle category had an average TC level at the desirable level (<5.172 mmol/L [<200 mg/dL]). Conversely, groups with high genetic risk (>90% PRS) in all lifestyle categories had an average TC level at high designation (≥6.206 mmol/L [≥240 mg/dL]). For groups with the same PRS grouping but different lifestyle categories, the average TC levels were mostly in the same National Cholesterol Education Program designation, but healthy lifestyle tended to pull the average TC levels to a healthier designation. The protective effect of healthy lifestyle was most pronounced for triglyceride. The normal designation for triglyceride is <1.694 mmol/L [150 mg/dL]. While none of the PRS groups with poor lifestyle had average levels below this threshold, the average triglyceride level was below this threshold for people with intermediate lifestyle and <40% PRS, whereas for people with healthy lifestyle, the average level was below the threshold for all the PRS groups except the one with the highest genetic risk, that is, >99%. As for the high triglyceride designation (≥2.258 mmol/L [≥200 mg/dL]), the average triglyceride level was above this threshold for individuals with >60% PRS in poor lifestyle group versus >99% PRS in the intermediate lifestyle group and none of the PRS groups in the ideal lifestyle group. A similar result was also observed in the LDL-C analysis, with <5% PRS, <10% PRS, and <20% PRS in the poor, intermediate, and ideal lifestyle groups, respectively, were at the optimal or near-optimal levels of LDL-C (<3.362 mmol/L [<130 mg/dL]). Moreover, 5% to 60% PRS, 10% to 60% PRS, and 20% to 80% PRS in the respective groups were at borderline high levels (3.362–4.138 mmol/L [130–160 mg/dL]), and the remaining intervals of PRS had LDL-C levels designated as high (>4.138 mmol/L [>160 mg/dL]).
The results in Figure 4 suggest that lifestyle and genetic factors may exert independent effects on the blood lipid levels, as healthy lifestyle reduced the mean lipid levels similarly regardless of PRS grouping. We subsequently tested statistical interactions between PRS and combined lifestyle categories for all 3 types of lipid (Tables X and XI in the Data Supplement). In the analyses based on the deciles of PRS, we observed no significant interactions after Bonferroni correction, as suggested by Figure 4. However, in the analyses based on continuous PRS, we observed a significant positive interaction between PRS and lifestyle for triglyceride (P=0.0027), suggesting that people at higher PRS could benefit more from lifestyle adherence.
We also studied the joint effects of genetic risk and individual lifestyle factors on lipid levels (Figures IX through XI in the Data Supplement). Individual lifestyle factors still exerted effects on lipid levels independent of genetic risk, with BMI as the prominent lifestyle factor for all 3 types of lipid and smoking playing an important role specifically for triglyceride.
Sex Differences
We also analyzed the associations of diseases incidence and lipid levels with PRS and combined lifestyle stratified by sex as shown in Figures XII through XIV in the Data Supplement. The results were generally similar to the results from the combined analyses. However, given the same PRS quantile and combined lifestyle status, males tended to have a higher level of triglyceride compared with females, and the larger ARRs among males suggested that males might have more absolute benefits from lifestyle adherence regarding to the prevention of CAD, AF, and T2D.
Discussion
By using enhanced PRSs in this large-scale study of around 300 000 UKBB participants, no significant multiplicative interactions were found between genetic risk and lifestyle for CAD, AF, and T2D incidence under the Cox proportional hazard model. However, we observed a significant association between PRS grouping and ARR from lifestyle adherence for CAD, AF, and T2D, where healthy lifestyle decreased the absolute risk much more significantly in individuals at high PRS percentiles. We also found a significant positive interaction between PRS and lifestyle for triglyceride.
In general, the impact of lifestyle factors and PRS on CAD, AF, and T2D incidence observed in our study is in line with previous reports by Khera et al18 and Said et al.19 Both poorer lifestyle and higher genetic risk could lead to higher risks of developing CAD, AF, and T2D. However, our study has extended the investigations of the previous studies in several ways.
First, we investigated the effect of genetic factor and its interaction with lifestyle based on more accurate quantification of genetic risks. More specifically, by leveraging functional annotations in genetic risk prediction, we developed enhanced PRS for CAD, AF, T2D, TC, triglyceride, and LDL-C using AnnoPred.15 Applying to UKBB data set, we showed that our enhanced PRS outperformed PRS developed by other state-of-the-art methods (P+T, LDpred, PRS-CS) for all 6 disease/traits with higher AUCs/R2. It is worth noting that in both Khera et al’s18 and Said et al’s19 studies, they used restrictive PRS which could be regarded as a special case of P+T PRS. They generated the PRS based on empirical clumping and thresholding parameters without tuning; hence the performance of the PRS they used was expected to be even worse than the P+T PRS as we showed in Figure 1. Given the significant improvement of the prediction performance of AnnoPred PRS over P+T PRS, our investigation results based on the enhanced PRS would be more accurate and robust.
Second, we studied both multiplicative (HR) and additive (AR) interactions between PRS and lifestyle for CAD, AF, and T2D with a more refined population stratification. We observed that participants with higher PRS could get similar relative but greater absolute benefits by leading a healthy lifestyle. Previous work by Khera et al18 also found that people with high PRS (>80% PRS) could get larger ARR from lifestyle adherence compared with other PRS groups. However, based on their restricted PRS and coarse population stratification, they reported that of the 7814 participants in the ARIC (Atherosclerosis Risk in Communities) study, the ARR of CAD from healthy lifestyle was 2.7%, 2.5%, and 5.6% in low (<20% PRS), intermediate (20%–80% PRS), and high PRS (>80% PRS) groups, respectively; from which they could hardly conclude whether the large ARR only existed in the high PRS group, or there was a trend that higher PRS would lead to larger ARR. Our finer stratification allowed us to observe such a trend clearly as illustrated in Figures 2 and 3, where the HR lines were parallel to each other and ARRs increased along with PRS for each of the 3 diseases. These results also provided the insight that PRS was not only valuable to identify the high PRS group, but also informative for further stratification among participants with low to intermediate PRS and among participants within the commonly categorized high PRS group (>80% PRS).
In this study, we also examined the characteristics of the extreme tails of the PRS distributions (PRS >99%). We observed a sharply increased risk in this extremely high PRS group that deviates greatly from the rest of the population, which was consistent with previous reports.15,25 Interestingly, the trend that higher PRS could lead to larger ARR from lifestyle adherence still held for this extremely high PRS group. And strikingly, based on our enhanced PRS, we were able to identify 1% (PRS >99%) of the population that could get 4.5%, 7.2%, and 12.4% ARR from lifestyle adherence for CAD, AF, and T2D, respectively.
Besides, in addition to the traits analyzed by Khera et al18 and Said et al,19 we also considered intermediate traits, namely blood lipid levels (LDL-C, triglyceride, TC). Although blood lipid levels have been implicated in CVD risk for some time,30 to our knowledge, our current study is the first to investigate the interactions between genetic risk and combined lifestyle on blood lipid levels through PRS. Consistent with previous studies,14,31 poorer lifestyle and higher PRS could lead to higher levels of LDL-C, triglyceride, and TC. Among 3 types of lipids, being the only one of the diagnosis items of metabolic syndrome,32 triglyceride appeared to be most responsive to lifestyle modification. This was also in line with the American College of Cardiology/American Heart Association Guideline where the main target of lifestyle therapies was metabolic syndrome.32 And among the 4 individual lifestyle components, BMI played the most prominent role, as long recognized.33 Additionally, we observed that people in all groups, either with higher or lower PRS, could benefit to an extent from lifestyle adherence in terms of lipids reduction. More specifically, we identified a positive significant interaction between continuous PRS and lifestyle for triglyceride (Table IX in Data Supplement), suggesting people with higher PRS could benefit more from lifestyle modification in terms of triglyceride reduction. While the effect of lifestyle factors on triglyceride is well-established and unsurprising,32 it is not expected that the effect is dependent on genetic factors and that this is not present for cholesterol (which is also influenced by genetic and lifestyle factors32). Furthermore, we also observed that the genetic burden on these lipid levels could not be completely overcome by lifestyle modification. For the group with extremely high PRS (>99%), even with an ideal lifestyle, the mean levels of TC and LDL-C were still within an undesired range with high risk, and the mean level of triglyceride was also at a broadline high designation. These findings suggest that lifestyle modification is likely to be adequate for people with low-middle PRS, especially for the management of triglyceride; but more interventions (eg, frequent surveillance, pharmaceutical interventions, and more intense lifestyle interventions) are required for people with high PRS to manage LDL-C and TC. This also adds the justification that for high risk individuals, the strategy combining lifestyle and lipid lowering drug treatment since the start may be superior to the strategy stepping from lifestyle to drug therapy when the former fails.34,35
We note several limitations of this study. First, although we have used enhanced PRSs with better prediction performance than other PRSs, the prediction capacities of PRSs are still moderate. The AUCs of using PRS alone in our study for three diseases ranged from 0.63 to 0.65, which were much lower than AUCs (0.75–0.80) of using comprehensive clinical prediction models as reported by previous studies.36–38 And the lipids PRS could only explain <10% variation of lipid levels, which were also relatively poor. Further development of risk prediction models incorporating other predictors could improve the performance and making these models relevant for clinical studies. Second, a causal relationship cannot be inferred between lifestyle and cardiovascular phenotypes given this study design,39 especially in the analysis of blood lipid, where the chronological order of lifestyle status and lipid levels was unknown. Third, since large-scale GWAS summary statistics for stroke and heart failure independent of UKBB were unavailable, we only considered CAD and AF within the list of CVDs. Thus, future research is needed to investigate the interactions between PRS and lifestyle for all-cause CVD. Another limitation is that we used self-reported characteristics for lifestyle factors such as physical activity, smoking, and diet status, which might be inaccurate and reduced the power of our study. In addition, although we analyzed the association of diseases/traits with lifestyle in PRS groups stratified by sex and found the results were generally similar to the combined analysis, PRS constructed based on sex-specific GWAS are required to further investigate possible sex-differences in the interactions between genetic risk and lifestyle factors.40–42 Finally, the present analyses were performed only on individuals of White British descent, and the UKBB participants were reported to be possibly healthier than the general population,43 which together would decrease the generalizability of our results to other study populations.
In conclusion, genetic risk and combined lifestyle are independently associated with the risks of CAD, AF, and T2D with regard to the log scale of HR. However, individuals at high genetic risk could derive greater benefit from lifestyle adherence in terms of the management of lipid levels and diseases ARRs.
Acknowledgments
We conducted the research using the UK Biobank resource under an approved data request (ref: 29900). We sincerely thank many genome-wide association studies (GWAS) consortia for making their GWAS summary data publicly accessible (http://www.kp4cd.org/datasets/mi).
Footnote
Nonstandard Abbreviations and Acronyms
- AF
- atrial fibrillation
- ARIC
- Atherosclerosis Risk in Communities
- ARR
- absolute risk reduction
- BMI
- body mass index
- CAD
- coronary artery disease
- CS
- continuous shrinkage
- CVD
- cardiovascular diseases
- GWAS
- genome-wide association studies
- HR
- hazard ratio
- LDL-C
- low-density lipoprotein cholesterol
- PRS
- polygenic risk scores
- T2D
- type 2 diabetes
- TC
- total cholesterol
- UKBB
- UK Biobank
Supplemental Material
References
1.
Huxley RR, Misialek JR, Agarwal SK, Loehr LR, Soliman EZ, Chen LY, Alonso A. Physical activity, obesity, weight change, and risk of atrial fibrillation: the Atherosclerosis Risk in Communities study. Circ Arrhythm Electrophysiol. 2014;7:620–625. doi: 10.1161/CIRCEP.113.001244
2.
Mozaffarian D. Dietary and policy priorities for cardiovascular disease, diabetes, and obesity: a comprehensive review. Circulation. 2016;133:187–225. doi: 10.1161/CIRCULATIONAHA.115.018585
3.
Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Carson AP, Chamberlain AM, Chang AR, Cheng S, Das SR, et al; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics-2019 update: a report from the American Heart Association. Circulation. 2019;139:e56–e528. doi: 10.1161/CIR.0000000000000659
4.
Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, Hahn EJ, Himmelfarb CD, Khera A, Lloyd-Jones D, McEvoy JW, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;140:e596–e646. doi: 10.1161/CIR.0000000000000678
5.
Kathiresan S, Srivastava D. Genetics of human cardiovascular disease. Cell. 2012;148:1242–1257. doi: 10.1016/j.cell.2012.03.001
6.
Khera AV, Kathiresan S. Genetics of coronary artery disease: discovery, biology and clinical translation. Nat Rev Genet. 2017;18:331–344. doi: 10.1038/nrg.2016.160
7.
Levin MG, Rader DJ. Polygenic risk scores and coronary artery disease: ready for prime time? Circulation. 2020;141:637–640. doi: 10.1161/CIRCULATIONAHA.119.044770
8.
Nikpay M, Goel A, Won HH, Hall LM, Willenborg C, Kanoni S, Saleheen D, Kyriakou T, Nelson CP, Hopewell JC, et al. A comprehensive 1000 Genomes–based genome-wide association meta-analysis of coronary artery disease. Nat Genet. 2015;47:1121–1130. doi: 10.1038/ng.3396
9.
Scott RA, Scott LJ, Mägi R, Marullo L, Gaulton KJ, Kaakinen M, Pervjakova N, Pers TH, Johnson AD, Eicher JD, et al; DIAbetes Genetics Replication And Meta-analysis (DIAGRAM) Consortium. An expanded genome-wide association study of type 2 diabetes in Europeans. Diabetes. 2017;66:2888–2902. doi: 10.2337/db16-1253
10.
Willer CJ, Schmidt EM, Sengupta S, Peloso GM, Gustafsson S, Kanoni S, Ganna A, Chen J, Buchkovich ML, Mora S, et al; Global Lipids Genetics Consortium. Discovery and refinement of loci associated with lipid levels. Nat Genet. 2013;45:1274–1283. doi: 10.1038/ng.2797
11.
Christophersen IE, Rienstra M, Roselli C, Yin X, Geelhoed B, Barnard J, Lin H, Arking DE, Smith AV, Albert CM, et al; METASTROKE Consortium of the ISGC, Neurology Working Group of the CHARGE Consortium; AFGen Consortium. Erratum: Large-scale analyses of common and rare variants identify 12 new loci associated with atrial fibrillation. Nat Genet. 2017;49:1286. doi: 10.1038/ng0817-1286c
12.
Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015;4:7. doi: 10.1186/s13742-015-0047-8
13.
Vilhjálmsson BJ, Yang J, Finucane HK, Gusev A, Lindström S, Ripke S, Genovese G, Loh PR, Bhatia G, Do R, et al; Schizophrenia Working Group of the Psychiatric Genomics Consortium, Discovery, Biology, and Risk of Inherited Variants in Breast Cancer (DRIVE) study. Modeling linkage disequilibrium increases accuracy of polygenic risk scores. Am J Hum Genet. 2015;97:576–592. doi: 10.1016/j.ajhg.2015.09.001
14.
Ge T, Chen CY, Ni Y, Feng YA, Smoller JW. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nat Commun. 2019;10:1776. doi: 10.1038/s41467-019-09718-5
15.
Hu Y, Lu Q, Powles R, Yao X, Yang C, Fang F, Xu X, Zhao H. Leveraging functional annotations in genetic risk prediction for human complex diseases. PLoS Comput Biol. 2017;13:e1005589. doi: 10.1371/journal.pcbi.1005589
16.
Dudbridge F. Power and predictive accuracy of polygenic risk scores. PLoS Genet. 2013;9:e1003348. doi: 10.1371/journal.pgen.1003348
17.
Inouye M, Abraham G, Nelson CP, Wood AM, Sweeting MJ, Dudbridge F, Lai FY, Kaptoge S, Brozynska M, Wang T, et al; UK Biobank CardioMetabolic Consortium CHD Working Group. Genomic risk prediction of coronary artery disease in 480,000 adults: implications for primary prevention. J Am Coll Cardiol. 2018;72:1883–1893. doi: 10.1016/j.jacc.2018.07.079
18.
Khera AV, Emdin CA, Drake I, Natarajan P, Bick AG, Cook NR, Chasman DI, Baber U, Mehran R, Rader DJ, et al. Genetic risk, adherence to a healthy lifestyle, and coronary disease. N Engl J Med. 2016;375:2349–2358. doi: 10.1056/NEJMoa1605086
19.
Said MA, Verweij N, van der Harst P. Associations of combined genetic and lifestyle risks with incident cardiovascular disease and diabetes in the UK Biobank study. JAMA Cardiol. 2018;3:693–702. doi: 10.1001/jamacardio.2018.1717
20.
Hindy G, Wiberg F, Almgren P, Melander O, Orho-Melander M. Polygenic risk score for coronary heart disease modifies the elevated risk by cigarette smoking for disease incidence. Circ Genom Precis Med. 2018;11:e001856. doi: 10.1161/CIRCGEN.117.001856
21.
Merino J, Jablonski KA, Mercader JM, Kahn SE, Chen L, Harden M, Delahanty LM, Araneta MRG, Walford GA, Jacobs SBR, et al; Diabetes Prevention Program Research Group. Interaction between type 2 diabetes prevention strategies and genetic determinants of coronary artery disease on cardiometabolic risk factors. Diabetes. 2020;69:112–120. doi: 10.2337/db19-0097
22.
Khera AV, Chaffin M, Wade KH, Zahid S, Brancale J, Xia R, Distefano M, Senol-Cosar O, Haas ME, Bick A, et al. Polygenic prediction of weight and obesity trajectories from birth to adulthood. Cell. 2019;177:587–596.e9. doi: 10.1016/j.cell.2019.03.028
23.
Mavaddat N, Michailidou K, Dennis J, Lush M, Fachal L, Lee A, Tyrer JP, Chen TH, Wang Q, Bolla MK, et al; ABCTB Investigators; kConFab/AOCS Investigators; NBCS Collaborators. Polygenic risk scores for prediction of breast cancer and breast cancer subtypes. Am J Hum Genet. 2019;104:21–34. doi: 10.1016/j.ajhg.2018.11.002
24.
Natarajan P, Young R, Stitziel NO, Padmanabhan S, Baber U, Mehran R, Sartori S, Fuster V, Reilly DF, Butterworth A, et al. Polygenic risk score identifies subgroup with higher burden of atherosclerosis and greater relative benefit from statin therapy in the primary prevention setting. Circulation. 2017;135:2091–2101. doi: 10.1161/CIRCULATIONAHA.116.024436
25.
Khera AV, Chaffin M, Aragam KG, Haas ME, Roselli C, Choi SH, Natarajan P, Lander ES, Lubitz SA, Ellinor PT, et al. Genome-wide polygenic scores for common diseases identify individuals with risk equivalent to monogenic mutations. Nat Genet. 2018;50:1219–1224. doi: 10.1038/s41588-018-0183-z
26.
Baigent C, Keech A, Kearney PM, Blackwell L, Buck G, Pollicino C, Kirby A, Sourjina T, Peto R, Collins R, et al; Cholesterol Treatment Trialists’ (CTT) Collaborators. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet. 2005;366:1267–1278. doi: 10.1016/S0140-6736(05)67394-1
27.
Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, Downey P, Elliott P, Green J, Landray M, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12:e1001779. doi: 10.1371/journal.pmed.1001779
28.
Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults. Circulation. 2002;106:3143–3421.
29.
Grundy SM, Cleeman JI, Merz CN, Brewer HB, Clark LT, Hunninghake DB, Pasternak RC, Smith SC, Stone NJ; National Heart, Lung, and Blood Institute; American College of Cardiology Foundation; American Heart Association. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation. 2004;110:227–239. doi: 10.1161/01.CIR.0000133317.49796.0E
30.
Shepherd J, Cobbe SM, Ford I, Isles CG, Lorimer AR, MacFarlane PW, McKillop JH, Packard CJ. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. West of Scotland Coronary Prevention Study Group. N Engl J Med. 1995;333:1301–1307. doi: 10.1056/NEJM199511163332001
31.
Mannu GS, Zaman MJ, Gupta A, Rehman HU, Myint PK. Evidence of lifestyle modification in the management of hypercholesterolemia. Curr Cardiol Rev. 2013;9:2–14. doi: 10.2174/157340313805076313
32.
Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, Braun LT, de Ferranti S, Faiella-Tommasino J, Forman DE, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;139:e1082–e1143. doi: 10.1161/CIR.0000000000000625
33.
Keys A. Effects of different dietary fats on plasma-lipid levels. Lancet. 1965;1:318–319. doi: 10.1016/s0140-6736(65)91053-6
34.
Bonaccio M, Di Castelnuovo A, Costanzo S, Persichillo M, De Curtis A, Cerletti C, Donati MB, de Gaetano G, Iacoviello L; Moli-sani Study Investigators. Interaction between Mediterranean diet and statins on mortality risk in patients with cardiovascular disease: findings from the Moli-sani Study. Int J Cardiol. 2019;276:248–254. doi: 10.1016/j.ijcard.2018.11.117
35.
Stone NJ, Robinson JG, Lichtenstein AH, Bairey Merz CN, Blum CB, Eckel RH, Goldberg AC, Gordon D, Levy D, Lloyd-Jones DM, et al; American College of Cardiology/American Heart Association Task Force on Practice Guidelines. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014;129(25 suppl 2):S1–45. doi: 10.1161/01.cir.0000437738.63853.7a
36.
Elliott J, Bodinier B, Bond TA, Chadeau-Hyam M, Evangelou E, Moons KGM, Dehghan A, Muller DC, Elliott P, Tzoulaki I. Predictive accuracy of a polygenic risk score-enhanced prediction model vs a clinical risk score for coronary artery disease. JAMA. 2020;323:636–645. doi: 10.1001/jama.2019.22241
37.
Alonso A, Krijthe BP, Aspelund T, Stepas KA, Pencina MJ, Moser CB, Sinner MF, Sotoodehnia N, Fontes JD, Janssens AC, et al. Simple risk model predicts incidence of atrial fibrillation in a racially and geographically diverse population: the CHARGE-AF consortium. J Am Heart Assoc. 2013;2:e000102. doi: 10.1161/JAHA.112.000102
38.
Wilkinson L, Yi N, Mehta T, Judd S, Garvey WT. Development and validation of a model for predicting incident type 2 diabetes using quantitative clinical data and a Bayesian logistic model: a nationwide cohort and modeling study. PLoS Med. 2020;17:e1003232. doi: 10.1371/journal.pmed.1003232
39.
Vandenbroucke JP, Broadbent A, Pearce N. Causality and causal inference in epidemiology: the need for a pluralistic approach. Int J Epidemiol. 2016;45:1776–1786. doi: 10.1093/ije/dyv341
40.
Zeng Y, Nie C, Min J, Chen H, Liu X, Ye R, Chen Z, Bai C, Xie E, Yin Z, et al. Sex differences in genetic associations with longevity. JAMA Netw Open. 2018;1:e181670. doi: 10.1001/jamanetworkopen.2018.1670
41.
Winham SJ, de Andrade M, Miller VM. Genetics of cardiovascular disease: importance of sex and ethnicity. Atherosclerosis. 2015;241:219–228. doi: 10.1016/j.atherosclerosis.2015.03.021
42.
Hartiala JA, Tang WH, Wang Z, Crow AL, Stewart AF, Roberts R, McPherson R, Erdmann J, Willenborg C, Hazen SL, et al. Genome-wide association study and targeted metabolomics identifies sex-specific association of CPS1 with coronary artery disease. Nat Commun. 2016;7:10558. doi: 10.1038/ncomms10558
43.
Fry A, Littlejohns TJ, Sudlow C, Doherty N, Adamska L, Sprosen T, Collins R, Allen NE. Comparison of sociodemographic and health-related characteristics of UK Biobank participants with those of the general population. Am J Epidemiol. 2017;186:1026–1034. doi: 10.1093/aje/kwx246
Information & Authors
Information
Published In
Copyright
© 2021 American Heart Association, Inc.
Versions
You are viewing the most recent version of this article.
History
Received: 29 June 2020
Accepted: 15 December 2020
Published online: 12 January 2021
Published in print: February 2021
Keywords
Subjects
Authors
Disclosures
Disclosures None.
Sources of Funding
This study was supported in part by the Yale World Scholars Program sponsored by the China Scholarship Council, National Institutes of Health grants R01 GM122078 and R01 GM134005, and National Science Foundation (NSF) grants DMS 1713120 and DMS 1902903.
Metrics & Citations
Metrics
Citations
Download Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Select your manager software from the list below and click Download.
- Bisphenol-A (BPA) Exposure as a Risk Factor for Non-Communicable Diseases, Bisphenols - New Environmental, Pathophysiological and Social Perspectives [Working Title], (2024).https://doi.org/10.5772/intechopen.112623
- Population Heterogeneity and Selection of Coronary Artery Disease Polygenic Scores, Journal of Personalized Medicine, 14, 10, (1025), (2024).https://doi.org/10.3390/jpm14101025
- Genetic Susceptibility to Arrhythmia Phenotypes in a Middle Eastern Cohort of 14,259 Whole-Genome Sequenced Individuals, Journal of Clinical Medicine, 13, 4, (1102), (2024).https://doi.org/10.3390/jcm13041102
- Evaluating the Efficacy of Type 2 Diabetes Polygenic Risk Scores in an Independent European Population, International Journal of Molecular Sciences, 25, 2, (1151), (2024).https://doi.org/10.3390/ijms25021151
- Genomic risk prediction of cardiovascular diseases among type 2 diabetes patients in the UK Biobank, Frontiers in Bioinformatics, 3, (2024).https://doi.org/10.3389/fbinf.2023.1320748
- Synergistic effect of genetic predisposition and lifestyle for coronary heart disease, European Journal of Preventive Cardiology, (2024).https://doi.org/10.1093/eurjpc/zwae077
- En route to conquer the silent killer “hypertension”: Integration of polygenic risk score with non-genetic determinants, Hypertension Research, (2024).https://doi.org/10.1038/s41440-024-01826-0
- Modification of coronary artery disease clinical risk factors by coronary artery disease polygenic risk score, Med, 5, 5, (459-468.e3), (2024).https://doi.org/10.1016/j.medj.2024.02.015
- Polygenic risk scores in atrial fibrillation: Associations and clinical utility in disease prediction, Heart Rhythm, 21, 6, (913-918), (2024).https://doi.org/10.1016/j.hrthm.2024.02.006
- Leveraging Functional Annotations Improves Cross-Population Genetic Risk Prediction, Statistics in Precision Health, (453-471), (2024).https://doi.org/10.1007/978-3-031-50690-1_18
- See more
Loading...
View Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginPurchase Options
Purchase this article to access the full text.
eLetters(0)
eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.
Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.