Supplemental Material Table of Contents
Supplemental Table 1. Genetic Instruments for FGF23 used in the Mendelian Randomization Analyses.
Supplemental Table 2. Definition and adjudication of heart failure cases across participating studies in the HERMES Consortium.
Supplemental Table 3. Mendelian Randomization Estimates for the Effect of Genetically- Predicted FGF23 on Heart Failure and Heart Failure Subtypes in analyses including rs2769071 (ABO).
Supplemental Table 4. Mendelian Randomization Estimates for the Effect of Genetically-
Predicted FGF23 on Heart Failure and Heart Failure Subtypes in analyses excluding rs17216707
(CYP24A1) and rs11741640 (RGS14).
Supplemental Table 1. Genetic Instruments for FGF23 used in mendelian randomization analyses
1SNP Nearest
gene
αChr Position
FGF23 Increasing
Allele
Other Allele
FGF23 Increasing Allele
Frequency
βBeta
λSE P-value F-
statistic
rs17216707 CYP24A1 20 52732362 T C 0.80 0.054 0.005 3.0 × 10
-24103.5
rs11741640 RGS14 5 176792743 G A 0.73 0.039 0.005 1.6 × 10
-1668.2
rs17479566 LINC01506 9 71198013 T C 0.22 0.031 0.005 2.0 × 10
-0936.0
rs9925837 LINC01229 16 79927303 G A 0.13 0.035 0.006 5.1 × 10
-0934.2
rs2769071* ABO 9 136145974 G A 0.37 0.037 0.005 6.1 × 10
-1770.1
Chr, chromosome; SE, standard error; SNP, single nucleotide polymorphism.
*Excluded from primary analyses due to pleiotropic effects of SNP.
α
Nearest gene by physical distance to the lead SNP.
β
Allele frequency data from 1000 Genome Phase 1 genotype data.
λ
Beta coefficients from meta-analysis of genome-wide association studies of log-transformed FGF23.
Supplemental Table 2. Definition and adjudication of heart failure cases across participating studies in the HERMES Consortium
2, 3Study Definition and adjudication of heart failure cases
1 Atherosclerosis Risk in Communities (ARIC)
Incident heart failure (HF) was defined as the first HF hospitalization or presence of HF code on death certificate since baseline visit through 2013. Discharge records and death certificates that showed a HF code in any position with International Classification of Diseases Code, Ninth Revision (ICD-9) code 428.x, and deaths with ICD-9/10 codes of either 428.x or I50 were considered as HF. Prevalent HF was ascertained at first visit. If the participant reported to have taken any medication for HF or qualifies for the Gothenburg Criteria, then the participant had prevalent HF at baseline.
2
A systems BIOlogy Study to TAilored Treatment in Chronic Heart Failure
(BIOSTAT-CHF)
Heart failure was defined based on physician diagnosis, previous documented admission with HF requiring diuretic treatment, treatment with furosemide ≥20 mg/day or equivalent.
3 Cardiovascular Health Study (CHS)
Incident HF events were identified by self-report or administrative data validated by
physician’s review of medical records, as described in previous reports (PMID: 1669507). In brief, HF was defined on the basis of cardiomegaly and pulmonary edema on chest X-ray; or dilated ventricle and wail-motion abnormalities by echocardiography or contrast
ventriculography; or congestive HF diagnosed by physician, plus receiving medical treatment (diuretic plus either digitalis, vasodilator or angiotensin converting enzyme inhibitor).
4 Copenhagen Cardiovascular Genetic study (COGEN)
Inclusion criteria: patients ≥18 years, Left ventricular ejection fraction (LVEF) <40% or symptoms of clinical HF assessed by a physician including NYHA staging > 1.
Exclusion criteria: patients with cardiac valvular pathology (e.g., aortic stenosis), HTx, PAH or other structural heart disease.
5 deCODE Heart Failure Study (deCODE) Heart failure case status was assigned based on ICD-9 or ICD-10 codes for discharge diagnoses (ICD-10: I50 and subcodes, ICD-9: 428 and subcodes).
6 Estonian Genome Center at the University of Tartu (EGCUT)
Heart failure status was assigned based on ICD-10: I50 and subcodes.
7
Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study
(EPHESUS)
Patients were recruited 3 to 14 days after acute myocardial infarction according to the following criteria: acute myocardial infarction as documented according to standard criteria;
LV dysfunction as documented by a LVEF ≤ 40 percent on echocardiography, radionuclide angiography, or angiography of the left ventricle after the index acute myocardial infarction and before randomization; and HF as documented by the presence of pulmonary rales, chest radiography showing pulmonary venous congestion, or the presence of a third heart sound. In patients with diabetes who met the criteria for LV dysfunction after acute myocardial
infarction, symptoms of HF did not have to be demonstrated, since such patients have an increased risk of cardiovascular events similar to that of nondiabetic patients with symptoms of HF.
8 European Prospective Investigation of Cancer (EPIC)-Norfolk
Heart failure definition was based on hospital admission or death record listing HF code - ICD-10: I50 and subcodes.
9 Framingham Heart Study (FHS)
Criteria for defining HF in the FHS have been described previously (PMID: 5122894, 16837677). In brief, HF was considered to be present if two major or one major plus two minor criteria were present in the absence of an alternative explanation for the symptoms and
signs. Major criteria are defined as paroxysmal nocturnal dyspnea, orthopnea, jugular venous distention, hepatojugular reflux, pulmonary rales, radiographic evidence of cardiomegaly, acute pulmonary edema, third heart sound, central venous pressure >16 cm of water, and weight loss >4.5 kg during first 5 days of treatment for suspected heart failure. Minor criteria are defined as bilateral ankle edema, nocturnal cough, dyspnea on ordinary exertion,
hepatomegaly, pleural effusion, and heart rate >120 beats per minute.
10 FINRISK Study
Individuals with a diagnosis corresponding to HF in the nationwide hospital discharge or cause of death registers (ICD-10: I50, I110, I130 and I132; ICD-9: 4029B, 404, 4148, 428; ICD-7:
42700, 42710, 428) or special drug reimbursement for heart failure medications (requires a medical certificate that meets predefined criteria for heart failure).
11
Genetics of Diabetes Audit and Research Tayside Scotland
(GoDARTS)
HF cases were defined by the presence of at least one of the following criteria:
1. Echocardiographic evidence of LV systolic impairment and diuretic prescription 2. Admission to hospital with HF and receipt of a loop diuretic prescription
Patients who were never prescribed a loop diuretic were not classified as HF cases. Date of HF diagnosis was taken as either the date of the earliest echocardiogram or the date of the earliest admission to hospital for HF.
12 Genetic Risk Assessment of Defibrillator Events (GRADE)
Inclusion criteria were patients who were ≥18 years of age with a diagnosis of at least moderate systolic LV dysfunction (EF ≤30%), and who had an ICD at the University of Pittsburgh Medical Center, Emory University Medical Center, Massachusetts General Hospital, Ohio State University Medical Center, Mid-Ohio Cardiology or the Pittsburgh Veterans Affairs Medical Center. Subjects were excluded if they had intractable Class IV heart failure, and conditions (other than HF) that were expected to limit survival to less than 6 months.
13 LUdwigshafen RIsk and Cardiovascular Health study (LURIC)
A clinical diagnosis of (left) HF was defined by the combined presence of symptoms of dyspnea on exertion and cardiac disease with impaired left ventricular function. Left ventricular dysfunction or impaired left ventricular function was defined by imaging techniques, such as echocardiography and left ventricular angiography, and graded semi- quantitatively into normal, minimal, moderate and severe impairment of left ventricular pump function. Based on the underlying cardiac disease, heart failure/left ventricular dysfunction was either of ischemic (in case of CAD) or of non-ischemic origin (dilated cardiomyopathy being the most frequent non-CAD disease). Mis- or underreporting of heart failure and/or left ventricular dysfunction was avoided in LURIC since virtually all LURIC participants, except for family members without coronary angiography, underwent echocardiography and left ventricular angiography.
14 Malmö Diet and Cancer Study (MDCS)
Heart failure was ascertained from the Swedish Hospital Discharge Register using diagnosis codes 427.00, 427.10, and 428.99 for International Classification of Diseases-8th Revision (ICD-8), 428 for the 9th Revision (ICD-9), and I50 and I11.0 for the 10th Revision (ICD-10) as primary diagnosis, according to a previous validation study (PMID:15916919).
15 Penn Heart Failure Study (PHFS)
Prevalent heart failure diagnosed by a HF cardiologist based on clinical evaluation and cardiac imaging
16 Prospective Investigation of the Vasculature in Uppsala Seniors
The medical records for all individuals with heart failure diagnosis in any position in the Swedish hospital discharge register were reviewed by two physicians who were blinded to the
(PIVUS) baseline data. They classified the cases as definite, questionable, or miscoded according to the European Society of Cardiology recommendations. They considered ICD HF codes 427.00, 427.10, 428 (ICD-9), I50 (ICD-10) and hypertensive heart disease with heart failure, I11.0 (ICD-10) as possible diagnosis of HF. Further details in PMID:15916919.
17 Prevention of REnal and Vascular ENd-stage Disease (PREVEND)
Heart failure cases were ascertained using criteria in accordance with the HF Guidelines of the European Society of Cardiology. In- and outpatient files were inspected for the presence of HF at baseline and for new onset heart failure, by recording signs, symptoms, and objective evidence of HF. In total, 586 individual cases were identified as suspected HF. An endpoint adjudication committee of seven independent experts evaluated all suspected cases of new onset heart failure. Each case was validated by two different experts by reviewing anonymized clinical charts, hospitalization, and physician office records in order to ascertain the incidence of HF. In case of consensus, patients were classified as ‘definite new onset HF’, ‘definite no new onset HF’, or ‘definite HF, with date of onset before time of recruitment’. In case of difference of opinion about an individual case, the committee made a joint decision.
18
PROspective Study of Pravastatin in the Elderly at Risk for vascular disease
(PROSPER)
Cases were defined by hospitalization for HF with a definition based on a combination of symptoms (e.g., shortness of breath) and signs, including chest radiograph with fluid congestion or echocardiogram with severely diminished LV function. All outcomes were adjudicated by an expert committee blinded to randomized study medication and using pre‐
defined criteria.
19 Regeneron/Geisinger (DiscovEHR)
Heart failure status was assigned based on ICD-10: I50 and subcodes.
20 Rotterdam study 1
Prevalent HF at baseline was assessed using a validated score based on the European Society of Cardiology recommendation, identified from hospital discharge diagnoses, and
retrospective medical records screening. Cases of incident HF were obtained by continuously monitoring participants for the occurrence of HF during follow-up through general
practitioners’ records and hospital discharge diagnoses. The date of incident HF was defined as the day of the first occurrence of symptoms suggestive of HF, obtained from the medical records, or the day of receipt of a first prescription for a loop diuretic or an ACE inhibitor indicated for treatment of HF, whichever came first. The diagnosis of HF was classified as definite, probable, possible, or unlikely in accordance with the criteria from the European Society of Cardiology. Potential cases were ascertained by two research physicians and verified by a cardiologist. Only definite and probable cases were considered in the analyses.
21 Study of Health in Pomerania (SHIP)
Heart failure was defined according to a modified Rotterdam definition (PMID: 10213348).
Prevalent HF cases in SHIP were defined as having history of HF (either chest pain during exercise, bypass, heart transplant, atrial flutter or fibrillation, LV hypertrophy in individuals aged 45 or older, known MI) and HF symptoms (dyspnea at exercise or swollen legs at evening) that were not related to bronchitis (bronchitis that occurred recently or during the last 12 months).
22 Stabilization of Plaque using Darapladib-Thrombolysis in Myocardial Infarction 52
(SOLID-TIMI52)
Heart failure status at enrolment was identified from medical record with no specific definition. HF hospitalizations adjudicated during follow up were defined as admission to hospital or attendance at an acute health care facility for administration of intravenous diuretic treatment, escalation of diuretic doses, and/or inotropes. Confirmation of HF diagnosis was
obtained by chest imaging demonstrating pulmonary congestion or edema, or, in patients without available chest imaging, at least one of the following: Pulmonary edema, (i.e., rales
>1/3 up the lung fields thought to be of cardiac causes), pulmonary capillary wedge pressure
>18 mmHg or BNP >500 pg/ml (or NT-terminal prohormone BNP >2500 pg/ml).
23 TwinGene Heart failure status was assigned based on ICD-10: I50; ICD-8 and ICD-9 428
24 UK Biobank
Individuals with self-reported "HF/pulmonary edema" or "cardiomyopathy"; or who carry an International Classification of Diseases (ICD)-10 or ICD-9 billing code for heart/ventricular failure or cardiomyopathy (ICD-10: I11.0, I13.0, I13.2, I25.5, I42.0, I42.5, I42.8, I42.9, I50.0, I50.1, I50.9; ICD-9: 4254, 4280, 4281, 4289). Individuals with self-reported or an ICD-10 based classification of hypertrophic cardiomyopathy were excluded.
25 Uppsala Longitudinal Study of Adult Men (ULSAM)
The medical records for all individuals with heart failure diagnosis in any position in the Swedish hospital discharge register were reviewed by two physicians who were blinded to the baseline data. They classified the cases as definite, questionable, or miscoded according to the European Society of Cardiology recommendations. They considered ICD HF codes 427.00, 427.10, 428 (ICD-9), I50 (ICD-10) and hypertensive heart disease with heart failure, I11.0 (ICD-10) as possible diagnosis of HF.
26 Women’s Genome Health Study (WGHS)
Heart failure cases were ascertained by cardiologists from medical records. Cases of incident nonfatal HF were confirmed if either the Framingham Heart Study (mainly physical
examination and radiographic data) or Cardiovascular Health Study criteria (predominantly based on the treating physician's diagnosis and use of specific therapy) were met. Fatal HF cases included those not identified as a case of HF prior to death and classified into “Definite”
or “Probable” fatal HF based on medical records and death certificate with next-of-kin or physician confirmation.
Supplemental Table 3. Mendelian Randomization Estimates for the Effect of Genetically- Predicted FGF23 on Heart Failure and Heart Failure Subtypes in analyses including rs2769071 (ABO)
Outcome and data source Number of HF cases
Inverse Variance Weighted MR Odds Ratio (95%CI) P-value Heart Failure (HERMES)
Overall 47,309 1.39 (1.12, 1.72) 0.002
Heart Failure (BioVU)
Overall 18,415 1.31 (0.98, 1.75) 0.067
Low eGFR PRS
3,086 3.26 (1.61,6.58) 0.001
High eGFR PRS
15,329 1.09 (0.79, 1.49) 0.609
P-interaction* 0.005
HF Preserved Ejection Fraction (BioVU)
Overall 13,141 1.40 (1.00,1.96) 0.050
Low eGFR PRS 2,223 6.77 (2.97, 15.40) <0.001
High eGFR PRS 10,918 1.02 (0.70, 1.47) 0.933
P-interaction* 3.72x10
-5HF Reduced Ejection Fraction (BioVU)
Overall 3,394 0.90 (0.44, 1.83) 0.763
Low eGFR PRS 554 0.27 (0.05, 1.50) 0.135
High eGFR PRS 2,840 1.14 (0.52, 2.51) 0.739
P-interaction* 0.135
HF Mid-range Ejection Fraction (BioVU)
Overall 388 0.43 (0.06, 3.21) 0.411
Low eGFR PRS 70 1.14 (0.01,118.37) 0.955
High eGFR PRS 318 0.34 (0.04, 3.15) 0.341
P-interaction* 0.645
Low eGFR PRS: Estimated glomerular filtration rate polygenic risk score (eGFR PRS) less than one standard deviation below the mean.
High eGFR PRS: Estimated glomerular filtration rate polygenic risk score (eGFR PRS) greater than one standard deviation below the mean.
*P-interaction for the difference in the causal effect of genetically-predicted FGF23 on HF in low versus high eGFR PRS.
Abbreviations: CI, confidence interval; HF, heart failure; MR, Mendelian Randomization; PRS,
polygenic risk score
Supplemental Table 4. Mendelian Randomization Estimates for the Effect of Genetically- Predicted FGF23 on Heart Failure and Heart Failure Subtypes in analyses excluding rs17216707 (CYP24A1) and rs11741640 (RGS14)
Outcome and data source Number of
HF cases Inverse Variance Weighted MR Odds Ratio (95%CI) P-value Heart Failure (HERMES)
Overall 47,309 1.98 (1.43, 2.74) <0.001
Heart Failure (BioVU)
Overall 18,415 1.50 (0.99, 2.27) 0.058
Low eGFR PRS
3,086 3.90 (1.40, 10.83) 0.009
High eGFR PRS
15,329 1.24 (0.79, 1.95) 0.354
P-interaction* 0.04
HF Preserved Ejection Fraction (BioVU)
Overall 13,141 1.75 (1.08, 2.84) 0.023
Low eGFR PRS 2,223 5.96 (1.82, 19.52) 0.003
High eGFR PRS 10,918 1.38 (0.81, 2.35) 0.230
P-interaction* 0.03
HF Reduced Ejection Fraction (BioVU)
Overall 3,394 0.79 (0.29, 2.20) 0.656
Low eGFR PRS 554 0.41 (0.03, 5.13) 0.488
High eGFR PRS 2,840 0.90 (0.29, 2.74) 0.850
P-interaction* 0.58
HF Mid-range Ejection Fraction (BioVU)
Overall 388 0.16 (0.01, 2.95) 0.215
Low eGFR PRS 70 4.77 (0.01, 4515) 0.655
High eGFR PRS 318 0.07 (0.01, 1.92) 0.117
P-interaction* 0.28
Low eGFR PRS: Estimated glomerular filtration rate polygenic risk score (eGFR PRS) less than one standard deviation below the mean.
High eGFR PRS: Estimated glomerular filtration rate polygenic risk score (eGFR PRS) greater than one standard deviation below the mean.
*P-interaction for the difference in the causal effect of genetically-predicted FGF23 on HF in low versus high eGFR PRS.
Abbreviations. IVW, inverse-variance weighted; MR, Mendelian Randomization; PRS,
polygenic risk score
References
1. Robinson-Cohen C, Bartz TM, Lai D, Ikizler TA, Peacock M, Imel EA, Michos ED, Foroud TM, Akesson K, Taylor KD, Malmgren L, Matsushita K, Nethander M, Eriksson J, Ohlsson C, Mellström D, Wolf M, Ljunggren O, McGuigan F, Rotter JI, Karlsson M, Econs MJ, Ix JH, Lutsey PL, Psaty BM, Boer IHd, Kestenbaum BR: Genetic Variants Associated with Circulating Fibroblast Growth Factor 23. JASN, 29: 2583-2592, 2018 10.1681/ASN.2018020192
2. Shah S, Henry A, Roselli C, Lin H, Sveinbjörnsson G, Fatemifar G, Hedman Å, Wilk JB, Morley MP, Chaffin MD, Helgadottir A, Verweij N, Dehghan A, Almgren P, Andersson C, Aragam KG, Ärnlöv J, Backman JD, Biggs ML, Bloom HL, Brandimarto J, Brown MR, Buckbinder L, Carey DJ, Chasman DI, Chen X, Chung J, Chutkow W, Cook JP, Delgado GE, Denaxas S, Doney AS, Dörr M, Dudley SC, Dunn ME, Engström G, Esko T, Felix SB, Finan C, Ford I, Ghanbari M, Ghasemi S, Giedraitis V, Giulianini F, Gottdiener JS, Gross S, Guðbjartsson DF, Gutmann R, Haggerty CM, van der Harst P, Hyde CL, Ingelsson E, Jukema JW, Kavousi M, Khaw KT, Kleber ME, Køber L,
Koekemoer A, Langenberg C, Lind L, Lindgren CM, London B, Lotta LA, Lovering RC, Luan J, Magnusson P, Mahajan A, Margulies KB, März W, Melander O, Mordi IR, Morgan T, Morris AD, Morris AP, Morrison AC, Nagle MW, Nelson CP, Niessner A, Niiranen T, O'Donoghue ML, Owens AT, Palmer CNA, Parry HM, Perola M, Portilla- Fernandez E, Psaty BM, Rice KM, Ridker PM, Romaine SPR, Rotter JI, Salo P, Salomaa V, van Setten J, Shalaby AA, Smelser DT, Smith NL, Stender S, Stott DJ, Svensson P, Tammesoo ML, Taylor KD, Teder-Laving M, Teumer A, Thorgeirsson G,
Thorsteinsdottir U, Torp-Pedersen C, Trompet S, Tyl B, Uitterlinden AG, Veluchamy A,
Völker U, Voors AA, Wang X, Wareham NJ, Waterworth D, Weeke PE, Weiss R, Wiggins KL, Xing H, Yerges-Armstrong LM, Yu B, Zannad F, Zhao JH, Hemingway H, Samani NJ, McMurray JJV, Yang J, Visscher PM, Newton-Cheh C, Malarstig A, Holm H, Lubitz SA, Sattar N, Holmes MV, Cappola TP, Asselbergs FW, Hingorani AD, Kuchenbaecker K, Ellinor PT, Lang CC, Stefansson K, Smith JG, Vasan RS, Swerdlow DI, Lumbers RT, Center RG: Genome-wide association and Mendelian randomisation analysis provide insights into the pathogenesis of heart failure. Nat Commun, 11: 163, 2020 10.1038/s41467-019-13690-5
3. Lumbers RT, Shah S, Lin H, Czuba T, Henry A, Swerdlow DI, Mälarstig A, Andersson C,
Verweij N, Holmes MV, Ärnlöv J, Svensson P, Hemingway H, Sallah N, Almgren P,
Aragam KG, Asselin G, Backman JD, Biggs ML, Bloom HL, Boersma E, Brandimarto J,
Brown MR, Brunner-La Rocca HP, Carey DJ, Chaffin MD, Chasman DI, Chazara O,
Chen X, Chen X, Chung JH, Chutkow W, Cleland JGF, Cook JP, de Denus S, Dehghan
A, Delgado GE, Denaxas S, Doney AS, Dörr M, Dudley SC, Engström G, Esko T,
Fatemifar G, Felix SB, Finan C, Ford I, Fougerousse F, Fouodjio R, Ghanbari M,
Ghasemi S, Giedraitis V, Giulianini F, Gottdiener JS, Gross S, Guðbjartsson DF, Gui H,
Gutmann R, Haggerty CM, van der Harst P, Hedman Å K, Helgadottir A, Hillege H,
Hyde CL, Jacob J, Jukema JW, Kamanu F, Kardys I, Kavousi M, Khaw KT, Kleber ME,
Køber L, Koekemoer A, Kraus B, Kuchenbaecker K, Langenberg C, Lind L, Lindgren
CM, London B, Lotta LA, Lovering RC, Luan J, Magnusson P, Mahajan A, Mann D,
Margulies KB, Marston NA, März W, McMurray JJV, Melander O, Melloni G, Mordi
IR, Morley MP, Morris AD, Morris AP, Morrison AC, Nagle MW, Nelson CP, Newton-
Cheh C, Niessner A, Niiranen T, Nowak C, O'Donoghue ML, Owens AT, Palmer CNA,
Paré G, Perola M, Perreault LL, Portilla-Fernandez E, Psaty BM, Rice KM, Ridker PM, Romaine SPR, Roselli C, Rotter JI, Ruff CT, Sabatine MS, Salo P, Salomaa V, van Setten J, Shalaby AA, Smelser DT, Smith NL, Stefansson K, Stender S, Stott DJ, Sveinbjörnsson G, Tammesoo ML, Tardif JC, Taylor KD, Teder-Laving M, Teumer A, Thorgeirsson G, Thorsteinsdottir U, Torp-Pedersen C, Trompet S, Tuckwell D, Tyl B, Uitterlinden AG, Vaura F, Veluchamy A, Visscher PM, Völker U, Voors AA, Wang X, Wareham NJ, Weeke PE, Weiss R, White HD, Wiggins KL, Xing H, Yang J, Yang Y, Yerges-Armstrong LM, Yu B, Zannad F, Zhao F, Wilk JB, Holm H, Sattar N, Lubitz SA, Lanfear DE, Shah S, Dunn ME, Wells QS, Asselbergs FW, Hingorani AD, Dubé MP, Samani NJ, Lang CC, Cappola TP, Ellinor PT, Vasan RS, Smith JG: The genomics of heart failure: design and rationale of the HERMES consortium. ESC heart failure, 8:
5531-5541, 2021 10.1002/ehf2.13517
