All studies published in Clinical Gastroenterology and Hepatology are embargoed until 3PM ET of the day they are published as corrected proofs on-line. Studies cannot be publicized as accepted manuscripts or uncorrected proofs.
Clinical Gastroenterology and Hepatology 2014;-:-–-
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
Risk of Upper and Lower Gastrointestinal Bleeding in Patients Taking Nonsteroidal Anti-inflammatory Drugs, Antiplatelet Agents, or Anticoagulants Q11
Ángel Lanas,*,‡ Patricia Carrera,‡ Yolanda Arguedas,* Santiago Garcia,§ Luis Bujanda,‡,k Xavier Calvet,‡,¶ Julio Ponce,‡,# Ángeles Perez-Aísa,** Manuel Castro,‡,‡‡ Maria Muñoz,* Carlos Sostres,* and Luis A. García-Rodríguez§§ *Servicio de Aparato Digestivo, Hospital Clínico, University of Zaragoza, Zaragoza, Spain; ‡CIBERehd, Barcelona, Spain; § Servicio de Aparato Digestivo, Hospital Universiatrio Miguel Servet, Zaragoza, Spain; kHospital Donostia, San Sebastian, Spain; ¶Servicio de Aparato Digestivo, Corporació Sanitària Universitària, Sabadell Barcelona, Spain; #Servicio de Aparato Digestivo, Hospital La Fe, Valencia, Spain; **Servicio de Aparato Digestivo, Hospital Costa del Sol, Marbella, Spain; ‡‡ Gastroenterología, Hospital de Valme, Sevilla, Spain; and §§Centro Español de Investigación Farmacoepidemiológica, Madrid, Spain
Q7
BACKGROUND & AIMS:
Treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) or low-dose aspirin is associated with increased risk of upper gastrointestinal bleeding. There is little evidence on the risk of lower gastrointestinal bleeding with NSAIDs, antiplatelet agents (APAs), or anticoagulants. We aimed to quantify the relative risk (RR) of upper and lower gastrointestinal bleeding associated with use of NSAIDs, APAs, or anticoagulants.
METHODS:
We performed a case-control study that used data collected from consecutive patients hospitalized for gastrointestinal bleeding (563 upper, mean age, 63.6 – 16.7 years and 415 lower, mean age, 70.8 – 13.8 years), confirmed by endoscopy or other diagnostic procedures. Unhospitalized patients were used as controls (n [ 1008) and matched for age, hospital, and month of admission. Drug use was considered current when taken within 7 days or less before hospitalization. RRs and 95% confidence intervals (CIs) were estimated by unconditional logistic regression analysis.
RESULTS:
Use of anticoagulants, low-dose aspirin, and other drugs (non-aspirin-APA, 82.3% thienopiridines) was associated with upper and lower gastrointestinal bleeding; the risk was 2-fold higher for anticoagulants (RR, 4.2; 95% CI, 2.9–6.2) than for low-dose aspirin (RR, 2.1; 95% CI, 1.4–3.3) or other non-aspirin-APA drugs (RR, 2.0; 95% CI, 1.6–2.6). NSAID use was also associated with increased risk of gastrointestinal bleeding and greater for upper (RR, 2.6; 95% CI, 2.0–3.5) than lower gastrointestinal bleeding (RR, 1.4; 95% CI, 1.0–1.9). Use of proton pump inhibitors was associated with reduced risk of upper, but not lower, gastrointestinal bleeding.
CONCLUSIONS:
Anticoagulants, low-dose aspirin, NSAIDs, and other non-aspirin-APA drugs are associated with increased risk of upper and lower gastrointestinal bleeding. Use of anticoagulants appears to be the strongest risk factor for gastrointestinal bleeding.
Keywords: Stomach; Intestine; Peptic Ulcer; Small Bowel; Colon; Side Effect; Complication.
t is well-established that nonsteroidal antiinflammatory drug (NSAID) treatment is associated with increased risk of peptic ulcer and non-variceal upper gastrointestinal bleeding (UGIB).1,2 However, growing evidence suggests that NSAIDs can also damage the lower gastrointestinal (GI) tract.3,4 Several studies have shown that NSAID use is associated with mucosal damage of the small bowel identified as mucosal breaks.5,6 Reports also suggest that NSAIDs can also damage the colon.7,8 New evidence is becoming available concerning small bowel or colonic damage
I
associated with aspirin use, but the clinical relevance of these lesions is still uncertain.9 Evidence on the risk of lower gastrointestinal bleeding (LGIB) associated Abbreviations used in this paper: APA, antiplatelet agent; ASA, low-dose aspirin; CI, confidence interval; GI, gastrointestinal; H2RA, H2 receptor antagonist; LGIB, lower gastrointestinal bleeding; NSAID, nonsteroidal anti-inflammatory drug; PPI, proton pump inhibitor; RR, relative risk; UGIB, upper gastrointestinal bleeding. © 2014 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2014.11.007
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
2
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
with other non-aspirin antiplatelet or anticoagulant agents is related only to diverticular bleeding and is also very scarce,10 whereas it is clear that use of these drugs is growing in a progressively elderly population.11 Different studies have shown that the use of proton pump inhibitors (PPIs) prevents upper GI damage and the risk of upper GI complications but is unable to prevent NSAID-associated small bowel mucosal damage.1,6 No data are available concerning the impact of PPI use on lower GI complications. During the last decade, there were reports showing a decreasing trend in hospitalizations that are due to upper GI complications, whereas lower GI complications showed a small but sustained increasing trend.12 Whether these changes are, at least in part, due to change in prescription habits (eg, higher PPI use) and a result of increased risk of LGIB associated with use of NSAIDs or low-dose aspirin (ASA), nonaspirin antiplatelet agents (APAs), or anticoagulant agents is not known. This study was designed to provide evidence on the risk of both upper and lower GI bleeding associated with NSAIDs, ASA, non-aspirin APAs, anticoagulants, and PPIs. By providing this information we should be able to evaluate simultaneously and compare the risk of bleeding associated with the use of these drugs in both the upper and the lower GI tract and provide some evidence to further explain current trends in hospitalizations that are due to GI complications, which should help to design potential prevention strategies in these patients.
Methods Study Design and Population Case-control study with prospective case ascertainment and data collection was carried out between 2009 and mid-2013. Cases and controls were collected through a network of general hospitals integrated within the Spanish Association of Gastroenterology and the Biomedical Investigation Network Center of hepatic and digestive diseases (CIBERehd). Overall, eligible participants were 20–90 years old with non-variceal GI bleeding who had been free of liver disease, coagulation disorders, or malignancies for the previous 5 years.
Definitions A case is a patient hospitalized because of GI bleeding (hematemesis, melena, hematochezia, or red blood per rectum), which was confirmed by hospital personnel. A bleeding event was considered to be UGIB event if hematemesis was observed by hospital staff and/or there was either blood in the stomach or a lesion with stigmata of bleeding was present at the time of the upper
-,
No.
-
GI endoscopy. A bleeding event was considered to be LGIB event when a lesion with stigmata was found below the angle of Treitz by endoscopic or radiologic procedures, or when no lesions were identified in the upper GI tract at the endoscopic procedure performed within 24 hours of emergency hospital admission and no hematemesis was reported or evidenced. All other events not complying with these definitions and without a clearly identifed site were considered as unspecified GI bleeding events. Cases with the following conditions were excluded: (1) bleeding caused by gastroesophageal or intestinal varices, GI cancer, Mallory–Weiss lesions, associated coagulopathy, and esophagitis; (2) patients with unreliable sources of information; (3) patients refusing to participate; and (4) in-hospital bleeding patients. Controls matched by age (5 years), gender, hospital, and month of admission were selected. Controls were obtained from people accompanying or visiting hospitalized patients. When the identification of a control under this condition was unsuccessful, unselected people referred to external general outpatient’s laboratory office for blood extraction (as part of routine general analysis) were used. For exposure, drug use was considered to be current when the drug was taken up to 7 days before the index date. It was considered to be past when drug use ended more than 1 week before the index date. Non-use was considered in individuals not reporting use. The index date for cases was the first day when the GI bleeding episode was objectively noticed, and for controls it was the day of interview. We analyzed the effects of individual NSAIDs among current single users. ASA was defined as any dose no greater than 300 mg/day. Non-aspirin APA drugs were clopidogrel, ticlopidine, dipyridamole, and trifusal. PPI use included all types of PPIs available on the market during the study period. All doses were considered to be PPI use. H2 receptor antagonist (H2RA) included any dose of ranitidine and famotidine, because cimetidine use was negligible. For GI disorder history, a person was defined as having no history of upper GI disorder if he/she reported no history of dyspepsia or ulcer (uncomplicated or complicated) before the index date. A person was defined as having a history of dyspepsia only if he or she did not also report a history of peptic ulcer. Finally, a person was defined as having a history of peptic ulcer without or with complications (bleeding or perforation) when he or she reported so and found or provided previous hospital reports that confirmed the diagnosis. All of these groups were mutually exclusive. In the same way, we classified cases and controls as having or not having a history of lower GI diseases (complicated and uncomplicated). A person was defined as having a history of lower GI complications if he/she reported bleeding or perforation history caused by lesion (GI cancers were excluded, as defined in the exclusion
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232
-
233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290
2014
Upper and Lower GI Bleeding and Toxic Drugs
criteria) located in the small bowel, colon, or rectum (not the anus) before the index date. In the same way, patients were considered to have a non-complicated lower GI history if they had been previously diagnosed with any non-complicated GI problem that was due to a visible organic lesion in the small bowel, colon, or rectum.
Data Collection Cases and controls were interviewed by the same persons (a gastroenterologist or gastroenterology trainee) in each participating center within 48 hours of admission (for cases). To ensure reliable data collection, cases and controls were accompanied during the interview by a relative or someone who lived with them. A structured questionnaire of marketed drugs and a careful review of prescriptions were used. Interview data were completed with the patient’s family and hospital clinical records concerning the bleeding episode. To avoid bias, interviewers were unaware of the actual objectives of the study. A committee was designated to respond to any questions arising during the study, and all data were introduced into a single database. Bleeding patients were managed according to each participating center protocols.
Statistical Analysis All data were introduced into a single database by one experienced data manager. Final data validation was performed twice by 2 investigators (Y.A., P.C.) who rechecked drug names, doses, and medical history. An initial exploratory analysis of all variables was carried out. Continuous variables were expressed as mean with standard deviation, whereas qualitative variables were expressed as frequencies and percentages. Bivariate analysis of variables was performed by c2 test for categorical variables and with t test for continuous data and was run before the construction of logistic regression models. For the overall analysis, we performed both unconditional and conditional regression (with the matching variables introduced in the model) and obtained similar results. Most reported analyses compute the relative risk (RR) of upper/lower GIB and the 95% confidence intervals (CIs). The final multivariate model included age, sex, hospital, calendar semester, GI history, smoking status, oral anticoagulants, APAs, acid-suppressing drugs, NSAIDs, coxibs, and aspirin. All the analyses were performed with SPSS 14.0 for Windows (SPSS Ibérica, Madrid, Spain). The sample size calculation and analysis were designed to collect information about drugs with low use (power, 80%; a value, 0.05) in the general population, which indicates that the sample size is adequate for most types of drugs reported in this study.
3
Table 1. Frequency and Crude Estimates of RRs and 95% CIs of Overall GI Bleeding Associated With Several Clinical Risk Factors
Variable UGI historya LGI historyb Dyspepsia history Current smoker Alcohol consumptionc
Controls (n ¼ 1008), n (%) 85 11 61 140 25
(8.4) (1.1) (6.1) (13.9) (2.5)
Cases (n ¼ 1008), n (%) 218 49 66 186 79
(21.6) (4.9) (6.5) (18.5) (7.8)
Crude RR (95% CI) 3.0 4.6 1.1 1.4 3.3
(2.3–3.9) (2.4–8.9) (0.8–1.6) (1.1–1.8) (2.1–5.3)
a
Includes both complicated and uncomplicated peptic ulcer history. Includes both complicated and uncomplicated lower GI lesion history. c Includes at least moderate alcohol consumption (more than 3 glasses per day of beer, wine, or liquor). b
Results Demographics and Clinical Characteristics of Patients A total of 1008 cases with GI bleeding and 1008 controls were collected from the participating centers. Men were 60.7% of cases and controls, and about 83.2% of the cases were aged >50 years. The mean age was 66.6 16.0 and 65.6 15.5 years in cases and controls, respectively. The corresponding figures for UGIB cases were 63.6 16.7 years and for LGIB cases were 70.8 13.8 years; also, 71.8% and 46.3% of patients, respectively, were men. Table 1 shows the frequency and crude estimate of RRs associated with factors usually considered associated with GI bleeding or with potential confounders that were included in the logistic regression models. In 563 of cases (55.9%), the bleeding lesion was located in the upper GI tract, whereas in 415 cases (41.2%), the bleeding was considered to be located in the lower GI tract. In the remaining 30 cases (3.0%), a location could not be determined. Table 2 summarizes the main lesions identified in cases with either upper or lower GI bleeding. The most common lesion of UGIB events was duodenal ulcer (41.9%), whereas the most frequent one in LGIB events was diverticular lesion (39.3%). Tests for Helicobacter pylori infection were noted in 78.5% of UGIB cases, with a positive result in 62.9%. In lesions from the lower GI tract, stigmata of bleeding were found in 342 cases (82.4%), whereas in the remaining cases no stigmata was reported.
Drug Use in Upper and Lower Gastrointestinal Bleeders and Controls Overall, 256 patients (24.3%) had taken at least 1 NSAID in the week before the first evidence of GI bleeding event compared with 16.8% among controls
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348
4
349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
Table 2. Causes of Upper and Lower GI Bleeding n (%) UGIB cases (n ¼ 563) Gastric bleeding due to duodenal ulcer Gastric bleeding due to gastric ulcer Q9 Gastric bleeding due to acute gastroduodenal Vascular gastroduodenal lesions Gastric polyps LGIB cases (n ¼ 415) Diverticular lesions Intestinal bleeding (small/large bowel) No specified Ischemic colitis Vascular lesions Inflammatory colitis Other causes Erosive lesions Large bowel polyps Hemorrhoidal or anal bleeding (hemoglobin drop >2 g/dL)
236 180 123 20 4
(41.9) (32.0) (21.8) (3.6) (0.7)
163 (39.3) 73 (17.6) 49 34 30 26 19 13 8
(11.8) (8.2) (7.2) (6.3) (4.6) (3.1) (1.8)
before the interview (P < .05). A similar percentage could be observed in patients and controls concerning ASA use. Table 3 summarizes current drug use in cases and controls. In addition to NSAID and ASA use, cases also had taken other APAs (mostly clopidogrel) and anticoagulants more frequently, but no differences were found in PPI use. The most frequent NSAID used was ibuprofen (12.3% in cases vs 9.8% in controls), followed by diclofenac (4.5% vs 2.0%), high-dose aspirin (5.0 vs 0.8%), naproxen (1.6% vs 1.5%), and coxibs (1.3% vs 1.4%). The crude and adjusted RRs for each individual NSAID are reported in Supplementary Table 1. The mean dose used for ibuprofen was 909.8 465.4 mg/day,
Table 3. Overall Frequency, Crude Estimates of RRs and 95% CIs of GI Bleeding Associated With Current Drug Use
Variable NSAIDs ASA PPI H2RA Anticoagulants NSAID or ASA All APA Non-ASA APAa
Controls (n ¼ 1008), n (%) 169 175 350 11 58 318 202 57
(16.8) (17.4) (34.7) (1.1) (5.8) (31.5) (20.0) (5.7)
Cases (n ¼ 1008), n (%) 256 247 381 23 156 458 304 101
(25.4) (24.5) (37.8) (2.3) (15.5) (45.4) (30.2) (10.0)
Crude RR (95% CI) 1.7 1.5 1.1 2.1 3.0 1.8 1.7 1.9
(1.4–2.1) (1.2–1.9) (1.0–1.4) (1.0–4.4) (2.2–4.1) (1.5–2.2) (1.4–2.1) (1.3–2.6)
a OR and 95% CI for clopidogrel (82.3% of this group) were 1.8 (1.3–2.7); 26 controls (2.6%) and 45 cases (4.5%) were taking a combination of ASA and NSAIDs (RR, 1.7; 1.1.–2.9); 29 controls and 44 cases were taking a combination of non-ASA APA together with ASA (RR, 1.5; 1.0–2.5); 9 controls and 11 cases were taking a combination of non-ASA APA together with NSAIDs (RR, 1.2; 0.5–3.0).
Figure 1. Unconditional adjusted estimates of RRs and 95% CIs of (A) UGIB or (B) LGIB associated with current drug use. RRs were adjusted for hospital, age, gender, calendar semester, GI lesion history, dyspepsia history, smoking status, alcohol consumption, anticoagulants, PPI, non-aspirin APA use, H2RA, NSAID, and ASA use.
diclofenac was 124.7 287.3 mg/day, high-dose aspirin was 595.4 312.3 mg/day, and naproxen was 684.5 323.0 mg/day. The crude RR associated with high (>1200 mg/day) vs low dose of ibuprofen, the most frequent NSAID used in our study, was 3.9 (95% CI, 1.5–10.0). Figure 1 shows the unconditional adjusted RR of UGIB and LGIB associated with different drug use. Overall, NSAID use was associated with increased risk of upper and lower GI bleeding, although the estimate was 2 times higher for the upper compared with the lower GI tract. When the risk of upper and lower GI bleeding was divided by type of NSAID, the increased risk was present for ibuprofen, high-dose aspirin, and diclofenac in the upper GI tract but not in the lower GI tract (Supplementary Table 1). For other individual NSAIDs or coxibs, the low frequency of use limited the power of these analyses. A similar conclusion can be reached for the risk of LGIB associated with these NSAIDs. ASA use was also associated similarly with increased risk of both upper and lower GI bleeding. A similar conclusion can be reached concerning the risk associated with other non-aspirin APAs. We conducted sensitivity analysis excluding patients with bleeding that was due to ischemic colitis or inflammatory colitis, because use of NSAID or ASA might introduce some confounding by indication effect in these patients; however, the analysis
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
-
465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522
2014
Upper and Lower GI Bleeding and Toxic Drugs
did not show any significant change in the results (data not shown). In this study, the highest risk associated with both upper and lower GI bleeding was observed with anticoagulant use. The estimates were similar for lower and upper GI bleeding. Supplementary Table 2 shows the frequency of drug use in cases and controls and the crude and conditional adjusted RRs for that use in UGIB and LGIB patients. Current PPI use was present in 381 cases (37.8%) and in 350 controls (34.7%), but the distribution of use was different in cases with upper and lower GI bleeding events. PPI use was associated with decreased risk of UGIB (adjusted RR, 0.4; 95% CI, 0.3–0.5), whereas it was associated with a small increased risk of LGIB (adjusted RR, 1.5; 95% CI, 1.1–1.9). We have also analyzed the effects of PPI combined with these drugs. As shown in Supplementary Table 3, the combination of NSAIDs, ASA, other APAs, and anticoagulants with PPI was not associated with upper GI bleeding.
Discussion It is well-established that NSAIDs and ASA are associated with increased risk of UGIB,1,2,13,14 but data on their effects on the risk of bleeding in lower GI tract are less clear, although the evidence has been growing in recent years.8–10 Also, the effect of non-aspirin APA drugs and anticoagulants on the risk of LGIB is much less explored. This study has evaluated the risk of GI bleeding associated with these compounds and compares the risks of either upper or lower GI bleeding in the same study within the same source population and same temporal setting. We have observed that NSAIDs are associated with increased risk of both upper and lower GI bleeding and that the risk is clearly higher in the upper GI tract. Different animal and healthy volunteer studies have shown that NSAIDs can damage the mucosa of the small intestine and even the colon,3–7 but the clinical implications of those findings were unclear. A few clinical trials conducted in patients with different rheumatic conditions have shown a higher than expected incidence of lower GI complications with naproxen, diclofenac, or etoricoxib (MEDAL and VIGOR studies).15,16 The CONDOR17 and GI-REASON18 trials have shown that patients taking NSAIDs with/without PPI have a higher incidence of GI events (basically hemoglobin drops without clinically overt bleeding) compared with celecoxib, when taking into consideration both the upper and lower GI tract together. However, none of these studies have provided a clear estimate of the risk of major LGIB compared with non-NSAID use. Our study suggests that the increased risk of LGIB is lower than that seen for UGIB. In our population, NSAID use was mainly due to ibuprofen and diclofenac, whereas other compounds had much less use. On the basis of these data, we would
5
suggest that our overall estimates of risk for NSAIDs are driven mainly by ibuprofen (for the most part at analgesic doses). The lack of increased risk of GI bleeding for naproxen or coxibs has to be interpreted with caution because of the large variability of the estimates of risk. Another interesting finding of this study was that ASA was associated with a similar risk of upper and lower GI bleeding. Because there are previous studies that suggest that even ASA can induce mucosal damage in both the upper and lower GI tract,19 the effect on GI bleeding with this compound cannot rule out that the bleeding was due to the induction of mucosal damage and subsequent bleeding. However, ASA could have also induced bleeding from preexisting lesions (eg, angiodysplasia, diverticula). This antiplatelet mechanism is reinforced by the fact that non-ASA APA use (mostly clopidogrel) was associated with UGIB and LGIB, and that this increased risk was similar to that observed with ASA. Because these compounds have not been found to induce mucosal damage, the effect on bleeding should be mainly mediated by its antiplatelet effect. Anticoagulants (dicumarinics) were associated with a risk of UGIB and LGIB greater than that observed with antiplatelet drugs and NSAIDs. These compounds do not affect directly the GI mucosa but have a higher antihemostatic effect compared with antiplatelet drugs. We believe that this is a significant finding because the proportion of GI bleeders associated with anticoagulant use is growing, which may be a significant problem in the future, because the elderly population continues to increase in Western countries. Furthermore, PPI co-therapy, which decreases the risk of mucosal damage and complications in the gastroduodenal mucosa, will also have no protective effect on the lower GI tract. We have confirmed that PPI use was associated with >50% risk reduction in UGIB, but as expected, it was not found to reduce the risk of LGIB. On the contrary, we found a small increase in that risk with PPI use. We did not interpret this finding as a direct effect of PPI on the lower GI tract but as an effect of residual “confounding by indication,” despite controlling for a number of covariates. Yet, PPI use is usually a strong marker of worse overall comorbidity.11 Nevertheless, new studies should further investigate this effect, because recent experimental animal data suggest that PPI can alter the microbiota of the small bowel and potentiate the damaging effect of NSAIDs on the intestinal mucosa.20 This study has strengths and weaknesses. Among the strengths we include the relatively high number of patients included in the study and the fact that all were consecutive patients with documented upper or lower GI bleeding, which permitted the evaluation of the risk of upper and lower GI bleeding in the same cohort while obtaining comparable data. All bleeding cases were investigated, and the data were prospectively collected. The use of over-the-counter drugs by cases and controls was also collected. Among the limitations, the study
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580
6
581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
design (interview-based) could have introduced selection bias. The recollection of data, including the temporal use and dose of the drugs, is subject to bias because patients may not recall this properly in a life-threatening situation. Also, no additional validation of drug use was performed by using computerized databases. Although some of these problems can be solved or minimized by collecting the information with family members, the absence of prescription data at primary care level and pharmacy data may have affected the actual drug use. Bias cannot be completely prevented, especially when some underlying condition can be linked both to exposure and to the event of interest. Also, the sample size of the study did not allow providing reliable odds ratios for NSAIDs with low frequency of use and by dose, with exception made for ibuprofen. Finally, the controls used may not reflect the ideal population, which may affect the differences of drug use between cases and controls. In summary, our study provides evidence that NSAIDs, ASA, non-ASA APAs, and anticoagulants increase the risk of both upper and lower GI bleeding. The magnitude of this increase is similar for upper and lower GI tract for ASA, other APA, and anticoagulant use, whereas this increase in the risk seems to be higher in the upper compared with the lower GI tract for NSAID use, which may suggest different mechanisms of action. These results may have clinical implications, especially for prevention strategies. The risk of UGIB in susceptible patients can be reduced with PPI co-prescription, but this is not the case for the risk of LGIB. Because NSAID, antithrombotic, and antiplatelet therapy is growing in a progressively older population, special caution should be placed in those patients with a history of lower GI disease. We suggest that because of growing evidence in the field, current guidelines should be reconsidered, taking into consideration not only the upper but also the lower GI tract and not only NSAIDs and ASA but also include non-aspirin APAs and anticoagulant therapy.
Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org, and at http://dx.doi.org/10.1016/j.cgh.2014.11.007.
References 1. Lanza FL, Chan FK, Quigley EM. Guidelines for prevention of NSAID-related ulcer complications. Am J Gastroenterol 2009; 104:728–738. 2. García Rodríguez LA, Barreales Tolosa L. Risk of upper gastrointestinal complications among users of traditional NSAIDs and COXIBs in the general population. Gastroenterology 2007;132:498–506. 3. Bjarnason I, Hayllar J, Macpherson AJ, et al. Side effects of nonsteroidal anti–inflammatory drugs on the small and large intestine in humans. Gastroenterology 1993;104:1832–1847.
-,
No.
-
4. Graham DY, Opekun AR, Willingham FF, et al. Visible small– intestinal mucosal injury in chronic NSAID users. Clin Gastroenterol Hepatol 2005;3:55–59. 5. Maiden L, Thjodleifsson B, Seigal A, et al. Long-term effects of nonsteroidal anti-inflammatory drugs and cyclooxygenase-2 selective agents on the small bowel: a cross-sectional capsule enteroscopy study. Clin Gastroenterol Hepatol 2007;5: 1040–1045. 6. Goldstein JL, Eisen GM, Lewis B, et al. Small bowel mucosal injury is reduced in healthy subjects treated with celecoxib compared with ibuprofen plus omeprazole, as assessed by video capsule endoscopy. Aliment Pharmacol Ther 2007; 25:1211–1222. 7. Kurahara K, Matsumoto T, Lida M, et al. Clinical and endoscopic features of nonsteroidal anti-inflammatory drug-induced colonic ulcerations. Am J Gastroenterol 2001;96:473–480. 8. Lanas A, Sekar MC, Hirschowitz BI. Objective evidence of aspirin use in both ulcer and non–ulcer upper and lower gastrointestinal bleeding. Gastroenterology 1992;103:862–869. 9. Yuhara H, Corley DA, Nakahara F, et al. Aspirin and non-aspirin NSAIDs increase risk of colonic diverticular bleeding: a systematic review and meta-analysis. J Gastroenterol 2014;49: 992–1000. 10. Nagata N, Niikura R, Aoki T, et al. Colonic diverticular hemorrhage associated with the use of NSAIDs, low-dose aspirin, antiplatelet drugs, and dual therapy. J Gastroenterol Hepatol Q8 (in press). 11. Abraham NS, Hartman C, Richardson P, et al. Risk of lower and upper gastrointestinal bleeding, transfusions, and hospitalizations with complex antithrombotic therapy in elderly patients. Circulation 2013;128:1869–1877. 12. Lanas A, García-Rodríguez LA, Polo-Tomás M, et al. Time trends and impact of upper and lower gastrointestinal bleeding and perforation in clinical practice. Am J Gastroenterol 2009; 104:1633–1641. 13. Antithrombotic Trialists’ (ATT) Collaboration. Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials. Lancet 2009;373:1849–1860. 14. Sostres C, Lanas A. Gastrointestinal effects of aspirin. Nat Rev Gastroenterol Hepatol 2011;8:385–394. 15. Laine L, Curtis SP, Cryer B, et al. MEDAL Steering Committee: assessment of upper gastrointestinal safety of etoricoxib and diclofenac in patients with osteo-arthritis and rheumatoid arthritis in the Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) programme—a randomised comparison. Lancet 2007;369:465–473. 16. Laine L, Connors LG, Reicin A, et al. Serious lower gastrointestinal clinical events with nonselective NSAID or coxib use. Gastroenterology 2003;124:288–292. 17. Chan FK, Lanas A, Scheiman J, et al. Celecoxib versus omeprazole and diclofenac in patients with osteoarthritis and rheumatoid arthritis (CONDOR): a randomised trial. Lancet 2010; 376:173–179. 18. Cryer B, Li C, Simon LS, et al. GI-REASONS: a novel 6-month, prospective, randomized, open-label, blinded endpoint (PROBE) trial. Am J Gastroenterol 2013;108:392–400. 19. Smecuol E, Pinto-Sanchez MI, Suarez AA, et al. Low-dose aspirin affects the small bowel mucosa: results of a pilot study with a multidimensional assessment. Clin Gastroenterol Hepatol 2009;7:524–529.
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696
-
2014
Upper and Lower GI Bleeding and Toxic Drugs
20. Wallace JL, Syer S, Denou E, et al. Proton pump inhibitors 697 exacerbate NSAID-induced small intestinal injury by inducing 698 dysbiosis. Gastroenterology 2011;141:1314–1322. 699 700 Reprint requests 701 Address requests for reprints to: Angel Lanas, Servicio de Aparato Digestivo 702 Q1 Q2 Hospital Clínico Universitario, Zaragoza, Spain. e-mail: angel.lanas@gmail. com; fax: 34-976-765787. 703 704 Q3 Acknowledgments 705 Q4 The authors thank the following people who helped in their work: M. A Aranguren, Hospital Donostia, San Sebastian, Spain; Judith Millastre, Hospital 706 Universitario Miguel Servet, Zaragoza, Spain; Marta Gracia, Hospital Universitario Miguel Servet, Zaragoza, Spain; Ana Chaves, Hospital Costa del Sol, 707 Marbella, Spain; Marta Ponce, Hospital La Fe, Valencia, Spain; Meritxell 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
7
Sánchez-Lloansi, Corporación Sanitaria Universitaria, Sabadell, Spain; Pedro Lacarta, Hospital Clínico Universitario, Zaragoza, Spain; and Alberto Lúe, Hospital Clínico Universitario, Zaragoza, Spain. Conflicts of interest These authors disclose the following: L. A. García Rodríguez has received research funding from AstraZeneca R&D and Bayer and has also received honoraria for serving on scientific advisory boards for Bayer. The institution of Angel Lanas has received funds from Bayer from investigator-initiated research and Angel Lanas has been advisor to Pfizer and Bayer and received speaking fees from Bayer and AstraZeneca. The remaining authors disclose no conflicts. Q5 Funding The study was funded by the Official Spanish Agency, Instituto de Salud Carlos III-FIS, grant PI08/1301, and the CIBERehd, which is supporting the salary of Q6 Patricia Carrera.
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812
7.e1
813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
Supplementary Table 1. Crude and Conditional Adjusted RRs of UGIB and LGIB by Type of NSAID Use Overall Ibuprofen Aspirin (>300 mg) Diclofenac Naproxen Dexketoprofen Other NSAIDs Coxibs
UGIB Ibuprofen Aspirin (>300 mg) Diclofenac Naproxen Dexketoprofen Other NSAIDs Coxibs
LGIB Ibuprofen Aspirin (>300 mg) Diclofenac Naproxen Dexketoprofen Other NSAIDs Coxibs
Controls (n ¼ 1008), n (%) 99 8 20 15 7 12 14
(9.8) (0.8) (2.0) (1.5) (0.7) (1.2) (1.4)
Controls (n ¼ 563), n (%) 49 4 7 8 1 6 8
(8.7) (0.7) (1.2) (1.4) (0.2) (1.1) (1.4)
Controls (n ¼ 415), n (%) 49 3 13 6 5 6 6
(11.8) (0.7) (3.1) (1.4) (1.2) (1.4) (1.4)
Cases (n ¼ 1008), n (%) 124 50 45 16 11 16 13
(12.3) (5.0) (4.5) (1.6) (1.1) (1.6) (1.3)
Cases (n ¼ 563), n (%) 84 42 28 8 6 6 5
(14.9) (7.5) (5.0) (1.4) (1.1) (1.1) (0.9)
Cases (n ¼ 415), n (%) 36 8 17 7 5 10 7
(8.7) (1.9) (4.1) (1.7) (1.2) (2.4) (1.7)
Total 223 58 65 31 18 28 27
(11.1) (2.9) (3.2) (1.5) (0.9) (1.4) (1.3)
Total 133 46 35 16 7 12 13
(11.8) (4.1) (3.1) (1.4) (0.6) (1.1) (1.2)
Total 85 11 30 13 10 16 13
(10.2) (1.3) (3.6) (1.6) (1.2) (1.9) (1.6)
P value .088
Clinical Gastroenterology and Hepatology 2014;-:-–-
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
Risk of Upper and Lower Gastrointestinal Bleeding in Patients Taking Nonsteroidal Anti-inflammatory Drugs, Antiplatelet Agents, or Anticoagulants Q11
Ángel Lanas,*,‡ Patricia Carrera,‡ Yolanda Arguedas,* Santiago Garcia,§ Luis Bujanda,‡,k Xavier Calvet,‡,¶ Julio Ponce,‡,# Ángeles Perez-Aísa,** Manuel Castro,‡,‡‡ Maria Muñoz,* Carlos Sostres,* and Luis A. García-Rodríguez§§ *Servicio de Aparato Digestivo, Hospital Clínico, University of Zaragoza, Zaragoza, Spain; ‡CIBERehd, Barcelona, Spain; § Servicio de Aparato Digestivo, Hospital Universiatrio Miguel Servet, Zaragoza, Spain; kHospital Donostia, San Sebastian, Spain; ¶Servicio de Aparato Digestivo, Corporació Sanitària Universitària, Sabadell Barcelona, Spain; #Servicio de Aparato Digestivo, Hospital La Fe, Valencia, Spain; **Servicio de Aparato Digestivo, Hospital Costa del Sol, Marbella, Spain; ‡‡ Gastroenterología, Hospital de Valme, Sevilla, Spain; and §§Centro Español de Investigación Farmacoepidemiológica, Madrid, Spain
Q7
BACKGROUND & AIMS:
Treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) or low-dose aspirin is associated with increased risk of upper gastrointestinal bleeding. There is little evidence on the risk of lower gastrointestinal bleeding with NSAIDs, antiplatelet agents (APAs), or anticoagulants. We aimed to quantify the relative risk (RR) of upper and lower gastrointestinal bleeding associated with use of NSAIDs, APAs, or anticoagulants.
METHODS:
We performed a case-control study that used data collected from consecutive patients hospitalized for gastrointestinal bleeding (563 upper, mean age, 63.6 – 16.7 years and 415 lower, mean age, 70.8 – 13.8 years), confirmed by endoscopy or other diagnostic procedures. Unhospitalized patients were used as controls (n [ 1008) and matched for age, hospital, and month of admission. Drug use was considered current when taken within 7 days or less before hospitalization. RRs and 95% confidence intervals (CIs) were estimated by unconditional logistic regression analysis.
RESULTS:
Use of anticoagulants, low-dose aspirin, and other drugs (non-aspirin-APA, 82.3% thienopiridines) was associated with upper and lower gastrointestinal bleeding; the risk was 2-fold higher for anticoagulants (RR, 4.2; 95% CI, 2.9–6.2) than for low-dose aspirin (RR, 2.1; 95% CI, 1.4–3.3) or other non-aspirin-APA drugs (RR, 2.0; 95% CI, 1.6–2.6). NSAID use was also associated with increased risk of gastrointestinal bleeding and greater for upper (RR, 2.6; 95% CI, 2.0–3.5) than lower gastrointestinal bleeding (RR, 1.4; 95% CI, 1.0–1.9). Use of proton pump inhibitors was associated with reduced risk of upper, but not lower, gastrointestinal bleeding.
CONCLUSIONS:
Anticoagulants, low-dose aspirin, NSAIDs, and other non-aspirin-APA drugs are associated with increased risk of upper and lower gastrointestinal bleeding. Use of anticoagulants appears to be the strongest risk factor for gastrointestinal bleeding.
Keywords: Stomach; Intestine; Peptic Ulcer; Small Bowel; Colon; Side Effect; Complication.
t is well-established that nonsteroidal antiinflammatory drug (NSAID) treatment is associated with increased risk of peptic ulcer and non-variceal upper gastrointestinal bleeding (UGIB).1,2 However, growing evidence suggests that NSAIDs can also damage the lower gastrointestinal (GI) tract.3,4 Several studies have shown that NSAID use is associated with mucosal damage of the small bowel identified as mucosal breaks.5,6 Reports also suggest that NSAIDs can also damage the colon.7,8 New evidence is becoming available concerning small bowel or colonic damage
I
associated with aspirin use, but the clinical relevance of these lesions is still uncertain.9 Evidence on the risk of lower gastrointestinal bleeding (LGIB) associated Abbreviations used in this paper: APA, antiplatelet agent; ASA, low-dose aspirin; CI, confidence interval; GI, gastrointestinal; H2RA, H2 receptor antagonist; LGIB, lower gastrointestinal bleeding; NSAID, nonsteroidal anti-inflammatory drug; PPI, proton pump inhibitor; RR, relative risk; UGIB, upper gastrointestinal bleeding. © 2014 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2014.11.007
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
2
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
with other non-aspirin antiplatelet or anticoagulant agents is related only to diverticular bleeding and is also very scarce,10 whereas it is clear that use of these drugs is growing in a progressively elderly population.11 Different studies have shown that the use of proton pump inhibitors (PPIs) prevents upper GI damage and the risk of upper GI complications but is unable to prevent NSAID-associated small bowel mucosal damage.1,6 No data are available concerning the impact of PPI use on lower GI complications. During the last decade, there were reports showing a decreasing trend in hospitalizations that are due to upper GI complications, whereas lower GI complications showed a small but sustained increasing trend.12 Whether these changes are, at least in part, due to change in prescription habits (eg, higher PPI use) and a result of increased risk of LGIB associated with use of NSAIDs or low-dose aspirin (ASA), nonaspirin antiplatelet agents (APAs), or anticoagulant agents is not known. This study was designed to provide evidence on the risk of both upper and lower GI bleeding associated with NSAIDs, ASA, non-aspirin APAs, anticoagulants, and PPIs. By providing this information we should be able to evaluate simultaneously and compare the risk of bleeding associated with the use of these drugs in both the upper and the lower GI tract and provide some evidence to further explain current trends in hospitalizations that are due to GI complications, which should help to design potential prevention strategies in these patients.
Methods Study Design and Population Case-control study with prospective case ascertainment and data collection was carried out between 2009 and mid-2013. Cases and controls were collected through a network of general hospitals integrated within the Spanish Association of Gastroenterology and the Biomedical Investigation Network Center of hepatic and digestive diseases (CIBERehd). Overall, eligible participants were 20–90 years old with non-variceal GI bleeding who had been free of liver disease, coagulation disorders, or malignancies for the previous 5 years.
Definitions A case is a patient hospitalized because of GI bleeding (hematemesis, melena, hematochezia, or red blood per rectum), which was confirmed by hospital personnel. A bleeding event was considered to be UGIB event if hematemesis was observed by hospital staff and/or there was either blood in the stomach or a lesion with stigmata of bleeding was present at the time of the upper
-,
No.
-
GI endoscopy. A bleeding event was considered to be LGIB event when a lesion with stigmata was found below the angle of Treitz by endoscopic or radiologic procedures, or when no lesions were identified in the upper GI tract at the endoscopic procedure performed within 24 hours of emergency hospital admission and no hematemesis was reported or evidenced. All other events not complying with these definitions and without a clearly identifed site were considered as unspecified GI bleeding events. Cases with the following conditions were excluded: (1) bleeding caused by gastroesophageal or intestinal varices, GI cancer, Mallory–Weiss lesions, associated coagulopathy, and esophagitis; (2) patients with unreliable sources of information; (3) patients refusing to participate; and (4) in-hospital bleeding patients. Controls matched by age (5 years), gender, hospital, and month of admission were selected. Controls were obtained from people accompanying or visiting hospitalized patients. When the identification of a control under this condition was unsuccessful, unselected people referred to external general outpatient’s laboratory office for blood extraction (as part of routine general analysis) were used. For exposure, drug use was considered to be current when the drug was taken up to 7 days before the index date. It was considered to be past when drug use ended more than 1 week before the index date. Non-use was considered in individuals not reporting use. The index date for cases was the first day when the GI bleeding episode was objectively noticed, and for controls it was the day of interview. We analyzed the effects of individual NSAIDs among current single users. ASA was defined as any dose no greater than 300 mg/day. Non-aspirin APA drugs were clopidogrel, ticlopidine, dipyridamole, and trifusal. PPI use included all types of PPIs available on the market during the study period. All doses were considered to be PPI use. H2 receptor antagonist (H2RA) included any dose of ranitidine and famotidine, because cimetidine use was negligible. For GI disorder history, a person was defined as having no history of upper GI disorder if he/she reported no history of dyspepsia or ulcer (uncomplicated or complicated) before the index date. A person was defined as having a history of dyspepsia only if he or she did not also report a history of peptic ulcer. Finally, a person was defined as having a history of peptic ulcer without or with complications (bleeding or perforation) when he or she reported so and found or provided previous hospital reports that confirmed the diagnosis. All of these groups were mutually exclusive. In the same way, we classified cases and controls as having or not having a history of lower GI diseases (complicated and uncomplicated). A person was defined as having a history of lower GI complications if he/she reported bleeding or perforation history caused by lesion (GI cancers were excluded, as defined in the exclusion
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232
-
233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290
2014
Upper and Lower GI Bleeding and Toxic Drugs
criteria) located in the small bowel, colon, or rectum (not the anus) before the index date. In the same way, patients were considered to have a non-complicated lower GI history if they had been previously diagnosed with any non-complicated GI problem that was due to a visible organic lesion in the small bowel, colon, or rectum.
Data Collection Cases and controls were interviewed by the same persons (a gastroenterologist or gastroenterology trainee) in each participating center within 48 hours of admission (for cases). To ensure reliable data collection, cases and controls were accompanied during the interview by a relative or someone who lived with them. A structured questionnaire of marketed drugs and a careful review of prescriptions were used. Interview data were completed with the patient’s family and hospital clinical records concerning the bleeding episode. To avoid bias, interviewers were unaware of the actual objectives of the study. A committee was designated to respond to any questions arising during the study, and all data were introduced into a single database. Bleeding patients were managed according to each participating center protocols.
Statistical Analysis All data were introduced into a single database by one experienced data manager. Final data validation was performed twice by 2 investigators (Y.A., P.C.) who rechecked drug names, doses, and medical history. An initial exploratory analysis of all variables was carried out. Continuous variables were expressed as mean with standard deviation, whereas qualitative variables were expressed as frequencies and percentages. Bivariate analysis of variables was performed by c2 test for categorical variables and with t test for continuous data and was run before the construction of logistic regression models. For the overall analysis, we performed both unconditional and conditional regression (with the matching variables introduced in the model) and obtained similar results. Most reported analyses compute the relative risk (RR) of upper/lower GIB and the 95% confidence intervals (CIs). The final multivariate model included age, sex, hospital, calendar semester, GI history, smoking status, oral anticoagulants, APAs, acid-suppressing drugs, NSAIDs, coxibs, and aspirin. All the analyses were performed with SPSS 14.0 for Windows (SPSS Ibérica, Madrid, Spain). The sample size calculation and analysis were designed to collect information about drugs with low use (power, 80%; a value, 0.05) in the general population, which indicates that the sample size is adequate for most types of drugs reported in this study.
3
Table 1. Frequency and Crude Estimates of RRs and 95% CIs of Overall GI Bleeding Associated With Several Clinical Risk Factors
Variable UGI historya LGI historyb Dyspepsia history Current smoker Alcohol consumptionc
Controls (n ¼ 1008), n (%) 85 11 61 140 25
(8.4) (1.1) (6.1) (13.9) (2.5)
Cases (n ¼ 1008), n (%) 218 49 66 186 79
(21.6) (4.9) (6.5) (18.5) (7.8)
Crude RR (95% CI) 3.0 4.6 1.1 1.4 3.3
(2.3–3.9) (2.4–8.9) (0.8–1.6) (1.1–1.8) (2.1–5.3)
a
Includes both complicated and uncomplicated peptic ulcer history. Includes both complicated and uncomplicated lower GI lesion history. c Includes at least moderate alcohol consumption (more than 3 glasses per day of beer, wine, or liquor). b
Results Demographics and Clinical Characteristics of Patients A total of 1008 cases with GI bleeding and 1008 controls were collected from the participating centers. Men were 60.7% of cases and controls, and about 83.2% of the cases were aged >50 years. The mean age was 66.6 16.0 and 65.6 15.5 years in cases and controls, respectively. The corresponding figures for UGIB cases were 63.6 16.7 years and for LGIB cases were 70.8 13.8 years; also, 71.8% and 46.3% of patients, respectively, were men. Table 1 shows the frequency and crude estimate of RRs associated with factors usually considered associated with GI bleeding or with potential confounders that were included in the logistic regression models. In 563 of cases (55.9%), the bleeding lesion was located in the upper GI tract, whereas in 415 cases (41.2%), the bleeding was considered to be located in the lower GI tract. In the remaining 30 cases (3.0%), a location could not be determined. Table 2 summarizes the main lesions identified in cases with either upper or lower GI bleeding. The most common lesion of UGIB events was duodenal ulcer (41.9%), whereas the most frequent one in LGIB events was diverticular lesion (39.3%). Tests for Helicobacter pylori infection were noted in 78.5% of UGIB cases, with a positive result in 62.9%. In lesions from the lower GI tract, stigmata of bleeding were found in 342 cases (82.4%), whereas in the remaining cases no stigmata was reported.
Drug Use in Upper and Lower Gastrointestinal Bleeders and Controls Overall, 256 patients (24.3%) had taken at least 1 NSAID in the week before the first evidence of GI bleeding event compared with 16.8% among controls
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348
4
349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
Table 2. Causes of Upper and Lower GI Bleeding n (%) UGIB cases (n ¼ 563) Gastric bleeding due to duodenal ulcer Gastric bleeding due to gastric ulcer Q9 Gastric bleeding due to acute gastroduodenal Vascular gastroduodenal lesions Gastric polyps LGIB cases (n ¼ 415) Diverticular lesions Intestinal bleeding (small/large bowel) No specified Ischemic colitis Vascular lesions Inflammatory colitis Other causes Erosive lesions Large bowel polyps Hemorrhoidal or anal bleeding (hemoglobin drop >2 g/dL)
236 180 123 20 4
(41.9) (32.0) (21.8) (3.6) (0.7)
163 (39.3) 73 (17.6) 49 34 30 26 19 13 8
(11.8) (8.2) (7.2) (6.3) (4.6) (3.1) (1.8)
before the interview (P < .05). A similar percentage could be observed in patients and controls concerning ASA use. Table 3 summarizes current drug use in cases and controls. In addition to NSAID and ASA use, cases also had taken other APAs (mostly clopidogrel) and anticoagulants more frequently, but no differences were found in PPI use. The most frequent NSAID used was ibuprofen (12.3% in cases vs 9.8% in controls), followed by diclofenac (4.5% vs 2.0%), high-dose aspirin (5.0 vs 0.8%), naproxen (1.6% vs 1.5%), and coxibs (1.3% vs 1.4%). The crude and adjusted RRs for each individual NSAID are reported in Supplementary Table 1. The mean dose used for ibuprofen was 909.8 465.4 mg/day,
Table 3. Overall Frequency, Crude Estimates of RRs and 95% CIs of GI Bleeding Associated With Current Drug Use
Variable NSAIDs ASA PPI H2RA Anticoagulants NSAID or ASA All APA Non-ASA APAa
Controls (n ¼ 1008), n (%) 169 175 350 11 58 318 202 57
(16.8) (17.4) (34.7) (1.1) (5.8) (31.5) (20.0) (5.7)
Cases (n ¼ 1008), n (%) 256 247 381 23 156 458 304 101
(25.4) (24.5) (37.8) (2.3) (15.5) (45.4) (30.2) (10.0)
Crude RR (95% CI) 1.7 1.5 1.1 2.1 3.0 1.8 1.7 1.9
(1.4–2.1) (1.2–1.9) (1.0–1.4) (1.0–4.4) (2.2–4.1) (1.5–2.2) (1.4–2.1) (1.3–2.6)
a OR and 95% CI for clopidogrel (82.3% of this group) were 1.8 (1.3–2.7); 26 controls (2.6%) and 45 cases (4.5%) were taking a combination of ASA and NSAIDs (RR, 1.7; 1.1.–2.9); 29 controls and 44 cases were taking a combination of non-ASA APA together with ASA (RR, 1.5; 1.0–2.5); 9 controls and 11 cases were taking a combination of non-ASA APA together with NSAIDs (RR, 1.2; 0.5–3.0).
Figure 1. Unconditional adjusted estimates of RRs and 95% CIs of (A) UGIB or (B) LGIB associated with current drug use. RRs were adjusted for hospital, age, gender, calendar semester, GI lesion history, dyspepsia history, smoking status, alcohol consumption, anticoagulants, PPI, non-aspirin APA use, H2RA, NSAID, and ASA use.
diclofenac was 124.7 287.3 mg/day, high-dose aspirin was 595.4 312.3 mg/day, and naproxen was 684.5 323.0 mg/day. The crude RR associated with high (>1200 mg/day) vs low dose of ibuprofen, the most frequent NSAID used in our study, was 3.9 (95% CI, 1.5–10.0). Figure 1 shows the unconditional adjusted RR of UGIB and LGIB associated with different drug use. Overall, NSAID use was associated with increased risk of upper and lower GI bleeding, although the estimate was 2 times higher for the upper compared with the lower GI tract. When the risk of upper and lower GI bleeding was divided by type of NSAID, the increased risk was present for ibuprofen, high-dose aspirin, and diclofenac in the upper GI tract but not in the lower GI tract (Supplementary Table 1). For other individual NSAIDs or coxibs, the low frequency of use limited the power of these analyses. A similar conclusion can be reached for the risk of LGIB associated with these NSAIDs. ASA use was also associated similarly with increased risk of both upper and lower GI bleeding. A similar conclusion can be reached concerning the risk associated with other non-aspirin APAs. We conducted sensitivity analysis excluding patients with bleeding that was due to ischemic colitis or inflammatory colitis, because use of NSAID or ASA might introduce some confounding by indication effect in these patients; however, the analysis
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
-
465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522
2014
Upper and Lower GI Bleeding and Toxic Drugs
did not show any significant change in the results (data not shown). In this study, the highest risk associated with both upper and lower GI bleeding was observed with anticoagulant use. The estimates were similar for lower and upper GI bleeding. Supplementary Table 2 shows the frequency of drug use in cases and controls and the crude and conditional adjusted RRs for that use in UGIB and LGIB patients. Current PPI use was present in 381 cases (37.8%) and in 350 controls (34.7%), but the distribution of use was different in cases with upper and lower GI bleeding events. PPI use was associated with decreased risk of UGIB (adjusted RR, 0.4; 95% CI, 0.3–0.5), whereas it was associated with a small increased risk of LGIB (adjusted RR, 1.5; 95% CI, 1.1–1.9). We have also analyzed the effects of PPI combined with these drugs. As shown in Supplementary Table 3, the combination of NSAIDs, ASA, other APAs, and anticoagulants with PPI was not associated with upper GI bleeding.
Discussion It is well-established that NSAIDs and ASA are associated with increased risk of UGIB,1,2,13,14 but data on their effects on the risk of bleeding in lower GI tract are less clear, although the evidence has been growing in recent years.8–10 Also, the effect of non-aspirin APA drugs and anticoagulants on the risk of LGIB is much less explored. This study has evaluated the risk of GI bleeding associated with these compounds and compares the risks of either upper or lower GI bleeding in the same study within the same source population and same temporal setting. We have observed that NSAIDs are associated with increased risk of both upper and lower GI bleeding and that the risk is clearly higher in the upper GI tract. Different animal and healthy volunteer studies have shown that NSAIDs can damage the mucosa of the small intestine and even the colon,3–7 but the clinical implications of those findings were unclear. A few clinical trials conducted in patients with different rheumatic conditions have shown a higher than expected incidence of lower GI complications with naproxen, diclofenac, or etoricoxib (MEDAL and VIGOR studies).15,16 The CONDOR17 and GI-REASON18 trials have shown that patients taking NSAIDs with/without PPI have a higher incidence of GI events (basically hemoglobin drops without clinically overt bleeding) compared with celecoxib, when taking into consideration both the upper and lower GI tract together. However, none of these studies have provided a clear estimate of the risk of major LGIB compared with non-NSAID use. Our study suggests that the increased risk of LGIB is lower than that seen for UGIB. In our population, NSAID use was mainly due to ibuprofen and diclofenac, whereas other compounds had much less use. On the basis of these data, we would
5
suggest that our overall estimates of risk for NSAIDs are driven mainly by ibuprofen (for the most part at analgesic doses). The lack of increased risk of GI bleeding for naproxen or coxibs has to be interpreted with caution because of the large variability of the estimates of risk. Another interesting finding of this study was that ASA was associated with a similar risk of upper and lower GI bleeding. Because there are previous studies that suggest that even ASA can induce mucosal damage in both the upper and lower GI tract,19 the effect on GI bleeding with this compound cannot rule out that the bleeding was due to the induction of mucosal damage and subsequent bleeding. However, ASA could have also induced bleeding from preexisting lesions (eg, angiodysplasia, diverticula). This antiplatelet mechanism is reinforced by the fact that non-ASA APA use (mostly clopidogrel) was associated with UGIB and LGIB, and that this increased risk was similar to that observed with ASA. Because these compounds have not been found to induce mucosal damage, the effect on bleeding should be mainly mediated by its antiplatelet effect. Anticoagulants (dicumarinics) were associated with a risk of UGIB and LGIB greater than that observed with antiplatelet drugs and NSAIDs. These compounds do not affect directly the GI mucosa but have a higher antihemostatic effect compared with antiplatelet drugs. We believe that this is a significant finding because the proportion of GI bleeders associated with anticoagulant use is growing, which may be a significant problem in the future, because the elderly population continues to increase in Western countries. Furthermore, PPI co-therapy, which decreases the risk of mucosal damage and complications in the gastroduodenal mucosa, will also have no protective effect on the lower GI tract. We have confirmed that PPI use was associated with >50% risk reduction in UGIB, but as expected, it was not found to reduce the risk of LGIB. On the contrary, we found a small increase in that risk with PPI use. We did not interpret this finding as a direct effect of PPI on the lower GI tract but as an effect of residual “confounding by indication,” despite controlling for a number of covariates. Yet, PPI use is usually a strong marker of worse overall comorbidity.11 Nevertheless, new studies should further investigate this effect, because recent experimental animal data suggest that PPI can alter the microbiota of the small bowel and potentiate the damaging effect of NSAIDs on the intestinal mucosa.20 This study has strengths and weaknesses. Among the strengths we include the relatively high number of patients included in the study and the fact that all were consecutive patients with documented upper or lower GI bleeding, which permitted the evaluation of the risk of upper and lower GI bleeding in the same cohort while obtaining comparable data. All bleeding cases were investigated, and the data were prospectively collected. The use of over-the-counter drugs by cases and controls was also collected. Among the limitations, the study
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580
6
581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
design (interview-based) could have introduced selection bias. The recollection of data, including the temporal use and dose of the drugs, is subject to bias because patients may not recall this properly in a life-threatening situation. Also, no additional validation of drug use was performed by using computerized databases. Although some of these problems can be solved or minimized by collecting the information with family members, the absence of prescription data at primary care level and pharmacy data may have affected the actual drug use. Bias cannot be completely prevented, especially when some underlying condition can be linked both to exposure and to the event of interest. Also, the sample size of the study did not allow providing reliable odds ratios for NSAIDs with low frequency of use and by dose, with exception made for ibuprofen. Finally, the controls used may not reflect the ideal population, which may affect the differences of drug use between cases and controls. In summary, our study provides evidence that NSAIDs, ASA, non-ASA APAs, and anticoagulants increase the risk of both upper and lower GI bleeding. The magnitude of this increase is similar for upper and lower GI tract for ASA, other APA, and anticoagulant use, whereas this increase in the risk seems to be higher in the upper compared with the lower GI tract for NSAID use, which may suggest different mechanisms of action. These results may have clinical implications, especially for prevention strategies. The risk of UGIB in susceptible patients can be reduced with PPI co-prescription, but this is not the case for the risk of LGIB. Because NSAID, antithrombotic, and antiplatelet therapy is growing in a progressively older population, special caution should be placed in those patients with a history of lower GI disease. We suggest that because of growing evidence in the field, current guidelines should be reconsidered, taking into consideration not only the upper but also the lower GI tract and not only NSAIDs and ASA but also include non-aspirin APAs and anticoagulant therapy.
Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org, and at http://dx.doi.org/10.1016/j.cgh.2014.11.007.
References 1. Lanza FL, Chan FK, Quigley EM. Guidelines for prevention of NSAID-related ulcer complications. Am J Gastroenterol 2009; 104:728–738. 2. García Rodríguez LA, Barreales Tolosa L. Risk of upper gastrointestinal complications among users of traditional NSAIDs and COXIBs in the general population. Gastroenterology 2007;132:498–506. 3. Bjarnason I, Hayllar J, Macpherson AJ, et al. Side effects of nonsteroidal anti–inflammatory drugs on the small and large intestine in humans. Gastroenterology 1993;104:1832–1847.
-,
No.
-
4. Graham DY, Opekun AR, Willingham FF, et al. Visible small– intestinal mucosal injury in chronic NSAID users. Clin Gastroenterol Hepatol 2005;3:55–59. 5. Maiden L, Thjodleifsson B, Seigal A, et al. Long-term effects of nonsteroidal anti-inflammatory drugs and cyclooxygenase-2 selective agents on the small bowel: a cross-sectional capsule enteroscopy study. Clin Gastroenterol Hepatol 2007;5: 1040–1045. 6. Goldstein JL, Eisen GM, Lewis B, et al. Small bowel mucosal injury is reduced in healthy subjects treated with celecoxib compared with ibuprofen plus omeprazole, as assessed by video capsule endoscopy. Aliment Pharmacol Ther 2007; 25:1211–1222. 7. Kurahara K, Matsumoto T, Lida M, et al. Clinical and endoscopic features of nonsteroidal anti-inflammatory drug-induced colonic ulcerations. Am J Gastroenterol 2001;96:473–480. 8. Lanas A, Sekar MC, Hirschowitz BI. Objective evidence of aspirin use in both ulcer and non–ulcer upper and lower gastrointestinal bleeding. Gastroenterology 1992;103:862–869. 9. Yuhara H, Corley DA, Nakahara F, et al. Aspirin and non-aspirin NSAIDs increase risk of colonic diverticular bleeding: a systematic review and meta-analysis. J Gastroenterol 2014;49: 992–1000. 10. Nagata N, Niikura R, Aoki T, et al. Colonic diverticular hemorrhage associated with the use of NSAIDs, low-dose aspirin, antiplatelet drugs, and dual therapy. J Gastroenterol Hepatol Q8 (in press). 11. Abraham NS, Hartman C, Richardson P, et al. Risk of lower and upper gastrointestinal bleeding, transfusions, and hospitalizations with complex antithrombotic therapy in elderly patients. Circulation 2013;128:1869–1877. 12. Lanas A, García-Rodríguez LA, Polo-Tomás M, et al. Time trends and impact of upper and lower gastrointestinal bleeding and perforation in clinical practice. Am J Gastroenterol 2009; 104:1633–1641. 13. Antithrombotic Trialists’ (ATT) Collaboration. Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials. Lancet 2009;373:1849–1860. 14. Sostres C, Lanas A. Gastrointestinal effects of aspirin. Nat Rev Gastroenterol Hepatol 2011;8:385–394. 15. Laine L, Curtis SP, Cryer B, et al. MEDAL Steering Committee: assessment of upper gastrointestinal safety of etoricoxib and diclofenac in patients with osteo-arthritis and rheumatoid arthritis in the Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) programme—a randomised comparison. Lancet 2007;369:465–473. 16. Laine L, Connors LG, Reicin A, et al. Serious lower gastrointestinal clinical events with nonselective NSAID or coxib use. Gastroenterology 2003;124:288–292. 17. Chan FK, Lanas A, Scheiman J, et al. Celecoxib versus omeprazole and diclofenac in patients with osteoarthritis and rheumatoid arthritis (CONDOR): a randomised trial. Lancet 2010; 376:173–179. 18. Cryer B, Li C, Simon LS, et al. GI-REASONS: a novel 6-month, prospective, randomized, open-label, blinded endpoint (PROBE) trial. Am J Gastroenterol 2013;108:392–400. 19. Smecuol E, Pinto-Sanchez MI, Suarez AA, et al. Low-dose aspirin affects the small bowel mucosa: results of a pilot study with a multidimensional assessment. Clin Gastroenterol Hepatol 2009;7:524–529.
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696
-
2014
Upper and Lower GI Bleeding and Toxic Drugs
20. Wallace JL, Syer S, Denou E, et al. Proton pump inhibitors 697 exacerbate NSAID-induced small intestinal injury by inducing 698 dysbiosis. Gastroenterology 2011;141:1314–1322. 699 700 Reprint requests 701 Address requests for reprints to: Angel Lanas, Servicio de Aparato Digestivo 702 Q1 Q2 Hospital Clínico Universitario, Zaragoza, Spain. e-mail: angel.lanas@gmail. com; fax: 34-976-765787. 703 704 Q3 Acknowledgments 705 Q4 The authors thank the following people who helped in their work: M. A Aranguren, Hospital Donostia, San Sebastian, Spain; Judith Millastre, Hospital 706 Universitario Miguel Servet, Zaragoza, Spain; Marta Gracia, Hospital Universitario Miguel Servet, Zaragoza, Spain; Ana Chaves, Hospital Costa del Sol, 707 Marbella, Spain; Marta Ponce, Hospital La Fe, Valencia, Spain; Meritxell 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
7
Sánchez-Lloansi, Corporación Sanitaria Universitaria, Sabadell, Spain; Pedro Lacarta, Hospital Clínico Universitario, Zaragoza, Spain; and Alberto Lúe, Hospital Clínico Universitario, Zaragoza, Spain. Conflicts of interest These authors disclose the following: L. A. García Rodríguez has received research funding from AstraZeneca R&D and Bayer and has also received honoraria for serving on scientific advisory boards for Bayer. The institution of Angel Lanas has received funds from Bayer from investigator-initiated research and Angel Lanas has been advisor to Pfizer and Bayer and received speaking fees from Bayer and AstraZeneca. The remaining authors disclose no conflicts. Q5 Funding The study was funded by the Official Spanish Agency, Instituto de Salud Carlos III-FIS, grant PI08/1301, and the CIBERehd, which is supporting the salary of Q6 Patricia Carrera.
FLA 5.2.0 DTD YJCGH54064_proof 1 December 2014 1:04 pm ce CLR
755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812
7.e1
813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870
Lanas et al
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
Supplementary Table 1. Crude and Conditional Adjusted RRs of UGIB and LGIB by Type of NSAID Use Overall Ibuprofen Aspirin (>300 mg) Diclofenac Naproxen Dexketoprofen Other NSAIDs Coxibs
UGIB Ibuprofen Aspirin (>300 mg) Diclofenac Naproxen Dexketoprofen Other NSAIDs Coxibs
LGIB Ibuprofen Aspirin (>300 mg) Diclofenac Naproxen Dexketoprofen Other NSAIDs Coxibs
Controls (n ¼ 1008), n (%) 99 8 20 15 7 12 14
(9.8) (0.8) (2.0) (1.5) (0.7) (1.2) (1.4)
Controls (n ¼ 563), n (%) 49 4 7 8 1 6 8
(8.7) (0.7) (1.2) (1.4) (0.2) (1.1) (1.4)
Controls (n ¼ 415), n (%) 49 3 13 6 5 6 6
(11.8) (0.7) (3.1) (1.4) (1.2) (1.4) (1.4)
Cases (n ¼ 1008), n (%) 124 50 45 16 11 16 13
(12.3) (5.0) (4.5) (1.6) (1.1) (1.6) (1.3)
Cases (n ¼ 563), n (%) 84 42 28 8 6 6 5
(14.9) (7.5) (5.0) (1.4) (1.1) (1.1) (0.9)
Cases (n ¼ 415), n (%) 36 8 17 7 5 10 7
(8.7) (1.9) (4.1) (1.7) (1.2) (2.4) (1.7)
Total 223 58 65 31 18 28 27
(11.1) (2.9) (3.2) (1.5) (0.9) (1.4) (1.3)
Total 133 46 35 16 7 12 13
(11.8) (4.1) (3.1) (1.4) (0.6) (1.1) (1.2)
Total 85 11 30 13 10 16 13
(10.2) (1.3) (3.6) (1.6) (1.2) (1.9) (1.6)
P value .088
