CLINICAL AND SYSTEMATIC REVIEWS 519

CME

Discontinuing Long-Term PPI Therapy: Why, With

REVIEW
Whom, and How?
Laura Targownik, MD, MSHS, FRCPC1

Proton pump inhibitors (PPIs) are among the most widely used class of drugs prescribed over the long term in all
of clinical medicine with 8–10% of ambulatory adults have been prescribed a PPI in the past 30 days. However,
numerous studies have raised doubts about the long term safety of PPI use. The purpose of this review is threefold:
(i) To provide an overview of the current evidence demonstrating associations between PPI use and adverse health
outcomes and the likelihood of the associations being causal (Why?); (ii) To be able to identify long-term PPI users in
whom the intensity of PPI therapy could be reduced or in whom PPIs could be eliminated outright (Who?); and (iii) To
provide strategies on how to reduce or stop chronic PPI therapy while maintaining symptom control and reducing the
risk for symptom or upper GI disease recurrence (How?).

Am J Gastroenterol 2018; 113:519–528; doi:10.1038/ajg.2018.29; published online 20 March 2018

INTRODUCTION balancing of the risks and benefits of reducing or eliminating PPI

The introduction of proton pump inhibitors (PPIs) in the mid-
1980s into the marketplace represented a major step forward in
the management of acid-peptic-related disease. By irreversibly
blocking the proton pump in the gastric oxyntic mucosa, PPIs
irreversibly block active proton pumps in the gastric oxyntic
mucosa, resulting in a profound reduction in basal and stimu-
lated gastric acid output (1). As a result, the intragastric pH can
be maintained at a sufficiently high level to allow for consistent
healing and prevention of esophagitis and peptic ulcer disease,
and more rapid control of associated symptoms (2,3). PPIs are
also perceived to have a favorable side effect profile, and are often
prescribed not only for established acid-related upper gastroin-
testinal (GI) disease but also empirically for nonspecific upper GI
symptoms (4–6). As a result, PPIs have become among the most
widely used class of drugs prescribed over the long term in all
of clinical medicine. It is estimated that 8–10% of ambulatory
adults have been prescribed a PPI in the past 30 days (7). PPI use
is particularly prevalent in the elderly; people over the age of 60
years are 3.5 times more likely to be using PPIs than those under
60 years (8). In 2009, over US$7 billion was spent on PPI pre-
scriptions in the United States, and over US$13 billion was spent
worldwide (9) and this does not include money spent on over-the-
counter PPIs (10).
However, numerous studies have raised doubts about the long-
term safety of PPI use (11,12). As a result, many practitioners are
looking for guidance of whether patients who have been using
PPIs chronically should be curtailing their use. This decision is
dependent of the indications for PPI use, and an assessment and
use in an individual patient. Also, if the decision is made to dis-
continue chronic PPI therapy, consideration needs to be given as
how this process should be undertaken. The purpose of this review
is threefold:
1. To provide an overview of the current evidence demonstrating
associations between PPI use and adverse health outcomes and
the likelihood of the associations being causal (Why?).
2. To be able to identify long-term PPI users in whom the
intensity of PPI therapy could be reduced or outright elimi-
nated (Who?).
3. To provide strategies on how to reduce or stop chronic PPI
therapy while maintaining symptom control and reducing
the risk for symptom or upper GI disease recurrence (How?).
WHY SHOULD WE BE CONCERNED ABOUT LONG-
TERM PPI USE?
To date, published studies have reported association between
the long-term use of PPIs and over two dozen unique complica-
tions (13–24) (Figure 1). The specific details regarding the studies
reporting associations between PPI use and these conditions will
not be described in detail in the article; the reader is directed to
several excellent reviews of this topic ((25,26)).
However, one key commonality of these studies demonstrating
PPI-associated adverse events is that they are nearly all observa-
tional studies, with the majority of the most cited and noteworthy
studies having been performed by retrospectively analyzing large

1
Section of Gastroenterology, Department of Internal Medicine, Max Rady School of Medicine at University of Manitoba, Winnipeg, Manitoba, Canada.
Correspondence: Laura Targownik, MD, MSHS, FRCPC, Section of Gastroenterology, Department of Internal Medicine, Max Rady School of Medicine at University
of Manitoba, 805G-715 McDermot Avenue, Winnipeg, Manitoba, Canada R3E 3P4. E-mail: laura.targownik@med.umanitoba.ca

© 2018 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY

 520 Targownik

Stroke

Dementia

REVIEW
Lungs
(pneumonia) Heart
(MI)

Subacute cutaneous
lupus erythematosus
(SCLE)

Muscle
(rhabdomyolysis)

Liver
(hepatic
encephalopathy)

Blood
(anemia)

Stomach
(fundic gland polyps)
Spontaneous
bacterial peritonitis
(SBP)

Acute interstitial
nephritis (AIN)

ESRD/
chronic renal
insufficiency

Small bowel
(bacterial overgrowth) Colon (C difficile,
Salmonella,
Campylobacter
colitis)
Bone
(hip fracture/ Microscopic
osteoporosis) colitis

Figure 1. Adverse events associated with proton pump inhibitor use. Original: Complications of proton pump inhibitor therapy. Gastroenterology 2017
Jul;153(1):35–48. Used with permission of Michael Vaezi.

health-care utilization data sets which had been collected primar- assigned, and these non-random factors that lead to PPI use (or
ily for administrative purposes. In an observational study, the non-use) may individually impact on the probability of that per-
exposure status (i.e., PPI user vs. PPI non-user) is not randomly son developing the outcome of concern. This issue of confound-

The American Journal of GASTROENTEROLOGY VOLUME 113 | APRIL 2018 www.nature.com/ajg

 Discontinuing Long-Term PPI Therapy 521

ing affects every observational trial, and it becomes impossible to cation, as is the case how we are comfortable stating that there is
definitely prove whether the PPI caused the outcome, or whether a causal association between cigarette smoking and lung cancer
the factors that led to PPI use caused the outcome. While statisti- even though there has never been a study where persons at risk

REVIEW
cal tools such as multivariable regression and propensity weighting for lung cancer were randomized to cigarette use or matching pla-
can be used to mitigate against the effects of confounding, there cebos (27).
will invariably remain some degree of residual confounding due to The Bradford Hill Criteria (Table 1) were conceived to provide
the presence of unmeasured differences in the exposed and non- a framework for determining the likelihood of causality where an
exposed groups. association between an exposure and outcome is detected (28).
To unequivocally prove that PPIs are directly responsible for While Hill originally proposed nine criteria for assessing the likeli-
any of the hypothesized complications, a randomized controlled hood of causality, the areas where there is universal agreement as
trial must be performed. However, it is exceedingly unlikely at this to their importance are those of strength of association, biologic
time that a randomized controlled trial will be undertaken that is plausibility, consistency, and temporality
sufficiently powered and of a sufficiently long duration to detect
a clinically and statistically significant increase in the incidence Strength of association
of these complications between PPI users and non-users. How- The greater the magnitude detected association between an expo-
ever, a causal relationship can potentially be presumed if there are sure and an outcome, the more likely that relationship is to be
other evidentiary pathways linking PPI use to a specific compli- causal and clinically relevant Strong associations that remain
between an exposure and outcome after adjusting for known con-
founders are less likely to due to the effect of residual unmeas-
ured confounders, and thus more likely to be real. Unmeasured
Table 1. Bradford Hill Criteria confounders are those factors that are associated with both PPI
use and the adverse event under study, but are either impossible
Criterion
to assess or difficult to measure accurately. As an example, global
Strength of Association Larger effect sizes are less likely to be due to frailty increases the risk of hip fractures and is associated with
chance or confounding degenerative diseases and senescence; people who are frail are
Biologic Plausibility Evidence for a plausible physiologic mechanism also more likely to be prescribed PPIs (29,30). While as clinicians
should exist, or be thought possible we may subjectively recognize when a patient is frail, it is difficult
Temporality The exposure must occur prior to the event to quantify precisely, and is therefore not generally reported in the
Consistency Results across studies in similar populations data sets in which these studies are performed.
should be similar This issue is distinct from that of statistical significance, which
Biologic Gradient Strength of Association should correlate with can be attained for small magnitude effects if the sample size of
exposure intensity the population under study if sufficiently large. This is particularly
Coherence Presence of association does not directly con- relevant given the epidemiologic studies reporting PPI-related
tradict pre-existing evidence complications are performed on data sets containing information
Analogy Similar exposures have been shown to be on thousands if not millions of people, meaning that highly sta-
cause similar disease processes tistically significant results can be reported for very small asso-
Specificity Other factors that would otherwise explain the ciations. Table 2 contains the reported effect sizes (odds ratios,
association are not present hazard, ratios, or risk ratios) of PPI exposure and the specific

Table 2. Most recent estimates of effect size for PPI-associated adverse events

Adverse event Effect size (95% CI) Reference Heterogeneity?

Enteral infection OR 2.55 (1.53–4.26) Leonard et al. (23) Yes

Community-acquired pneumonia OR 1.49 (1.16–1.92) Lambert et al. (36) Yes
Clostridium difficile-associated diarrhea OR 1.26 (1.12–1.29) Cao et al. (107) Yes
Hip fracture OR 1.26 (1.16–1.36) Zhou et al. (37) Yes
Dementia HR 1.44 (136–1.52) Gomm et al. (16) N/A
Vitamin B12 deficiency HR 1.83 (1.36–2.46) Jung et al. (108) Borderline
Chronic renal failure RR 1.36 (1.07–1.72) Nochaiwong et al. (109) Yes
Myocardial infarction OR 1.16 (1.09–1.24) Shah et al. (22) N/A
HR, hazard ratio; N/A, not applicable; OR, odds ratio; PPI, proton pump inhibitor; RR, relative risk.

© 2018 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY

 522 Targownik
adverse events in the most recently published meta-analysis as
of September 2017, or of single studies where just one study has
been published. For all adverse events, the pooled adjusted effect
REVIEW
size is <2, and for many of the most frequently studied associa-
tions, the reported effect sizes are <1.5. This level of association,
even if highly statistically significant, can be explained away by
the effects of unmeasured confounding, and do not let strong sup-
port for causality. Conversely, clinicians should be cognizant that
even small magnitude effects, if truly causal, can have enormous
impacts of population health if the exposure is highly prevalent,
and the outcome is morbid, and not rare. As an example, if 15%
of the geriatric population are chronic PPI users, and PPI use was
directly responsible for 25% increase in the risk of hip fracture
among chronic users, then PPI use would lead to an additional
45,000 hip fractures per year among people over age 65 years in
the United States (31) and would be responsible for 1,500 addi-
tional deaths each year (32).
Temporality
It is unimpeachably true that a patient must be exposed to a
putative risk factor before the occurrence of an adverse event
in order to be implicated in its cause. Essentially all studies that
have studied PPI-related adverse events attempt to assess PPI
exposure before the event of interest. However, the precise date
that a diagnosis of a putative complication occurs may not nec-
essarily be the same as what is recorded in the study database.
This is particularly the case for diagnoses, which are progressive
and degenerative like Alzheimer’s-type dementia or osteoporo-
sis, or those with a long preclinical prodrome, which may have
an onset months to years before being clinically recognized (33).
In more acutely occurring complications, early symptoms may
be misdiagnosed as similarly presenting conditions, as commu-
nity-acquired pneumonia may be initially misdiagnosed as an
upper respiratory infection, congestive heart failure, or, if chest
pain is a predominant feature, gastroesophageal reflux disease
(GERD) (34). If a PPI is dispensed during this prodromal period,
it may generate as association between PPI use and this adverse
outcome which appears to be temporal, but in fact violates
this assumption. This phenomenon is commonly referred to as
protopathic bias.
Consistency
In order for a result to be consistent, the results of the epidemio-
logic studies performed in different populations should fund
associations between PPI use and adverse events of comparable
magnitude. Studies of the most widely evaluated PPI-associated
adverse conditions (fracture, Clostridium difficile infection, com-
munity acquired pneumonia) have demonstrated some degree of
heterogeneity, and also have not shown a consistent dose response
effect (35–37). In addition, as there is not currently a requirement
for preregistration of epidemiologic studies, it is very likely that
studies which do not demonstrate an association between PPI
use and an adverse event are not completed or published. Com-
monly referred to as the “file-drawer effect ((38)),” this means that
positive studies tend to be available for meta-analysis, and the
The American Journal of GASTROENTEROLOGY
non-published negative studies do not get included, biasing the
findings in favor of there being an effect when none exists.
Biologic plausibility
If PPIs do indeed directly cause any of the adverse events for
which PPI use is associated, then it must result from their physi-
ologic effects in the person using them. The physiologic effects of
PPIs can be characterized into three categories: (1) effects result-
ing from decreased gastric acid output due to the direct effect of
inhibition of proton pump-mediated gastric acid secretion; (2)
inhibition of proton pumps found in other organ systems; (3)
effects due to alteration of drug metabolism, leading to changes in
the concentration and activity of other drugs or metabolites that
affect bone metabolism.
(1) Effects resulting from reduction of gastric acid secretion:
Of these categories, the most well understood effect is that of
PPIs on blocking gastric acid output. The presence of gastric acid
is often considered to be a nuisance due to its role in the patho-
physiology of GERD and peptic ulcer disease. Further, although
gastric acid is needed to convert inactive pepsinogen into the
gastric protease pepsin, ingested proteins will still be completely
digested in the absence of pepsin and gastric acid by pancreatic
proteases. Therefore, one might think that pharmacological in-
hibition of gastric acid secretion would have minimal negative
consequence.
However, the fact that gastric acid production has been evo-
lutionary retained across nearly all vertebrates suggests that the
ability to generate gastric acid provides a survival advantage, at
least until modern times (39). Most probably, the benefits of gas-
tric acid pertain to its ability to stabilize the gut microbiota by
preventing colonization with acid-sensitive organisms and path-
ogens; interestingly, carnivorous and scavenging animals, who
have a higher risk of exposure to pathogens, tend to have higher
levels of gastric acid production than do herbivores or grazers
(40). In humans, multiple studies have demonstrated that condi-
tions which lead to a reduction in gastric acid output (atrophic
gastritis, surgical gastrectomy) increase the risk of enteric infec-
tions with acid-sensitive organisms such as Salmonella and
Campylobacter (41).
PPI use has been strongly associated with an increased risk
of infection with C. difficile (35,42–44). However, ingested
C. difficile organisms are acid resistant, and therefore a reduc-
tion in gastric acid itself should not result directly in an increased
likelihood of C. difficile infection (45). However, PPIs have also led
to significant alterations of the gut microbiota (46), with decreases
in general biodiversity, increase in Firmicute phyla organisms,
and a decrease in Bacteroides. These changes may create an envi-
ronment where C. difficile is more likely proliferate leading to
symptomatic infection in people who are already colonized with
C. difficile organisms (47).
In addition, gastric acid has a role in facilitating the absorption
of iron, calcium, and vitamin B12. Of these, the effects of gastric
acid inhibition on calcium absorption are likely the most clini-
cally relevant (48–50), given the important role that calcium has
VOLUME 113 | APRIL 2018 www.nature.com/ajg

in bone homeostasis and in maintaining bone strength (51). While
calcium is more soluble in increasingly acidic solutions, the effects
of PPI use on calcium absorption are negligible in most circum-
stances (49,52,53). In addition, the majority of studies evaluating
the relationship between PPI use and bone mineral density have
not shown any effect of PPI use on bone strength or on the preva-
lence of osteoporosis (54,55). Therefore, there is minimal biologic
plausibility for PPIs to promote osteoporosis related fracture, at
least as mediated by reduction in gastric acid secretion.
(2) Effects from blockage of non-GI proton pumps: While PPIs pri-
marily inhibit the action of the proton pumps found in the gas-
tric oxyntic mucosa, vacuolar proton pumps that can be blocked
by PPIs are found elsewhere in the body, particularly in bone
osteoclasts and the distal tubules of the nephron (56,57). While
blockage of renal proton pumps in animal models may lead to de-
creased renal acid secretion and cause a mild metabolic acidosis,
it does not have any effect on glomerular function (58). Therefore,
it is unlikely that this mechanism is responsible for the associa-
tion between PPIs and chronic renal insufficiency. In the case of
osteoclast-based proton pumps, inhibition should decrease bone
resorption, resulting in increased bone mineral density.
(3) Other proposed mechanisms: In the studies that have linked
PPI use to Alzheimer’s disease, and chronic renal insufficiency
(16,24), the investigators have suggested pathophysiologic path-
ways that are independent of their main pharmacologic action. As
an example, the PPI lansoprazole has been shown to increase the
synthesis of certain amyloid-β protein, which is also elaborated in
persons with Alzheimer’s disease (59). However, the studies sup-
porting this and other mechanisms have not been demonstrated
in animal models or humans, and also have not be widely repli-
cated. In these cases, the evidence for biologic plausibility is the
most tenuous.
In addition, PPIs are metabolized in the liver by enzymes in
the cytochrome P450 family; competitive inhibition of these
cytochromes could interfere with the metabolism and/or clearance
of other drugs. The proposed mechanism for the apparent associa-
tion between myocardial infarction and PPIs among clopidogrel
users was that PPIs would prevent cytochrome P450 metabolized
conversion of the prodrug form of clopidogrel into the active drug
(60). However, the clinical significance of this mechanism is doubt-
ful (61), given that prospective trials demonstrated no difference
in adverse cardiovascular event rates between PPI users and non-
users among persons who were exposed to clopidogrel (62,63).
Overall, while correlations between chronic PPI use and a
diverse range of acute and chronic complications have been con-
sistently demonstrated, there is insufficient evidence supporting
a causal role for PPIs in the etiology of any of these conditions.
The supporting evidence for infectious complications, specifically
enteric infections, is the most compelling, whereas the evidence
for extragastrointestinal complications whose pathophysiology
should not be influenced by reduced acid secretion is especially
weak. At this point, guidelines that do not suggest chronic PPI
users should be screened for any of the PPI-associated disease,
© 2018 by the American College of Gastroenterology
Discontinuing Long-Term PPI Therapy 523
and it is not recommended to discontinue PPIs in persons who
are already afflicted with any of the PPI-associated adverse events,
nor should long-term PPI users receive specific treatment in an
REVIEW
attempt to prevent any of the complications (12,64). As an exam-
ple, it is not recommended that PPI users undergo screening for
osteoporosis or undergo a fracture risk assessment just because
they are PPI users, and there is no contraindication against PPI
use among people with established osteoporosis or a history of hip
fracture. However, one cannot prove that PPIs are entirely safe, and
the data sources that would be required to prove the safety of PPIs
are lacking. Therefore, it remains important not to inappropriately
use PPIs, and specifically, to limit their use to persons with condi-
tions where the benefits of PPI are well defined.
WHO SHOULD REMAIN ON LONG-TERM PPI
THERAPY
Definite indications for long-term PPI therapy
Treatment of erosive esophagitis and prevention of relapse: Among
persons with erosive esophagitis that completely healed on PPI
therapy, the risk of endoscopic relapse is 72% for persons who
stop using PPIs compared to 13% for those who continue PPI
therapy during this period (65). Furthermore, PPIs are more
effective than histamine-2 receptor antagonists (H2RAs) in main-
taining symptom-free remission among person with healed ero-
sive esophagitis (91% PPI vs. 62% H2RA, P<0.001) (65). The risk
of relapse with PPI discontinuation is particularly high among
persons with more severe forms of esophagitis. PPI discontinu-
ation could be considered among persons with milder forms of
erosive esophagitis (LA Class A/B), or in those persons where a
major factor that predisposed an individual to erosive esophagitis
(such as heavy alcohol use) has been eliminated.
Treatment of PPI-responsive esophageal eosinophilia (REE):
Approximately 50% of persons with apparent eosinophilic
esophagitis will have a clinical and histologic response to PPI
therapy (66). These patients are referred to as having PPI-REE, or
PPI-REE. Persons with PPI-REE have a high rate of symptomatic
and/or clinical relapse when effective therapy is discontinued.
Although the natural history of PPI-REE is not well defined, there
is evidence that untreated or undertreated eosinophilic esophagitis
may eventually result in chronic stricture (67). In addition, the
long term risks associated with PPI use are likely less clinically
significant than those resulting from long-term corticosteroid use,
the most commonly utilized medical therapy for EoE. Therefore,
the long-term use of PPIs in persons with eosinophilic esophagitis
would seem to be reasonable.
Prevention of peptic ulcer disease and its complications: PPIs are
highly efficacious in preventing gastric and duodenal ulceration
in chronic users of nonsteroidal anti-inflammatory drugs (68,69).
In epidemiologic studies, persons who used PPIs concomitantly
with nonsteroidal anti-inflammatory drug therapy reduce the
odds of being hospitalized for complicated peptic ulcer disease
by between 50% and 80% (70–72). PPI-based gastroprotection is
recommended for all chronic aspirin and/or nonsteroidal anti-
inflammatory drug users who have risk factors that increase the
The American Journal of GASTROENTEROLOGY
 524 Targownik
risk of peptic ulcer disease-related complications; these risk fac-
tors include concomitant use of antiplatelet and/or anticoagulants,
concomitant use of systemic corticosteroids, age over 70 years,
REVIEW
prior history of peptic ulcer disease, and the presence of severe
medical comorbidities (71,73,74).
Prevention of progression of Barrett’s esophagus: Acid-induced
esophageal injury is a major etiologic factor in the development
of intestinal metaplasia, and has been implicated in the progres-
sion to dysplasia and neoplasia (75). There are conflicting reports
about the benefits of PPIs in the prevention of esophageal adeno-
carcinoma, although there is a trend towards their being protec-
tive (76–78). All major guidelines on the management of Barrett’s
esophagus strongly recommend PPI therapy, even in the absence
of symptoms, to prevent progression of Barrett’s esophagus to dys-
plasia and esophageal adenocarcinoma (79,80)
Zollinger–Ellison Syndrome: Hypersecretion of gastric acid due
to the unregulated release of gastrin from a gastrinoma is the cardi-
nal feature of Zollinger–Ellison Syndrome. PPIs are the most effec-
tive therapy for control of complications of gastric acid secretion
(81–83) While removal of the gastrinomatous tissue may allow for
discontinuation of PPIs, this is not possible for persons with meta-
static disease, or those with underlying multiple endocrine neopla-
sia type 1 (84). In these patients, long-term continuation of PPIs is
of unquestioned value.
Possible indications for long-term PPI therapy
There are a number of indications where though PPI use may
be effective in the short term, the benefits of continuing therapy
are not well described. In these scenarios, withdrawal of PPIs in
would be unlikely to be result in significant adverse events. For
patients with these indications, n=1 trials of substitution of PPIs
with less potent anti-secretory therapies like H2RAs, pain modu-
lating agents like tricyclic antidepressants or selective serotonin
reuptake inhibitors, or outright discontinuation should be con-
sidered.
Functional dyspepsia: PPIs are frequently prescribed for patients
with dyspepsia, even if no ulcer is present. The most recent meta-
analysis of studies comparing PPIs with placebo for functional dys-
pepsia showed response rates of 34% for PPIs and 25% for placebo
(risk ratio (RR) 0.88, 95% confidence interval (CI): 0.82–0.94) (85).
However, the maximum duration of therapy evaluated is 8 weeks,
and it is unclear whether this efficacy benefit is maintained over
time. As functional dyspepsia is a chronic condition with recur-
rent symptoms, persons with functional dyspepsia likely will use
PPIs for longer periods of time, despite there being no evidence of
continued benefit over the longer term.
Endoscopy-negative reflux disease: Persons with reflux-like symp-
toms but with no endoscopic evidence of erosive esophagitis who
are given PPIs have higher levels of symptom relief when com-
pared to placebo (RR 0.69, 95% CI: 0.65–0.78); H2RAs (RR 0.78,
95% CI: 0.62–0.97) or prokinetics (RR 0.72, 95% CI: 0.56–0.92)
(86). However, there is no evidence for the benefit of PPI therapy
over the long-term among patients with endoscopy-negative reflux
disease. Persons with endoscopy-negative reflux disease are a het-
erogeneous group, including those who have non-erosive reflux
The American Journal of GASTROENTEROLOGY
disease and those with functional heartburn, who manifest reflux
symptoms consistent with reflux but do not have physiologic evi-
dence of reflux of acidic gastric contents in to the esophagus. PPIs
are likely effective in the long term for the former, but they are less
likely to be efficacious in those whose symptoms are not due to
esophageal acid exposure (87).
Symptoms presumed secondary to laryngopharyngeal reflux:
Laryngopharyngeal reflux (LPR) is often implicated in the etiol-
ogy of chronic symptoms of laryngopharyngeal irritation such as
hoarse voice, sore throat, or globus sensation (88). Current ENT
guidelines recommend PPI therapy for persons with laryngeal
symptoms presumed secondary to LPR, often at high dose and for
durations of at least 3–6 months (89). However, the diagnosis of
LPR is generally based on the finding of irritation of the posterior
larynx and arytenoid cartilage on visual inspection, which poorly
differentiates between laryngitis and those due to other causes,
such as post-nasal drip or environmental allergens and irritants
(90). Further, more accurate evaluations for physiologic reflux,
such as pH monitoring or multichannel impedance are rarely done
(91). The evidence supporting PPIs for treatment of LPR is at best
conflicted, with the most recently published meta-analyses dem-
onstrating PPIs were no better than placebo in achieving symptom
improvement (defined as a 50% reduction in laryngeal symptoms
(PPIs 42% vs. placebo 39%, difference 4%, 95% CI: 6–14)), and also
were no different than placebo in improving the signs of presumed
LPR on laryngoscopic examination (92).
Inappropriate PPI use
Inappropriate PPI use can be defined as use which either occurs in
the absence of a known indication, or for use for a non-absolute
indication where a trial of PPI cessation has not been considered
or attempted (93). All current guidelines recommend limiting
PPI use to specific indication, and to use the smallest dose for the
shortest duration required to adequately control symptoms. Even
in the absence of definitive evidence of medical harm associated
with PPI use, the excess cost associated with unnecessary therapy
should be enough reason to make attempts to curtail the inap-
propriate PPI use.
In 2010, Heidelbaugh et al. (94) assessed PPI use and their
indications in 946 patients in ambulatory care clinic, and found
that over one-third of PPI users did not have a documented indi-
cation for ongoing use, and 10% were using for extraesophageal
indications. Furthermore, nearly one-half patients did not have
any documentation of the response to therapy having been re-
evaluated at any point following PPI initiation. Patients without
appropriate indications who were using PPIs, without an indica-
tion, spent approximately US$1,500 per year on PPIs. Similarly,
Ladd et al. (95) found that among 201 persons who were newly
prescribed PPIs on a in-patient medical ward, three-quarters
had either no documented indication or an inappropriate indi-
cation for PPI, of whom 40% continued PPI therapy following
discharge.
The factors that are driving the inappropriate initiation and/or
failure to address the need for ongoing PPI therapy have not been
systematically evaluated, although a model characterizing patient
VOLUME 113 | APRIL 2018 www.nature.com/ajg

and physician behaviors that drive inappropriate PPI therapy can
be envisioned. First, in primary care, PPIs are often initiated empir-
ically for GI symptoms, as opposed to being used for an objectively
diagnosed condition, and there it is difficult if not impossible to
differentiate based on symptoms alone whether someone with
pyrosis has functional heartburn vs. erosive esophagitis, or func-
tional dyspepsia vs. peptic ulcer disease. Second, PPIs will lead to
complete healing of the active ulceration in persons with peptic
ulcer disease or erosive esophagitis, rendering it impossible to
determine whether the original symptoms arose from an organic
disease or a functional one. Third, PPI therapy in many instances
is initiated following endoscopy, or during an in-patient stay, by a
physician who does not maintain an ongoing relationship with the
patient. In these situations, the primary care physician may not be
certain as to whether PPI discontinuation is reasonable and safe.
Fourth, most upper GI symptoms arising from a functional syn-
drome wax and wane in intensity, and patients are more likely to
seek health care and remedies when symptoms are at their worst.
The symptoms for which PPIs are then prescribed would have
a high chance of having improved, even if the PPI had not been
used due to regression to the mean. Last, the absence of highly
effective competing drug options for upper GI symptomatology
compels prescribers to favor PPIs over other options. This model
provides an important framework for health-care prescribers and
symptoms who want to engage with reducing the burden of inap-
propriate PPI use
HOW SHOULD PPI THERAPY BE DEPRESCRIBED
PPI deprescribing refers to a patient or physician effort to either
discontinue PPIs outright, decrease exposure to PPIs by stepping
down to a lower dose or intermittent/on-demand treatment regi-
men, or substituting PPIs with a less costly and/or less potent form
of gastric acid inhibition. An attempt at PPI deprescription should
be considered for all PPI users who do not have definite indica-
tions for long-term therapy (severe erosive esophagitis, high risk
for aspirin/nonsteroidal anti-inflammatory drug-related compli-
cations, and Barrett’s esophagus), with the goal of therapy being to
use the lowest dose of PPI necessary to control acid-related symp-
toms and/or mucosa damage.
Effectiveness of step-down regimens
On-demand therapy, where a patient uses PPIs only when
symptoms occur and taken until symptoms are adequately
relieved, is significantly more effective than placebo in providing
adequate symptom relief (51% vs. 14%, P<0.0001) (96). A meta-
analysis that included randomized controlled trials comparing
continuing PPI use to on-demand PPI use or in persons with
mild treatment-responsive GERD symptoms found that per-
sons assigned to on-demand therapy has a significant increase
in the risk of loss of symptom control (16% vs. 10%, RR 1.71,
95% CI: 1.31–2.21), while decreasing weekly pill intake by nearly
3.8 PPI doses per week (97). In a primary care-based Norwegian
study of 2156 patients who had PPI-responsive GERD, persons
randomized to receive ongoing management with H2RAs had
© 2018 by the American College of Gastroenterology
Discontinuing Long-Term PPI Therapy 525
a significantly lower rate of symptom control than did those
maintained on regular PPI therapy (32% vs. 72%, P<0.0001)
(98). While these findings seem to favor continuing therapy as
REVIEW
opposed to stepping down, the majority of persons who were
decreased to on-demand therapy continued to report adequate
symptom relief, as did a meaningfully large minority of those
given H2RAs.
Inadomi et al. (99) prospectively demonstrated that 80% of per-
sons with controlled symptoms of GERD on higher-than-standard
doses of PPIs could be stepped down to standard dose PPIs with-
out developing significant recurrent symptoms. Aside from the
exceedingly rare Zollinger–Ellison syndrome, there are no clinical
scenarios that absolutely necessitate the use of higher-than-stand-
ard doses of PPIs. Therefore, dose reduction strategies should be
considered for all persons who are using PPIs chronically at high
doses.
Total discontinuation of PPIs
In persons without a clear indication for PPI therapy, or in those
persons who were treated empirically for upper GI symptoms, a
trial of discontinuation can be attempted. Not surprisingly, per-
sons with mild esophagitis who initially responded to PPI therapy
who were given placebo had a relapse rate between 48 and 100%
over the next 6 months, compared with 17–33%% for those who
continued on standard dose PPI (100).
For patients who are taking PPIs for unclear indications, or for
symptoms where there is little evidence for PPI use, common sense
would dictate that they should be discontinued without the expec-
tation of adverse consequences. However, the sudden discontinu-
ation of PPIs may cause a sudden and profound rise in gastric acid
output, a phenomenon known as rebound gastric acid hypersecre-
tion. The regular use of PPI raises intragastric pH; at high pH lev-
els, the secretion of the trophic hormone gastrin is not negatively
inhibited, predisposing to the development of hypergastrinemia
that in turn stimulates oxyntic cell proliferation and an increased
capacity to generate gastric acid (101). So long as the PPI is being
taken, this increased capacity for gastric acid generation is kept
in check. However, if PPI therapy is abruptly discontinued, large
amounts of gastric acid may be generated, which can theoretically
aggravate upper GI symptoms (102).
In a study assessing the clinical significance of this physiologic
phenomenon, Reimer et al. (103) took persons with no history
of GERD or dyspepsia and gave them all PPIs continuously for 8
weeks. At the eighth week mark, these persons were randomized
to either placebo or continuing the PPI for an additional 4 weeks;
40% of persons who discontinued the PPI reported dyspepsia in
the 4 weeks following discontinuation. In a similarly designed
study, Niklasson et al. (104) found that symptoms on PPI with-
drawal were significantly correlated with the degree of PPI-induced
hypergastrinemia. In a follow-up study where 78 PPI users not
having a documented indication were discontinued, nearly two-
thirds (51/78) developed upper GI symptoms within 6 months.
People who developed upper GI symptoms subsequently under-
went endoscopy, where 40% had endoscopic findings consistent
with their symptoms, mostly mild esophagitis (105).
The American Journal of GASTROENTEROLOGY
 526 Targownik
One strategy to mitigate against rebound acid hypersecretion-
related symptoms would be to taper off PPI use, as opposed to
sudden discontinuation. In a study by Inadomi et al. (106) where
REVIEW
PPI users were discontinued gradually (1/2 of a standard dose
for 2 weeks, then discontinuing), 60% were asymptomatic on no
medications or over-the-counter antacids in the year following.
Whether a more prolonged taper leads to fewer rebound symp-
toms is not known, but could be considered for those who develop
rebound symptoms with a more rapid taper. In my practice, I will
advise persons who wish to discontinue PPIs to take them every
other day for 2 weeks, twice per week for 2 weeks, and then to
discontinue. I also advise patients who discontinue PPIs to be
mindful of rebound symptoms, and not to immediately reintro-
duce PPIs at the first sign of recurrent symptoms, especially if they
occur in the first week after stopping therapy.
SUMMARY
While there is not yet definitive evidence of serious harm associated
with PPI use, it remains possible that further studies will be more
strongly suggestive of a causal relationship, especially for those associ-
ations where biologic plausibility exists. Given the current uncertainty,
it behooves the clinician to limit the use of long-term PPIs to those
with definite indications, and to actively identify and make attempts to
deprescribe PPIs in persons without absolute indications for contin-
ued use, or those in whom a reason for ongoing use cannot be deter-
mined. When discontinuing PPIs, consideration should be given to
step-down strategies, such as using H2RAs, or tapering the dose of
PPIs over time. Further, chronic PPI users should be cautioned that
upper GI symptoms may recur in the short term, and that the occur-
rence of these symptoms do not necessarily indicate a need for long-
term PPIs, but rather may be a sign of transient reversible gastric acid
hypersecretion. Last, health-care systems should aim to develop and
evaluate programs to assist and guide physicians and patients through
the deprescribing process with the goal of eliminating unnecessary
PPI use, and reduce superfluous health-care expenditures
CONFLICT OF INTEREST
Guarantor of the article: Laura Targownik, MD, MSHS, FRCPC.
Specific author contributions: All aspects of creation of this article
were undertaken by Laura Targownik without assistance from other
individuals
Financial support: No external financial support.
Potential competing interests: Laura Targownik has received sup-
port for research studies from Janssen Canada and Abbvie Canada.
In the past 24 months, she has served on advisory boards for Janssen
Canada, Takeda Canada, Merck Canada, Pfizer Canada, Ferring
Canada, and Mallinckrodt. She has served on the Speaker’s Panel for
Takeda Canada and Janssen Canada
REFERENCES
1. Wolfe MM, Soll AH. The physiology of gastric acid secretion. N Engl J
Med 1988;319:1707–15.
2. Hunt RH, Cederberg C, Dent J et al. Optimizing acid suppression for
treatment of acid-related diseases. Dig Dis Sci 1995;40:24S–49S.
The American Journal of GASTROENTEROLOGY
3. Li Y, Sha W, Nie Y et al. Effect of intragastric pH on control of peptic ulcer
bleeding. J Gastroenterol Hepatol 2000;15:148–54.
4. Austin PC, Stuart EA. Moving towards best practice when using inverse
probability of treatment weighting (IPTW) using the propensity score
to estimate causal treatment effects in observational studies. Stat Med
2015;34:3661–79.
5. Meineche-Schmidt V. Empiric treatment with high and standard dose of
omeprazole in general practice: two-week randomized placebo-controlled
trial and 12-month follow-up of health-care consumption. Am J Gastro-
enterol 2004;99:1050–8.
6. Rabeneck L, Souchek J, Wristers K et al. A double blind, randomized,
placebo-controlled trial of proton pump inhibitor therapy in patients with
uninvestigated dyspepsia. Am J Gastroenterol 2002;97:3045–51.
7. Rotman SR, Bishop TF. Proton pump inhibitor use in the U.S. ambulatory
setting, 2002–2009. PLoS ONE 2013;8:e56060.
8. Pottegard A, Broe A, Hallas J et al. Use of proton-pump inhibitors among
adults: a Danish nationwide drug utilization study. Ther Adv Gastroen-
terol 2016;9:671–8.
9. Katz MH. Failing the acid test: benefits of proton pump inhibitors may
not justify the risks for many users. Arch Intern Med 2010;170:747–8.
10. Shaikh M, Dutta Choudhury S, Mohanty J et al. Modulation of excited-
state proton transfer of 2-(2'-hydroxyphenyl)benzimidazole in a macro-
cyclic cucurbit[7]uril host cavity: dual emission behavior and pK(a) shift.
Chemistry 2009;15:12362–70.
11. Chen J, Yuan YC, Leontiadis GI et al. Recent safety concerns with proton
pump inhibitors. J Clin Gastroenterol 2012;46:93–114.
12. Gill JM, Player MS, Metz DC. Balancing the risks and benefits of proton
pump inhibitors. Ann Fam Med 2011;9:200–2.
13. Lam JR, Schneider JL, Zhao W et al. Proton pump inhibitor and histamine 2
receptor antagonist use and vitamin B12 deficiency. JAMA 2013;310:2435–42.
14. Bajaj JS, Zadvornova Y, Heuman DM et al. Association of proton pump
inhibitor therapy with spontaneous bacterial peritonitis in cirrhotic
patients with ascites. Am J Gastroenterol 2009;104:1130–4.
15. Xie Y, Bowe B, Li T et al. Risk of death among users of proton pump
inhibitors: a longitudinal observational cohort study of United States
veterans. BMJ Open 2017;7:e015735.
16. Gomm W, von Holt K, Thome F et al. Association of proton pump
inhibitors with risk of dementia: a pharmacoepidemiological claims data
analysis. JAMA Neurol 2016;73:410–6.
17. Laheij RJ, Sturkenboom MC, Hassing RJ et al. Risk of community-
acquired pneumonia and use of gastric acid-suppressive drugs. JAMA
2004;292:1955–60.
18. Yang YX, Lewis JD, Epstein S et al. Long-term proton pump inhibitor
therapy and risk of hip fracture. JAMA 2006;296:2947–53.
19. Targownik LE, Lix LM, Metge CJ et al. Use of proton pump inhibitors and
risk of osteoporosis-related fractures. CMAJ 2008;179:319–26.
20. Dial S, Delaney JA, Barkun AN et al. Use of gastric acid-suppressive
agents and the risk of community-acquired Clostridium difficile-associated
disease. JAMA 2005;294:2989–95.
21. Juurlink DN, Gomes T, Ko DT et al. A population-based study of the
drug interaction between proton pump inhibitors and clopidogrel. CMAJ
2009;180:713–8.
22. Shah NH, LePendu P, Bauer-Mehren A et al. Proton pump inhibitor usage
and the risk of myocardial infarction in the general population. PLoS
ONE 2015;10:e0124653.
23. Leonard J, Marshall JK, Moayyedi P. Systematic review of the risk of
enteric infection in patients taking acid suppression. Am J Gastroenterol
2007;102:2047–56.quiz 2057
24. Klatte DCF, Gasparini A, Xu H et al. Association between proton pump
inhibitor use and risk of progression of chronic kidney disease. Gastro-
enterology 2017;153:702–10.
25. Vaezi MF, Yang YX, Howden CW. Complications of proton pump inhibi-
tor therapy. Gastroenterology 2017;153:35–48
26. Johnson DA, Oldfield EC. Reported side effects and complications of
long-term proton pump inhibitor use: dissecting the evidence. Clin
Gastroenterol Hepatol 2013;11:458–64.quiz e37–8
27. Doll R. Smoking and carcinoma of the lung. Acta Unio Int Contra
Cancrum 1953;9:495–506
28. Hill AB. The environment and disease: association or causation? Proc R
Soc Med 1965;58:295–300
29. Woods NF, LaCroix AZ, Gray SL et al. Frailty: emergence and conse-
quences in women aged 65 and older in the Women's Health Initiative
Observational Study. J Am Geriatr Soc 2005;53:1321–30.
VOLUME 113 | APRIL 2018 www.nature.com/ajg

30. Teramura-Gronblad M, Hosia-Randell H, Muurinen S et al. Use of
proton-pump inhibitors and their associated risks among frail elderly
nursing home residents. Scand J Prim Health Care 2010;28:154–9.
31. Brauer CA, Coca-Perraillon M, Cutler DM et al. Incidence and mortality
of hip fractures in the United States. JAMA 2009;302:1573–9.
32. Haentjens P, Magaziner J, Colon-Emeric CS et al. Meta-analysis: excess
mortality after hip fracture among older women and men. Ann Intern
Med 2010;152:380–90.
33. Dubois B, Padovani A, Scheltens P et al. Timely diagnosis for Alzheimer’s
disease: a literature review on benefits and challenges. J Alzheimers Dis
2016;49:617–31.
34. Filion KB, Chateau D, Targownik LE et al. Proton pump inhibitors and
the risk of hospitalisation for community-acquired pneumonia: replicated
cohort studies with meta-analysis. Gut 2014;63:552–8.
35. Tariq R, Singh S, Gupta A et al. Association of gastric acid suppression
with recurrent Clostridium difficile infection: a systematic review and
meta-analysis. JAMA Intern Med 2017;177:784–91.
36. Lambert AA, Lam JO, Paik JJ et al. Risk of community-acquired pneumo-
nia with outpatient proton-pump inhibitor therapy: a systematic review
and meta-analysis. PLoS ONE 2015;10:e0128004.
37. Zhou B, Huang Y, Li H et al. Proton-pump inhibitors and risk of fractures:
an update meta-analysis. Osteoporos Int 2016;27:339–47.
38. Duyx B, Urlings MJE, Swaen GMH et al. Scientific citations favor posi-
tive results: a systematic review and meta-analysis. J Clin Epidemiol
2017;88:92–101.
39. Koelz HR. Gastric acid in vertebrates. Scand J Gastroenterol Suppl
1992;193:2–6.
40. Beasley DE, Koltz AM, Lambert JE et al. The evolution of stom-
ach acidity and its relevance to the human microbiome. PLoS ONE
2015;10:e0134116.
41. Giannella RA, Broitman SA, Zamcheck N. Influence of gastric acidity on
bacterial and parasitic enteric infections. A perspective. Ann Intern Med
1973;78:271–6.
42. Seto CT, Jeraldo P, Orenstein R et al. Prolonged use of a proton pump
inhibitor reduces microbial diversity: implications for Clostridium difficile
susceptibility. Microbiome 2014;2:42.
43. Freedberg DE, Salmasian H, Friedman C et al. Proton pump inhibitors
and risk for recurrent Clostridium difficile infection among inpatients. Am
J Gastroenterol 2013;108:1794–801.
44. Tleyjeh IM, Bin Abdulhak AA, Riaz M et al. Association between proton
pump inhibitor therapy and clostridium difficile infection: a contempo-
rary systematic review and meta-analysis. PLoS ONE 2012;7:e50836.
45. Wilson KH, Sheagren JN, Freter R. Population dynamics of ingested
Clostridium difficile in the gastrointestinal tract of the Syrian hamster. J
Infect Dis 1985;151:355–61.
46. Freedberg DE, Toussaint NC, Chen SP et al. Proton pump inhibitors alter
specific taxa in the human gastrointestinal microbiome: a crossover trial.
Gastroenterology 2015;149:883–885 e9.
47. Seekatz AM, Young VB. Clostridium difficile and the microbiota. J Clin
Invest 2014;124:4182–9.
48. Recker RR. Calcium absorption and achlorhydria. N Engl J Med
1985;313:70–3.
49. O'Connell MB, Madden DM, Murray AM et al. Effects of proton pump
inhibitors on calcium carbonate absorption in women: a randomized
crossover trial. Am J Med 2005;118:778–81.
50. Bo-Linn GW, Davis GR, Buddrus DJ et al. An evaluation of the impor-
tance of gastric acid secretion in the absorption of dietary calcium. J Clin
Invest 1984;73:640–7.
51. Ensrud KE, Duong T, Cauley JA et al. Low fractional calcium absorption
increases the risk for hip fracture in women with low calcium intake.
Study of Osteoporotic Fractures Research Group. Ann Intern Med
2000;132:345–53.
52. Wright MJ, Sullivan RR, Gaffney-Stomberg E et al. Inhibiting gastric acid
production does not affect intestinal calcium absorption in young, healthy
individuals: a randomized, crossover, controlled clinical trial. J Bone
Miner Res 2010;25:2205–11.
53. Hansen KE, Jones AN, Lindstrom MJ et al. Do proton pump inhibitors
decrease calcium absorption? J Bone Miner Res 2010;25:2510–9.
54. Fraser LA, Leslie WD, Targownik LE et al. The effect of proton pump
inhibitors on fracture risk: report from the Canadian Multicenter Osteo-
porosis Study. Osteoporos Int 2013;24:1161–8.
55. Targownik LE, Lix LM, Leung S et al. proton pump inhibitor use is not
associated with osteoporosis or accelerated bone mineral density loss.
Gastroenterology 2009;138:896–904
© 2018 by the American College of Gastroenterology
Discontinuing Long-Term PPI Therapy 527
56. Valles P, Lapointe MS, Wysocki J et al. Kidney vacuolar H+-ATPase:
physiology and regulation. Semin Nephrol 2006;26:361–74.
57. Tuukkanen J, Vaananen HK. Omeprazole, a specific inhibitor of H+-K+-
ATPase, inhibits bone resorption in vitro. Calcif Tissue Int 1986;38:
123–5.
REVIEW
58. Yang Q, Li G, Singh SK et al. Vacuolar H+-ATPase B1 subunit mutations
that cause inherited distal renal tubular acidosis affect proton pump as-
sembly and trafficking in inner medullary collecting duct cells. J Am Soc
Nephrol 2006;17:1858–66.
59. Badiola N, Alcalde V, Pujol A et al. The proton-pump inhibitor lansopra-
zole enhances amyloid beta production. PLoS ONE 2013;8:e58837.
60. Gilard M, Arnaud B, Cornily JC et al. Influence of omeprazole on the
antiplatelet action of clopidogrel associated with aspirin: the randomized,
double-blind OCLA (Omeprazole CLopidogrel Aspirin) study. J Am Coll
Cardiol 2008;51:256–60.
61. Bundhun PK, Teeluck AR, Bhurtu A et al. Is the concomitant use of clopi-
dogrel and proton pump inhibitors still associated with increased adverse
cardiovascular outcomes following coronary angioplasty?: a systematic
review and meta-analysis of recently published studies (2012–2016). BMC
Cardiovasc Disord 2017;17:3.
62. Douglas I, Evans S, Hingorani A et al. Clopidogrel and the interaction
with proton pump inhibitors: a comparison between cohort and within
person study designs. BMJ 2012;345:e4388.
63. O'Donoghue ML, Braunwald E, Antman EM et al. Pharmacodynamic
effect and clinical efficacy of clopidogrel and prasugrel with or without
a proton-pump inhibitor: an analysis of two randomised trials. Lancet
2009;374:989–97.
64. Freedberg DE, Kim LS, Yang YX. The risks and benefits of long-term
use of proton pump inhibitors: expert review and best practice advice
from the american gastroenterological association. Gastroenterology
2017;152:706–15.
65. Caro JJ, Salas M, Ward A. Healing and relapse rates in gastroesophageal
reflux disease treated with the newer proton-pump inhibitors lanso-
prazole, rabeprazole, and pantoprazole compared with omeprazole,
ranitidine, and placebo: evidence from randomized clinical trials. Clin
Ther 2001;23:998–1017.
66. Molina-Infante J, Ferrando-Lamana L, Ripoll C et al. Esophageal eosino-
philic infiltration responds to proton pump inhibition in most adults. Clin
Gastroenterol Hepatol 2011;9:110–7.
67. Schoepfer AM, Safroneeva E, Bussmann C et al. Delay in diagnosis of
eosinophilic esophagitis increases risk for stricture formation in a time-
dependent manner. Gastroenterology 2013;145:1230–6 e1-2.
68. Hawkey CJ, Karrasch JA, Szczepanski L et al. Omeprazole compared
with misoprostol for ulcers associated with nonsteroidal antiinflamma-
tory drugs. Omeprazole versus Misoprostol for NSAID-induced
Ulcer Management (OMNIUM) Study Group. N Engl J Med 1998;338:
727–34.
69. Yeomans ND, Tulassay Z, Juhasz L et al. A comparison of omeprazole
with ranitidine for ulcers associated with nonsteroidal antiinflammatory
drugs. Acid Suppression Trial: Ranitidine versus Omeprazole for NSAID-
associated Ulcer Treatment (ASTRONAUT) Study Group. N Engl J Med
1998;338:719–26.
70. Targownik LE, Metge CJ, Leung S et al. The relative efficacies of gastro-
protective strategies in chronic users of nonsteroidal anti-inflammatory
drugs. Gastroenterology 2008;134:937–44.
71. Andrade C, Sandarsh S, Chethan KB et al. Serotonin reuptake inhibi-
tor antidepressants and abnormal bleeding: a review for clinicians and a
reconsideration of mechanisms. J Clin Psychiatry 2010;71:1565–75.
72. Lanas A, Garcia-Rodriguez LA, Arroyo MT et al. Risk of upper gastro-
intestinal ulcer bleeding associated with selective cyclo-oxygenase-2
inhibitors, traditional non-aspirin non-steroidal anti-inflammatory drugs,
aspirin and combinations. Gut 2006;55:1731–8.
73. Lanza FL, Chan FK, Quigley EM. Guidelines for prevention of NSAID-
related ulcer complications. Am J Gastroenterol 2009;104:728–38.
74. Rostom A, Moayyedi P, Hunt R. Canadian consensus guidelines on
long-term nonsteroidal anti-inflammatory drug therapy and the need
for gastroprotection: benefits versus risks. Aliment Pharmacol Ther
2009;29:481–96.
75. Ouatu-Lascar R, Fitzgerald RC, Triadafilopoulos G. Differentiation and
proliferation in Barrett's esophagus and the effects of acid suppression.
Gastroenterology 1999;117:327–35.
76. Hu Q, Sun TT, Hong J et al. Proton pump inhibitors do not reduce the
risk of esophageal adenocarcinoma in patients with Barrett's esophagus: a
systematic review and meta-analysis. PLoS ONE 2017;12:e0169691.
The American Journal of GASTROENTEROLOGY
 528 Targownik
77. Masclee GM, Coloma PM, Spaander MC et al. NSAIDs, statins, low-dose
aspirin and PPIs, and the risk of oesophageal adenocarcinoma among pa-
tients with Barrett's oesophagus: a population-based case–control study.
BMJ Open 2015;5:e006640.
78. Singh S, Garg SK, Singh PP et al. Acid-suppressive medications and risk
REVIEW
of oesophageal adenocarcinoma in patients with Barrett's oesophagus: a
systematic review and meta-analysis. Gut 2014;63:1229–37.
79. Shaheen NJ, Falk GW, Iyer PG et al. ACG Clinical Guideline: diag-
nosis and management of Barrett's esophagus. Am J Gastroenterol
2016;111:30–50quiz 51
80. Spechler SJ, Sharma P, Souza RF et al. American Gastroenterological
Association medical position statement on the management of Barrett’s
esophagus. Gastroenterology 2011;140:1084–91.
81. Metz DC, Comer GM, Soffer E et al. Three-year oral pantoprazole admin-
istration is effective for patients with Zollinger–Ellison syndrome and
other hypersecretory conditions. Aliment Pharmacol Ther 2006;23:
437–44.
82. Maton PN, Vinayek R, Frucht H et al. Long-term efficacy and safety of
omeprazole in patients with Zollinger–Ellison syndrome: a prospective
study. Gastroenterology 1989;97:827–36.
83. Hirschowitz BI, Mohnen J, Shaw S. Long-term treatment with lansopra-
zole for patients with Zollinger–Ellison syndrome. Aliment Pharmacol
Ther 1996;10:507–22.
84. Weber HC, Venzon DJ, Lin JT et al. Determinants of metastatic rate and
survival in patients with Zollinger-Ellison syndrome: a prospective long-
term study. Gastroenterology 1995;108:1637–49.
85. Pinto-Sanchez MI, Yuan Y, Bercik P et al. Proton pump inhibitors for
functional dyspepsia. Cochrane Database Syst Rev 2017;3:CD011194.
86. Sigterman KE, van Pinxteren B, Bonis PA et al. Short-term treatment
with proton pump inhibitors, H2-receptor antagonists and prokinetics for
gastro-oesophageal reflux disease-like symptoms and endoscopy negative
reflux disease. Cochrane Database Syst Rev 2013;(5):CD002095.
87. Sifrim D, Zerbib F. Diagnosis and management of patients with reflux
symptoms refractory to proton pump inhibitors. Gut 2012;61:1340–54.
88. Vaezi MF. Laryngeal manifestations of gastroesophageal reflux disease.
Curr Gastroenterol Rep 2008;10:271–7.
89. Ford CN. Evaluation and management of laryngopharyngeal reflux.
JAMA 2005;294:1534–40.
90. Qadeer MA, Swoger J, Milstein C et al. Correlation between symptoms and
laryngeal signs in laryngopharyngeal reflux. Laryngoscope 2005;115:1947–52.
91. Friedman M, Maley A, Kelley K et al. Impact of pH monitoring on
laryngopharyngeal reflux treatment: improved compliance and symptom
resolution. Otolaryngol Head Neck Surg 2011;144:558–62.
92. Wei C. A meta-analysis for the role of proton pump inhibitor therapy
in patients with laryngopharyngeal reflux. Eur Arch Otorhinolaryngol
2016;273:3795–801.
93. Heidelbaugh JJ, Kim AH, Chang R et al. Overutilization of proton-pump
inhibitors: what the clinician needs to know. Ther Adv Gastroenterol
2012;5:219–32.
The American Journal of GASTROENTEROLOGY
94. Heidelbaugh JJ, Goldberg KL, Inadomi JM. Magnitude and economic
effect of overuse of antisecretory therapy in the ambulatory care setting.
Am J Manag Care 2010;16:e228–e234.
95. Ladd AM, Panagopoulos G, Cohen J et al. Potential costs of inappropriate
use of proton pump inhibitors. Am J Med Sci 2014;347:446–51.
96. Talley NJ, Lauritsen K, Tunturi-Hihnala H et al. Esomeprazole 20 mg
maintains symptom control in endoscopy-negative gastro-oesophageal
reflux disease: a controlled trial of “on-demand” therapy for 6 months.
Aliment Pharmacol Ther 2001;15:347–54.
97. Boghossian TA, Rashid FJ, Thompson W et al. Deprescribing versus
continuation of chronic proton pump inhibitor use in adults. Cochrane
Database Syst Rev 2017;3:CD011969.
98. Hansen AN, Wahlqvist P, Jorgensen E et al. Six-month management of
patients following treatment for gastroesophageal reflux disease symp-
toms—a Norwegian randomized, prospective study comparing the costs
and effectiveness of esomeprazole and ranitidine treatment strategies in a
general medical practitioners setting. Int J Clin Pract 2005;59:655–64.
99. Inadomi JM, McIntyre L, Bernard L et al. Step-down from multiple- to
single-dose proton pump inhibitors (PPIs): a prospective study of patients
with heartburn or acid regurgitation completely relieved with PPIs. Am J
Gastroenterol 2003;98:1940–4.
100. Armstrong D. Systematic review: persistence and severity in gastro-oe-
sophageal reflux disease. Aliment Pharmacol Ther 2008;28:841–53.
101. Waldum HL, Arnestad JS, Brenna E et al. Marked increase in gastric acid
secretory capacity after omeprazole treatment. Gut 1996;39:649–53.
102. Fossmark R, Johnsen G, Johanessen E et al. Rebound acid hypersecretion
after long-term inhibition of gastric acid secretion. Aliment Pharmacol
Ther 2005;21:149–54.
103. Reimer C, Sondergaard B, Hilsted L et al. Proton-pump inhibitor therapy
induces acid-related symptoms in healthy volunteers after withdrawal of
therapy. Gastroenterology 2009;137:80–7.87 e1
104. Niklasson A, Lindstrom L, Simren M et al. Dyspeptic symptom develop-
ment after discontinuation of a proton pump inhibitor: a double-blind
placebo-controlled trial. Am J Gastroenterol 2010;105:1531–7.
105. Reimer C, Bytzer P. Discontinuation of long-term proton pump inhibi-
tor therapy in primary care patients: a randomized placebo-controlled
trial in patients with symptom relapse. Eur J Gastroenterol Hepatol
2010;22:1182–8.
106. Inadomi JM, Jamal R, Murata GH et al. Step-down management of gas-
troesophageal reflux disease. Gastroenterology 2001;121:1095–100.
107. Cao F, Chen C, Wang M et al. An updated meta-analysis of controlled
observational studies: proton-pump inhibitors and risk of Clostridium
difficile infection. J Hosp Infect 2017;98:4–13.
108. Jung SB, Nagaraja V, Kapur A et al. Association between vitamin B12
deficiency and long-term use of acid-lowering agents: a systematic review
and meta-analysis. Intern Med J 2015;45:409–16.
109. Nochaiwong S, Ruengorn C, Awiphan R et al. The association between
proton pump inhibitor use and the risk of adverse kidney outcomes: a sys-
tematic review and meta-analysis. Nephrol Dial Transplant 2017;33:331–42.
VOLUME 113 | APRIL 2018 www.nature.com/ajg