Understanding the Natural History of the Disease

Introduction

Despite its high incidence, the natural history of diverticulitis remains incompletely understood. Although compelling associations have been described between inherent and environmental factors, understanding how these individual risks combine to produce different phenotypes of disease remains a challenge. Scientific knowledge gaps create clinical challenges. Physicians continue to struggle to understand the global variation of diverticular disease, the rising incidence in young patients, the proper role of antibiotics, and the ideal selection of surgical patients. Improved etiologic understanding of this disease has the potential to improve clinical care for this common disease.

Background

Diverticulosis is an asymptomatic, acquired condition which serves as the precursor lesion for diverticular disease. Asymptomatic diverticulosis is incidentally identified on colonoscopy, crosssectional imaging, or contrast enema studies. Pseudodiverticula herniate through the antimesenteric aspect of the circular muscle adjacent to small, penetrating blood vessels. Diverticula occur in the small bowel and rectum, but colonic diverticulosis is by far the most common and is the focus of this chapter.

Diverticulitis, defined as inflammation and superinfection of the diverticula, is the most common manifestation of diverticular disease. Diverticulitis includes a spectrum of acute and chronic presentations. Most patients with acute diverticulitis have an isolated episode of segmental colonic inflammation. However, 10-30% of these patients will have recurrent episodes, some life-limiting enough to warrant elective colectomy as a form of secondary prevention¹. Of patients requiring hospitalization, 10-25% will have an associated abscess. These patients are more likely to suffer recurrence and require surgery². A small minority present with free perforation of the colon and peritonitis necessitating urgent surgery. Diverticulitis patients may also present with chronic inflammatory complications such as colonic stricture and fistula, which require elective surgical intervention.

Two less common inflammatory manifestations include symptomatic uncomplicated diverticular disease (SUDD) and segmental colitis associated with diverticular disease (SCAD). These entities overlap with inflammatory bowel disease and irritable bowel syndrome. Diagnosis is difficult and the role of surgery is controversial. Diverticula may also manifest symptoms through non-inflammatory erosion of adjacent vessels, producing brisk, painless, typically self-limited hemorrhage³.

Epidemiology

Diverticular disease demonstrates wide global variation, being most common in North America and Europe. Diverticulosis is ubiquitous in older white individuals. In North America, >50% of patients over age 60 have diverticulosis on screening endoscopy⁴. Although <5% of individuals with diverticulosis will develop diverticulitis, the ubiquity of the precursor lesion leads to a high burden of disease. Every year in the United States 1.9 outpatient visits, 340,000 emergency room encounters, and nearly 200,000 inpatient admissions are attributed to diverticulitis, resulting in billions of dollars of expense⁵. Diverticulitis is considered a disease of industrialized society. Although population-wide retrospective comparisons are challenging, there is a clear overall increase in diverticulosis from a prevalence of <10% in autopsy studies from the early 20^th century. However, even in long-established industrialized societies incidence has not plateaued. Diverticulitis is increasing in the United States, particularly in younger patients. Between 1980 and 2007, diverticulitis increased by 132% among patients under age 50⁶. In the United States, incidence is higher at Northern latitudes, in urban areas⁷, and among white people, compared to Black, and Asian people⁸.

The relatively lower incidence of diverticular disease in Africa compared to Europe and North America has been noted since the 1960s. Population-wide screening endoscopy is not available in most African countries, so determining the true incidence of diverticulosis is challenging. In a small case series, diverticulosis was present in <10%⁹. As these individuals were symptomatic patients, mainly with hematochezia, the true incidence in the asymptomatic population is likely lower. However, diverticulitis is reported with increasing frequency at some African hospitals¹⁰.

Incidence of diverticulosis varies throughout Asia and is anatomically distinct. In European-ancestry individuals, diverticulosis is almost exclusively left-sided; the sigmoid colon is affected in >95% of patients. Diverticulosis is isolated to the sigmoid colon in 60% of cases and incidence decreases proximally along the colon, with 10% of patients having total colonic diverticulosis⁴. However, in Asian populations diverticulosis is right-sided in up to 80% of cases. The reason for this anatomic variation is unknown. Availability of population-wide screening endoscopy varies between countries, but reported incidence rates of diverticulosis range from <5% in mainland China¹¹, 12-14% in South Korea and Taiwan^12,13, 28% in Thailand¹⁴, and 32% in older adults in India¹⁵.

Etiology

Despite high disease incidence, the development of diverticulosis and the triggers that produce diverticulitis remain incompletely understood. Factors contributing to this knowledge gap include the overlapping causality between diverticulosis-risk and diverticulitis-risk, the inherent challenge in studying development of an asymptomatic lesion, and the lack of a practical animal model. Nonetheless, research has demonstrated compelling associations between potential risk factors and diverticular disease. Although diverticulitis is traditionally understood as an environmental disease triggered by diet and lifestyle, accumulating evidence suggests genetics play a substantial role. A more accurate paradigm therefore is that, like most common diseases, diverticulitis is a complex trait caused by the combined effects of inherent and environmental risk factors.

Genetics

Diverticulitis is associated with rare Mendelian connective tissue disorders including Marfan Syndrome, Ehler-Danlos Syndrome, polycystic kidney disease, and Williams Syndrome^16-18. Although these syndromes are rare, they indicate that connective tissue pathophysiology may be etiological. Diverticulitis is clinically associated with higher rates of common connective tissue diseases such as pelvic organ prolapse, hernia, and aortic aneurysm¹⁹. Family history of diverticulitis has been implicated as a risk factor for severe disease and recurrence²⁰.

Heritability of diverticulitis has been formally studied through twin studies, classic and powerful epidemiologic tools for discriminating between genetic and environmental risks. Twin pairs presumably share similar levels of environmental risk regardless if they are monozygotic or dizygotic. However, monozygotic twins share a much greater proportion of their DNA, making estimates of environmental versus genetic risk possible when large enough cohorts of twins are compared. Two large, modern twin studies of diverticular disease have been carried out in well-described Scandinavian populations. The Swedish Twin Registry of 104,452 twins contains 2,296 individuals with diverticular disease. Odds ratio for disease concordance was 7.15 in monozygotic twins versus 3.20 in same sex dizygotic twins²¹. Similarly, a Danish study found the relative risk of the second twin developing diverticular disease was 14.5 for monozygotic twins and 5.5 for dizygotic twins²². Taken together, the studies suggest that 40-53% of risk of diverticular disease is attributable to heritable factors. This is similar to heritability indices for Crohn’s disease and higher than that for ulcerative colitis in twin studies.

GWAS Data

Genetic contribution to disease is often thought of in terms of Mendelian traits like cystic fibrosis or familial adenomatous polyposis – relatively rare syndromes driven by single, powerful deleterious mutations. With rare exception²³, diverticulitis is not a Mendelian trait. Like most common diseases, such as coronary artery disease or diabetes, the genetic background is a risk factor, but not as a discrete heritable mutation. Appreciating the role of genetics in common diseases is nothing new; physicians have long collected family histories as a proxy for inherited risk. In a complex trait like diverticulitis therefore, revealing its genetic architecture is not a matter of identifying a deleterious driver mutation, but rather describing the potential contributions of multiple single nucleotide polymorphisms (SNPs) to development of this phenotype. The first genome-wide association study (GWAS) to attempt this was performed in Icelandic and Danish cohorts and identified three associated SNPs²⁴. A subsequent larger GWAS used the UK Biobank and a University of Michigan-based separate replication cohort to identify 39 more genome-wide significant SNPs²⁵. Associated genes included many with known roles in connective tissue biology, intestinal motility, inflammation, and vascular biology. Cell and pathway analysis revealed gene set enrichment in mesenchymal stem cells, mesodermal derivatives, and vascular development. These analyses provide many compelling possibilities for further study.

Structural/Functional Changes to the Colon

Structural and functional changes have been demonstrated in the colon of diverticular disease patients with a variety of experimental techniques. Surgeons appreciate stiffness and thickening of the colonic wall in diverticular resections, often using these changes to guide proximal colonic transection. Histologically, diverticular colons demonstrate elastosis and increased connective tissue deposition in the muscle of the colonic wall²⁶. Increased collagen cross-linking reduces intestinal compliance and decreases strength²⁷. Decreased numbers of interstitial cells of Cajal, altered components of the enteric nervous system, perturbation of serotonin signaling, altered neurotransmitters, and lower nitric oxide reactivity have all been described in colons afflicted with diverticular disease^28-30. Functional studies reveal higher intraluminal pressures, increased peristalsis, and segmental contractions of diverticular colons in vivo³¹. These studies employ diverse methodologies and study different sub-phenotypes of diverticular disease with, in most cases, little mechanistic follow up to clarify observed associations. However, the greatest challenge to interpreting these data is in determining which findings are causal factors and which are secondary to the disease itself. However, taken together, at a minimum they indicate that diverticular disease can alter the intestine across multiple cell types and physiologic functions.

Diet

Dietary fiber is by far the most established and recognized risk factor for diverticular disease. In 1969, Burkitt published his observations that middle-aged Ugandans suffered much less frequently from diseases such as diverticulitis, appendicitis, atherosclerosis, diabetes, varicose veins, and multiple cancers than their counterparts in England^32,33. He attributed this to the high fiber diet consumed by rural Africans. He and his colleagues further refined this hypothesis by demonstrating that a fiberdeficient diet led to decreased stool bulk, longer colon transit times, and, in some of the earliest microbiome work, altered colonic bacterial composition^33,34. Dietary fiber continues to be recognized as a protective factor for multiple cardiovascular, inflammatory, and neoplastic conditions. Microbial digestion of dietary fiber produces short chain fatty acids which exert local protective effects on the colonic mucosa and alter metabolism through enteric hormone production³⁵.

Dietary fiber, therefore could be hypothesized to prevent against the development of diverticulosis through improved colon mechanics or to prevent diverticulitis through an alternate protective mechanism. Despite the mechanistic paradigm of low fiber leading to constipation and straining, evidence for fiber’s prevention of diverticulosis is inconclusive. Large colonoscopy-based studies found no correlation between dietary fiber intake and asymptomatic diverticulosis³⁶. However, well-described longitudinal cohort studies have confirmed dietary fiber as protective against diverticulitis. Longitudinal follow up of more than 50,000 women in the Nurses Health Study found that those in the highest quintile of fiber intake were protected against incident diverticulitis, odds ratio = 0.86 (95% confidence interval: 0.78-0.95; P-trend = 0.002)³⁷. Separate, large cohort studies have confirmed that diets high in whole grains and low in red meat confer decreased risk of diverticulitis³⁸.

A key deficiency in diverticular disease research is the lack of a practical animal model. However, some of the only work with model organisms in diverticular disease has been performed to elucidate the role of fiber. Diverticulosis can be produced in rats fed low fiber diets over a period of 18 months. Similar to the protective effects of fiber seem in humans, rats on high fiber diets had lower total body weight and fewer sporadic tumors in addition to lacking diverticula. However, the role of fiber is complex. Development of diverticula in rats was also influenced by maternal diet. Offspring of mothers fed a low fiber diet had a significantly higher incidence of diverticulosis (42%) than offspring of mothers on a high fiber diet (21%, p <0.01)³⁹. If these data can be extrapolated to human disease, they suggest that the role of fiber is not purely mechanical and the role of heredity is not purely genetic.

Microbiome

Multiple compelling lines of reasoning suggest a role for the microbiome in diverticulitis. Colon bacteria are critical to metabolism of dietary fiber into the short chain fatty acids that maintain colonic health. High fiber diets promote increased diversity of colonic bacteria³⁵. Western diet alters the microbiome and is associated with diverticular disease. However, data to support the microbiome as a causal factor in diverticular disease are yet to be obtained. Challenges include reluctance to obtain colonoscopic samples during an acute attack, pre-treatment with antibiotics, challenges in timing sample collection, and multiple studies in the controversial clinical entity SUDD. At one end of the clinical spectrum, mucosal biopsies taken from asymptomatic patients undergoing first screening colonoscopy did not identify microbiome differences between individuals with and without diverticulosis⁴⁰. At the other end of severity, comparison of the microbiome existing in diseased versus normal portions of elective sigmoid colectomy specimens revealed distinct differences in the microbiota living in the adjacent tissues⁴¹, but it is impossible to determine whether these changes are causal or secondary to the disease. An anticipated clinical study of patients with asymptomatic, mild, and severe diverticulitis failed to accrue patients and was terminated. Similar to much of the pathophysiologic data in diverticular disease, the microbiome remains an interesting, but incompletely understood, potential risk factor.

Other Environmental Risk Factors

Exposures other than diet have been associated with diverticular disease. Case control and cohort studies of varying sizes and quality have demonstrated associations between diverticulitis and obesity, decreased physical activity, vitamin D levels, UV light exposure, seasonality, and use of common medications^1,7,42-44. It has been estimated that widespread adoption of a healthy lifestyle including <4 servings of red meat a week, 23g of daily fiber, vigorous physical activity, body mass index <25, and abstaining from smoking could prevent 50% of diverticulitis cases⁴⁵. Although these data have yet to be tested prospectively, they provide patients seeking to avoid diverticulitis with healthy lifestyle recommendations. Disentangling the multiple potential inherent and environmental risk factors contributing to diverticulitis remains an ongoing challenge for researchers.

Summary.

Diverticulitis is common disease, but despite this its pathophysiology remains incompletely understood. Multiple inherent and environmental associations have been described without clear delineation of causality. This lack of etiologic understanding has clinical ramifications. Diverticulitis cannot be predicted or prevented. Standard of care therapy, antibiotics, has been called in to serious question by randomized controlled trials. No disease-specific drugs have been developed or tested. Elective surgical recommendations are based on surgeon and patient factors without scientific data to aid in risk stratification. Therefore, further basic and translational research in diverticulitis represents a great opportunity to increase knowledge and improve care in this common and morbid disease.