The incidence of diverticulitis has been associated with geographic and seasonal variation. Low levels of circulating vitamin D are associated with diverticulitis. We investigated the association between UV light and diverticulitis.
We identified nonelective diverticulitis admissions in the Nationwide Inpatient Sample and linked hospital locations to UV data. We examined UV exposure in relation to risk of admission for diverticulitis. We identified geographic and seasonal trends among 226 522 nonelective admissions for diverticulitis. Compared with high-UV areas, low-UV areas had a higher rate of diverticulitis (751.8 vs 668.1 per 100 000 admissions; P < .001), diverticular abscess (12.0% vs 9.7%; P < .001), and colectomy (13.5% vs 11.5%; P < .001). We also observed significant seasonal variation, with a lower rate of diverticulitis in the winter (645 per 100 000) compared with the summer (748 per 100 000) (P < .001). The summer increase was more evident in areas with the greatest UV fluctuation vs areas with the least UV fluctuation (120 vs 70 per 100 000; P = .01).
Conclusions and Relevance
Low UV light exposure is associated with an increased rate of diverticulitis admissions and greater seasonal variation. Because UV exposure largely determines vitamin D status, these findings support a role for vitamin D in the pathogenesis of diverticulitis.
Diverticulitis is a common disease with incompletely understood pathophysiologic features. Diverticulitis results in 200 000 hospital admissions per year,1 and its sequela generate substantial morbidity. Diverticulosis is ubiquitous in older American adults,2 but the inciting event that leads to diverticulitis is unknown.
Geographic3 and seasonal4 variation in diverticulitis has been previously shown, with higher incidence in Western countries and in the summer. Prior studies have questioned traditional theoretical triggers for diverticulitis, such as nut and seed impaction in diverticula,5 and demonstrated novel associations with obesity,6 physical activity,7 and nonsteroidal anti-inflammatory drug use.8
Recently, we observed an association between low vitamin D levels and the development of diverticulitis.9 Vitamin D synthesis depends on UV light, which varies according to latitude, with greater annual fluctuation occurring farther from the equator. Associations exist between UV exposure, vitamin D levels, and several diseases. In the colon, low levels of vitamin D are associated with Crohn disease,10 colorectal cancer,11 and decreased cancer-specific survival.12 Residence in higher latitudes is associated with inflammatory bowel disease.13 In this study, we investigate a potential link between UV exposure as a determinant of vitamin D levels and diverticulitis.
We identified nonelective admissions for diverticulitis in the Nationwide Inpatient Sample, 2001-2005. We included patients who were older than 18 years with a nonelective admission associated with a billing code for diverticulitis (International classification of Diseases, Ninth Edition [ICD-9] code 562.11 or 562.13). Of the patients with diverticulitis, we identified those with intra-abdominal abscess (ICD-9 code 569.5 or 567.2x) or who required left-sided colorectal resection (ICD-9 code 45.75, 45.76, 48.62, or 48.63). We also identified a cohort of patients with diverticular hemorrhage (ICD-9 code 562.12). Institutional review board approval was given by Massachusetts General Hospital.
We linked Nationwide Inpatient Sample hospital zip codes to vitamin D action-spectrum UV irradiance to estimate the UV exposure experienced by an individual in the year before admission. We used a uniform measure of sun exposure based on the estimated number of minutes needed to achieve the UV equivalent of a 1000-IU oral dose of vitamin D (25-hydroxyvitamin D). This measurement has been previously described14 and was calculated using spectrophotometer-validated data. We categorized this measure into quartiles based on exposure times for Fitzpatrick skin type I.15 The time needed to achieve this dose during the summer is nearly uniform across the United States (2.7-5.7 minutes). Variation is greatest in the winter (6.7-103 minutes). Therefore, locations with the lowest winter irradiance experience the greatest seasonal fluctuation. Univariable analysis was performed with 2-tailed t tests. We used multivariable logistic regression to estimate the influence of season and UV exposure on admission for diverticulitis, adjusting for age, sex, race, and hospital location. Statistical analysis was performed using SAS Software, version 9.2 (SAS Institute).
From January 2001 through December 2005, we identified 32.4 million adult inpatient admissions within the National Inpatient Sample. Of these, 226 522 (0.7%) were nonelective admissions with a code for diverticulitis, yielding an overall rate of 699.8 per 100 000 total admissions. Baseline characteristics were similar between patients admitted to hospitals within different UV quartiles (eTable 1 in the Supplement). We observed higher rates of diverticulitis hospitalization among white patients, patients 60 years and older, and rural Americans (Table).
Rates of diverticulitis admissions were significantly higher in areas of low UV light (UV1 areas) compared with areas of high UV light (UV4 areas) (751.8 vs 668.1 per 100 000 admissions; P < .001). In subgroup analysis, this association was especially evident in 88 477 patients who were younger than 60 years (eTable 2 in the Supplement). Patients younger than 60 years in UV1 areas had 106 more diverticulitis admissions per 100 000 total admissions compared with those in UV4 areas. In contrast, patients 60 years and older in UV1 areas had only 37 more diverticulitis admissions per 100 000 total admissions compared with those in UV4 areas (Pheterogeneity <.001).
We also observed that rates of diverticulitis were highest in the third discharge quarter (June, July, and August) and lowest in the first (January, February, and March) (eTable 3 in the Supplement). The incidence of diverticulitis increased about 16% in the third quarter. Although seasonal variation was observed in all UV groups, we observed significantly greater variation in areas with lower winter UV irradiance and therefore greater annual fluctuation (eTable 3 in the Supplement and Figure, A). Admissions rose 17% in the summer in UV1 (darkest) areas vs 11% in UV4 (sunniest) areas (P = .01). Compared with individuals in UV1 areas, the multivariable odds ratio for admission was 0.94 (95% CI, 0.92-0.95) for those in UV4 areas and 0.82 (95% CI, 0.81-083) for those in UV3 areas. Tests of interaction between discharge quarter and UV exposure quartile demonstrated a greater effect of season on admissions in UV1 areas compared with UV4 areas (P = .047).
A, All patients. B, Patients 60 years and older vs younger than 60 years. C, Rural vs urban patients. UV1 indicates the darkest areas; UV4, the sunniest.
Certain subgroups had greater seasonal variation. White patients, patients 60 years and older, and rural patients demonstrated significantly higher rates of admission in the summer compared with their respective reference groups (eTable 3 in the Supplement and the Figure). Admissions among patients 60 years and older increased 20.7% in the summer compared with 13.3% among patients younger than 60 years (P < .001).
Rates of abscess and surgery among patients with diverticulitis were modestly but significantly higher in low UV areas. The incidence of diverticular abscess was significantly higher in UV1 areas (12.0%) compared with UV4 areas (9.7%) (P < .001). Similarly, rates of surgery were significantly higher in UV1 areas (13.5%) compared with UV4 areas (11.5%) (P < .001). We did not find seasonal variation in rates of surgery or abscess. No significant difference was found in diverticular hemorrhage admissions between UV1 and UV4 areas (282.0 vs 277.6 per 100 000; P = .80), and no seasonal variation was appreciated.
On multivariable analysis, increasing age, female sex, white race, and rural residence were significantly associated with admission for diverticulitis (eTable 3 in the Supplement). Compared with the third discharge quarter, all other quarters were inversely associated with admission (odds ratio, 0.87; 95% CI, 0.84-0.89 for the first quarter).
A large national database linked with detailed UV light data reveals significantly higher overall rates and greater seasonal variation in diverticulitis admissions in areas with low winter vitamin D–spectrum UV irradiance. Multivariable analysis supports greater influence of season in low UV areas, with a significant summer increase in admissions. Taken together, these data suggest that greater seasonal flux in UV irradiance is correlated with greater seasonal variation in diverticulitis admissions.
Previously, we have shown that low serum vitamin D levels are associated with subsequent diverticulitis. The underlying mechanism between UV irradiance and diverticulitis may be variation in vitamin D levels. In animal models, vitamin D deficiency worsens colitis16 and Salmonella infection.17 Vitamin D has been shown to reduce proinflammatory cytokines,18,19 and deficiency is associated with inflammation and neoplasia. Thus, the low levels of vitamin D may explain the increased rate of diverticulitis admissions, abscess, and colectomy observed in darker areas.
Our observed association between higher admission rates in the summer may appear inconsistent with the vitamin D hypothesis. However, an association between vitamin D deficiency and diverticulitis may require a prolonged latency. The association between low vitamin D level and diverticulitis may occur several months before clinical presentation. A similar pattern of low winter vitamin D levels followed by summer peaks of disease has been described for tuberculosis.20
We observed that there was greater geographic, yet lower seasonal, variation among younger patients. This finding may suggest that overall vitamin D deficiency is of greater significance in younger patients, while fluctuation has a greater effect on older patients. Seasonal variation was more pronounced among white than black patients, which could reflect greater annual fluctuation in vitamin D levels among white patients or may reflect differences in genetics, dietary patterns, or lifestyle habits that may mitigate the influence that UV exposure has on vitamin D levels.21 We also observed greater overall rates of admission and seasonality among rural hospitals. This finding may reflect greater variation in circulating vitamin D levels in rural areas due to more time spent outdoors.
Alternative explanations for the seasonal and geographic variation seen in this study could include an infectious etiologic aspect. Variation in the prevalence of diverticulosis could contribute to variation in diverticulitis, although we did not observe similar seasonal or geographic variation in diverticular hemorrhage. Variation in diet or other factors associated with diverticulitis, such as obesity, nonsteroidal anti-inflammatory drug use, and physical activity, could also potentially contribute to the observed trends. We controlled for race and age in our multivariable model, but other population-level differences could also influence our results.
The strengths of this study include use of a large national database, links with detailed UV irradiance data, and use of multivariable analyses to examine the interaction between season, geography, and diverticulitis. There are several limitations to this study. First, billing codes may misclassify some patients with diverticulitis. However, this proportion is unlikely to vary with geography or season. Second, our analysis only included inpatients, excluding the diverticulitis cases managed in the emergency department or outpatient setting. Third, important individual-level variables, such as diet and supplement use, as well as lifestyle habits, such as sun protection and socioeconomic status, are not well captured by the Nationwide Inpatient Sample.
Our study demonstrates that low winter UV light exposure is associated with higher rates of diverticulitis and that areas with greater annual UV fluctuation have larger seasonal variation. Because UV irradiance is associated with vitamin D levels, this study supports the hypothesis that low vitamin D levels and greater seasonal fluctuation in vitamin D levels may contribute to the pathophysiologic features of diverticulitis.
Accepted for Publication: August 7, 2014.
Corresponding Author: Andrew T. Chan, MD, MPH, Division of Gastroenterology, Massachusetts General Hospital, 55 Fruit St, Gray-Jackson 722, Boston, MA 02114 (firstname.lastname@example.org).
Published Online: November 19, 2014. doi:10.1001/jamasurg.2014.2049.
Author Contributions: Dr Chan had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Maguire, Song, Chan.
Drafting of the manuscript: Maguire, Chan.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: All authors.
Obtained funding: Chan.
Administrative, technical, or material support: Chan.
Supervision: Giovannucci, Chan.
Conflict of Interest Disclosures: Dr Chan reports having served as a consultant to Bayer HealthCare, Millennium Pharmaceuticals, Pfizer Inc, and Pozen Inc for work unrelated to the topic of this manuscript. No other conflicts of interest were reported.
Funding/Support: This work was supported by a Clinical Research Award from the American College of Gastroenterology (Dr Maguire) and grant K24 DK 098311 from the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, US Department of Health and Human Services (Dr Chan). Dr Chan is a Damon Runyon Clinical Investigator.
Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Previous Presentation: This study was presented at Digestive Diseases Week; May 3, 2014; Chicago, Illinois.
eTable 1. Characteristics of patients with nonelective admissions for diverticulitis by UV quartile.
eTable 2. Rate of nonelective admission for diverticulitis, by UV quartile and subgroup.
eTable 3. Rate of nonelective hospitalization for diverticulitis in summer (third quarter) vs winter (first quarter).
Maguire LH, Song M, Strate LL, Giovannucci EL, Chan AT. Association of Geographic and Seasonal Variation With Diverticulitis Admissions. JAMA Surg. 2015;150(1):74-77. doi:10.1001/jamasurg.2014.2049