Cohort of patients with a history of hemorrhoidal disease
Initiated by the Department of General and Thoracic Surgery and the biobank PopGen (22) from the Medical Faculty of Kiel University, a cohort of patients with symptomatic hemorrhoids and the need of invasive treatment was established, primarily to investigate the genetic underpinning of hemorrhoidal disease. In brief, between January 2016 and December 2017, individuals with a prior diagnosis of symptomatic and treated hemorrhoids were identified through medical records of 5 hospitals and offices in Northern Germany using the procedural codes for invasive treatment and surgery. Patients that were exclusively treated conservatively were not included in this study as the aim was to recruit patients with a strong phenotype of advanced hemorrhoids. Eligible patients were contacted via mail by their treating physicians and the PopGen biobank. After agreeing to participate, patients received a study set including a questionnaire about clinical and sociodemographic characteristics and a tube set for blood collection. A subset of study participants was asked to fill out an additional questionnaire concerning their dietary behavior (23) and usual physical activity (24). The study protocol was approved by the ethics committee of the Medical Faculty of Kiel University and written informed consent was obtained from all study participants.
A total of 1070 patients agreed to participate in the study and 393 individuals filled out the questionnaire on diet and physical activity. One participant was excluded because he had reported to have been suffering from anorexia nervosa and bulimia for several years, leading to a final analytical sample of 392 individuals for the nutrition-related analyses. For the analyses focusing on physical activity, an additional six patients had to be excluded due to missing information on some physical activities, thus, leading to an analytical sample size of 386 study participants.
Dietary behavior and physical activity of the hemorrhoid patients were compared to a reference (control) sample from the biobank PopGen. The study design of this control cohort has been reported in detail elsewhere (25,26). Briefly, the PopGen control cohort is a prospective cohort study that recruited individuals from the general population by random selection through local population registries and comprised a baseline and two follow-up assessments (2005-2018). Besides undergoing physical examinations in the PopGen study center, the study participants were asked to fill in a standardized questionnaire covering a broad range of clinical information, such as medical history, as well as sociodemographic and lifestyle factors. For comparison with the hemorrhoid patient cohort, data from the control cohort’s second follow-up (2016-2018) were used because at this examination cycle the same version of the diet and physical activity questionnaire as in the hemorrhoid study was applied (explained in detail below).
In the control cohort, information on hemorrhoids were obtained by questionnaire at the first follow-up which was about 5 years prior to the second follow-up, when the information on diet and physical activity was assessed. A total of 83 individuals from the reference cohort who reported hemorrhoids at first follow-up were excluded from the analyses. Unfortunately, information on hemorrhoids was not obtained at the second follow-up examination.
Controls were matched (1:1) to hemorrhoid patients based on age (using age categories of 5-year-ranges) and sex. For 297 and 288 hemorrhoids patients, appropriately matching controls could be identified and included for the diet and for the physical activity analyses, respectively.
Assessment of Diet and physical activity
Usual dietary intake was assessed with a standardized and validated semi-quantitative food frequency questionnaire (FFQ) that was developed and analyzed by the Department of Epidemiology at the German Institute of Human Nutrition in Potsdam-Rehbrücke (23). The FFQ measures the consumption frequencies of a list of foods and beverages in given portion sizes during the last 12 months. The frequencies were provided in four to eleven options ranging from ‘never’ and ‘once a day’ to ‘eleven times a day or more frequently’. The quantities were quoted as portions, grams, slices, spoons, or milliliters. Out of all FFQ data, intakes in grams per day for each food or food group, daily total energy intake and nutrient content were computed for each participant.
An additional validated questionnaire assessing usual physical activities during the last 12 months was applied (24). Responding to these questions, participants stated the average hours per week that they spent with different activities, including walking, cycling, sports, and gardening, each separately for summer and winter. Weekly hours of housework (e.g. washing, cleaning, cooking) and home repair (do-it-yourself work), as well as stair climbing defined as number of floors per day were enquired non-seasonally. In addition, hours of sleeping at night and at day, respectively, and hours per week spent watching TV were enquired for summer and winter times. To consider different intensities of different activity types, corresponding metabolic equivalent of task (MET) values (27), in accordance to the 2000 compendium of Physical Activity (28), were multiplied with the number of hours per week of each activity (mean number of hours from summer and winter activities, where applicable). One MET is defined as energy expenditure for sitting quietly (resting metabolic rate); and MET-values are the ratio of the metabolic rate for a specific activity divided by the resting metabolic rate (28). The applied MET-values were 3.0 for walking, 6.0 for cycling, 6.0 for sports, 4.0 for gardening, 3.0 for housework, 4.5 for home repair, and 8.0 for stair climbing (27,28). Summing up the MET-hours of walking, cycling, sports, gardening, housework, home repair, and stair climbing resulted in MET-hours per week of total physical activity.
Assessment of clinical and sociodemographic data
Most clinical and sociodemographic characteristics (age, sex, age at hemorrhoid diagnosis, body weight and height, pregnancy, grade of hemorrhoids, hemorrhoid therapy, current affliction, fecal incontinence) were self-reported by the study participants through filling in the general questionnaire. In a subgroup of 294 hemorrhoid patients, information on the grade of hemorrhoids was also available from clinical records so that for these patients, the physician-reported degree of hemorrhoids was used for analyses, whereas for all other patients self-reported data were taken into account. Out of the standardized questions regarding fecal incontinence, the Cleveland Clinic Fecal Incontinence Score/Wexner (CCFIS) was determined with a higher CCFIS (range 0-20) indicating a higher grade of incontinence (29). Body Mass Index (BMI) was defined as weight in kilograms divided by height in meters squared.
First, characteristics of the total cohort of hemorrhoid patients and of the matched hemorrhoid patient and control cohorts were determined, respectively.
Second, average intakes of nutrients, foods, and food groups, separately for men and women, as well as average amounts of physical activities were calculated as median and interquartile range (IQR) for the matched hemorrhoid and control cohorts. Differences in food consumption and physical activities between hemorrhoid patients and controls were tested for statistical significance with the Wilcoxon ranksum test.
Third, within the patient cohort with a history of hemorrhoidal disease, defined clinical characteristics were related to dietary intake and physical activity. Specifically, logistic regression analyses were applied to estimate the associations of hemorrhoid severity (grade 2-4 or unknown grade vs. grade 1), type of first hemorrhoid treatment (surgery vs. rubber band ligation/sclerotherapy), necessity of further treatment after hemorrhoid surgery (yes vs. no), willingness to undergo further hemorrhoid surgery if necessary (no vs. yes), current affliction at the anus (yes vs. no), and the CCFIS for fecal incontinence (6-10 or 11-15 or 16-20 vs. 0-5) with food intake and physical activity. The intake of selected nutrients, foods or food groups and the different physical activities were modeled as above vs. below the cohort-specific median (binary outcome) for these analyses. To control for confounding, all logistic regression analyses were adjusted for age, sex, and BMI. Dietary analyses were additionally adjusted for total energy intake.
The statistical analyses were conducted using SAS version 9.4 software (SAS Institute, Inc., NC, USA). Two-sided p values of < 0.05 were considered statistically significant.