Patients with inflammatory bowel diseases (IBDs), including Crohn's disease (CD) and ulcerative colitis (UC), may lose weight during flares of the disease because of intestinal inflammation and subsequent malabsorption or decreased oral intake (1,2). Nutritional status is therefore essential in treating patients with IBD, particularly those with extensive small bowel disease with a decreased ability to absorb nutrients. For patients with moderate-severe CD and UC, monoclonal antibodies, including antitumor necrosis factor-alpha (anti-TNF-α) medications, are often first-line therapies.
Although some patients with IBD are underweight because of the severity of bowel inflammation or other factors, weight gain can be an unwanted side effect of medications (3). Previous studies are small and reveal contradictory results concerning the effects of anti-TNF-α therapy on patients' body mass index (BMI). Wiese et al. followed 7 patients with CD during a 6-month period and found that infliximab therapy improved nutritional status. However, the authors were unable to determine whether weight gain was due to an increase in fat or muscle mass (4). A study that included 40 patients examined body composition in patients with IBD under treatment with different anti-TNF-α agents and detected a significant improvement in the nutritional status of patients with IBD and a significant increase in the baseline BMI during the anti-TNF-α therapy (5). A post hoc analysis of IBD clinical trials identified factors associated with weight gain, including race, inflammatory markers, medication dose, and sex, but did not control for BMI (6).
The effects of anti-TNF-α therapy on weight have been studied mostly among patients with rheumatologic diseases. Studies have shown that weight gain is observed more frequently among patients with rheumatoid arthritis prescribed anti-TNF-α medication compared with patients treated with non-anti-TNF-α immunosuppressing therapies (7), and this weight gain occurs not only among patients with normal BMI at initiation of anti-TNF-α therapy but also among patients who were overweight when starting therapy (8). The report of weight gain among patients with rheumatologic medical conditions and no malnutrition or inflammation of the gastrointestinal tract supports the hypothesis that treatment with anti-TNF-α medications results in weight gain and that weight gain among patients with IBD on these medications is not solely due to improved nutritional status or decreased inflammation.
Despite anecdotal reporting and small studies, the occurrence of unintentional weight gain after anti-TNF-α therapy has not been clarified and the effects of anti-TNF-α medication on weight in a large cohort of patients with IBD have not yet been studied. Therefore, we performed an observational study in an unselected cohort of Danish patients with CD and UC to evaluate the effect of anti-TNF-α therapy on weight among patients with IBD.
This observational register study included data from 4 national Danish health registries. In Denmark (population approximately 5.8 million), all citizens have free access to a tax‐supported healthcare system, and this uniform organization allowed the development of a population‐based study design based on an unselected Danish study population (9,10).
We used data from The Danish National Registry for Biological Therapy in Inflammatory Bowel Disease (Bio-IBD), which includes a cohort of 2,843 patients with UC and CD who were treated with biologic therapy in Denmark (11). The study cohort from this database was linked to information on treatment with biologic therapy, surgeries, and diseases from (i) The Danish National Patient Registry (DNPR) (10), (ii) The Danish National Prescription Registry (NPR) which includes data on conventional drug treatment (12), (iii) biochemical information from the Laboratory database (LAB-F) from the Danish Health Data Authority (13), and (iv) The Civil Registration System which collects data on death and immigration (14,15).
The Bio-IBD database is a nationwide clinical database containing administrative and clinical information for all Danish patients treated with biological therapy for CD or UC (11). It is mandatory for all public and private hospitals and clinics to report to the database. Reporting is performed by the clinics concurrently when patients receive biological therapy. The Bio-IBD database also contains physician's clinical scores of disease activity, including the Harvey-Bradshaw Index (HBI) for patients with CD and the Simple Clinical Colitis Index (SCCAI) for patients with UC, other ongoing IBD medications, adverse events regarding biologic medications, and treatment decisions on continuing, changing, or stopping a biologic. Patient weights are also routinely reported. The Bio-IBD database is further described in detail elsewhere (11). From the LAB-F database, we obtained C-reactive protein (CRP) results. The LAB-F database contains biochemical information from Danish laboratories except from the Central Denmark Region (approximately 21% of the Danish population) (13). At the time of this study, fecal calprotectin was not routinely collected in Denmark and not available for every patient in either database. After induction therapy, we defined patients as having active disease if their SCCAI score was ≥5 or CRP ≥10 mg/L and if the HBI was ≥5 or CRP ≥10 mg/L for patients diagnosed with UC and CD, respectively (16).
The DNPR includes records of all discharges from Danish hospitals, and based on patients' civil registration numbers, the DNPR includes information on hospital, department, admission and discharge dates, procedures performed, and discharge diagnoses using the International Classification of Diseases (13). It is mandatory for all hospitals to report all relevant patient diseases for each discharge. In Denmark, anti‐TNF-α therapies are administered only in public hospitals or in hospital‐based outpatient settings in accordance with the National Board of Health, and information on biologic therapy is given by procedure codes for dates of administration in the patient registry. The system permits a complete assessment of individuals receiving a biologic therapy and its associated consequences and is a very complete registry. We used the DNPR as an initial database and supplemented with detailed clinical data from the Bio-IBD.
Data on nonbiologic drug treatment (aminosalicylic acid, corticosteroid, azathioprine, and methotrexate) for UC and CD are available from the NPR (12). Since January 1, 1995, data on reimbursed outpatient drug prescriptions have been available from the NPR, and all pharmacies in Denmark send key data on outpatient drug prescriptions directly to the prescription database.
The Bio-IBD database includes 2,843 patients with UC or CD who were treated for the first time with anti-TNF-α therapy during the period from March 1, 2016, to February 28, 2019 (Figure 1). We found 1,428 biologic naive patients older than 18 years who were treated with at least 3 infusions of infliximab. According to Danish national clinical guidelines, infliximab is first-line treatment, and only a few biologic naïve patients were treated with other biologics such as adalimumab or vedolizumab. Because of the small number of patients who received adalimumab or vedolizumab as first-line therapy, we only included patients prescribed infliximab, leaving 1,290 biologic naïve patients with UC or CD who initiated treatment with infliximab. Furthermore, only patients with a registered weight and height measured at the start of infliximab treatment and during a minimum of 2 additional infusion encounters during the study period were included for final analysis. The baseline BMI of each patient was calculated at the start of initial biologic load treatment (day 0), and the primary outcome was change in weight during induction therapy (0–90 days) and during maintenance of biologic therapy (91–270 days). A total of 851 patients with IBD were included in analyses (Figure 1). They were all started on infliximab treatment in the study period and were followed for 9 months, including 3 months after induction treatment (0–90 days) and then another 6 months (91–270 days) to examine change in weight throughout the maintenance period.
Statistical analysis and possible confounders
Data on possible confounders (age, disease duration, IBD-related surgery, disease activity [at week 14], and use of corticosteroid [during days 91–270]) were obtained from the Bio-IBD database, and comedication information was obtained from the NPR. Statistical analysis was performed in descriptive statistics to summarize the data; frequencies and medians with interquartile ranges were used for categorical and continuous variables, respectively. The weight of each patient was collected at the time of initiation of therapy, and each time, biologics were administered. A linear spline multilevel regression model was used to estimate the association between the use of biologics and weight gain according to initial BMI (17). As the knot point for the linear spline, we used the end of initial treatment at day 90 followed by maintenance treatment day 91 to the end of the study period at day 270. Patients were grouped by the BMI category corresponding to the World Health Organization classification (underweight BMI <18.5, normal weight BMI 18.5–24.9, overweight BMI 25–30, and obese BMI >30). In addition, we did a subgroup analysis including outlier patients who gained more than 10% of their initial body weight during the maintenance period of 91–270 days.
Approval/ethical considerations and data availability
This study followed all currently applicable Danish laws regarding scientific research. This study was not interventional and did not require direct patient contact or influence on patients' treatment. This study was approved by the Danish Patient Safety Authority (ref MITW, jnr. 3-3013-3126/1). The use of the Bio-IBD database was approved by the steering committee of the registry (RKKP, jnr. BIOIBD-2018-12-21), and this study was notified at the Danish Data Protection Agency under the current joint notification of the Region of Southern Denmark (jnr. 18/43305).
This register-based noninterventional study followed the STROBE guidelines (18). The authors of this article have no special access privileges to the data used in this study, and other researchers may apply for access to data through an application to the Research Service at the Danish Health Data Authority.
A total of 362 patients with UC and 489 patients with CD were included in the study cohort (Table 1). The median follow-up time after the first infusion of infliximab was 214 days (interquartile range 187–239), and the median number of weight measurements during the follow-up period was 6 (interquartile range 5–7).
Women with a normal BMI (18.5–24.9) had an average weight increase of 1.3 kg (95% confidence interval [CI] 0.3–2.4) during the induction phase. Among women who were underweight or overweight, there was no statistically significant weight gain in the induction phase, and we found no statistically significant weight gain among women with UC during the maintenance phase, regardless of BMI at the start of treatment (Table 2). Women in all BMI categories had an average weight gain of less than 2 kg in both the induction and maintenance phases. Women with CD with an initial BMI of <18.5 had a weight gain of 2.0 kg (95% CI 1.4–2.7) during the induction phase. During the maintenance phase from days 91–270, they gained 7.5 kg (95% CI −0.3 to 15.4). Women with CD with an initial BMI ≥18.5 showed no statistically significant weight gain throughout the study period (Figure 2).
Men with UC who had a BMI <18.5 at the initiation of therapy had weight gain during the induction phase from 0 to 90 days, but no weight gain in the follow-up maintenance period. Instead, they were found to have a weight loss of 2.3 kg (95% CI −3.8 to −0.7) during the maintenance phase (Table 2). Overall, underweight men with CD and UC gained 2.9 kg (95% CI 2.1–3.6) and 2.9 kg (95% CI 1.9–3.9), respectively, in the first 90 days. In the maintenance period, underweight men with CD gained 1.9 kg (95% CI −0.5 to 4.3). Men with CD and a BMI of 25–30 had an average weight gain of 1.7 kg (95% CI 0.5–3.0) at the end of 9 months. Men with CD who had a normal BMI or those who were obese at the initiation of therapy showed no statistically significant weight gain in either the induction or maintenance periods (Figure 3).
In subanalysis, we identified 28 patients who gained >10% of their initial body weight during the maintenance period of 91–270 days (Table 3). Five of these patients had UC and 23 had CD. Two patients with UC and 15 with CD were on concomitant steroids during maintenance treatment. Two of the patients with UC were underweight at the onset of anti‐TNF-α therapy initiation. Of the patients with CD who gained >10% of their body weight, 2 were initially underweight with BMI <18.5, 6 had an initial BMI of 18.5–24.9, 2 had an initial BMI of 25–30, and 1 had an initial BMI of >30. In addition, among patients who had an initial BMI of <18.5, 60% switched from the underweight BMI group to the normal weight (BMI 18.5–25) group.
In this cohort of 851 patients with IBD followed up for 9 months after initiation of infliximab, there were no large changes in weight among patients with initial BMI ≥18.5 from the initiation of treatment and through the follow-up period after controlling for both disease activity and steroid use. Women with CD and men with both UC and CD who started therapy with an initial BMI of <18.5 had weight gain of 2–2.9 kg within the first 90 days, and for women and men with CD, this was seen to continue slightly during maintenance treatment. This is likely due to improved nutritional status and not because of the medication, but without additional details, reason for weight gain cannot be determined nor can intent of weight gain be assessed. Although the increase in weight is not statistically significant, it may be clinically significant for some patients, especially if this were unwanted weight gain. There were only 18 patients in this cohort, and additional studies among patients with CD are recommended.
Among the IBD population, it has been previously hypothesized that weight gain on anti-TNF-α therapy may be multifactorial, with a component of weight gain from the medication itself, in addition to improved nutritional status as the medication heals the gastrointestinal tract and patients are able to both consume more calories and better absorb nutrients. Our findings, however, do not suggest that infliximab contributes to long-term weight gain. Other than underweight women with CD, patients on infliximab were not shown to have weight gain >2 kg during the maintenance phase. Aside from women with UC and a normal BMI who had an average weight gain of 1.3 kg (95% CI 0.3–2.4) in the induction phase, neither women nor men with CD or UC who had a normal or high BMI had significant weight gain within the first 90 days after initiation of infliximab, suggesting no medication effect even if patients were feeling better, increasing their calorie intake, and had improved absorption as the mucosa healed. Although some patients who were underweight did have weight gain in the first 90 days, our data do not suggest that anti-TNF-α therapy itself is the main cause of their weight gain. In addition, even if men and women with both UC and CD who were underweight at the onset of this study had weight gain <3 kg in the first 90 days, they had no significant weight gain from 91 to 270 days.
There are limited data regarding change in weight on anti-TNF-α therapy among patients with IBD. A post hoc analysis by Christian et al. of the ACCENT I, ACCENT II, ACT I, and SONIC trials identified factors associated with weight gain; on univariate analysis, factors associated with weight gain included African American race, elevated CRP, dose, low hematocrit and albumin, and adjusted analysis that identified sex as a factor associated with increased weight gain, with women having 4.3% weight gain compared with men with 3.9% (6). Of note, this study did not control for BMI or disease activity. Another study of 269 patients comparing weight gain among different biologic therapies showed an increase in weight over the first year, but no significant differences between weight gain among patients on anti-TNF-α therapy, vedolizumab or ustekinumab (19).
The results from this nationwide study contrast with smaller studies of patients with other autoimmune conditions, which have shown an increase in weight on anti-TNF-α therapy compared with non-anti-TNF-α therapy. Serelis et al. (20) investigated the effects of anti-TNF-α treatment in 19 women with rheumatoid arthritis, showing no significant effect on body composition, although conversely, during a 21-month follow-up, Engvall et al. (21) observed that body fat mass increased significantly in 40 patients with rheumatoid arthritis receiving infliximab therapy. Di Renzo et al. (22) examined 40 patients who had psoriasis and psoriatic arthritis for 2 years and demonstrated that anti-TNF-α treatment was related to an increase in weight.
To the best of our knowledge, this study is the largest cohort of patients on anti-TNF-α therapy where weight is routinely measured. Data are based on an unselected nationwide population from Danish health registries, which have high completeness and validity on the diagnosis and biologic treatment (10). The data available across multiple national databases enabled us to control for many factors including disease activity and steroid use and also to investigate the effect of weight gain based on both sex and type of inflammatory bowel disease, in addition to looking for varying effects of weight gain regarding initial BMI. Furthermore, this study follows a large cohort of patients for 270 days after initiation of infliximab therapy to assess both short-term and long-term effects of infliximab on weight. There were only a small number of patients who had weight gain >10% of body weight. Additional studies with a larger cohort of patients with significant weight gain are needed to identify additional factors that may be associated with the increase in weight.
This study also has limitations. We had strict criteria for inclusion and therefore excluded 577 patients (40.4%) because of lack of information on the measures of height or weight. There were 138 patients with no recorded height or weight at initiation of therapy, and thus, the reason most of the patients were excluded was because they did not meet the criteria of 3 measures of weight during the study period. This group of patients was equally distributed in the normal weight and overweight groups, none of the patients were underweight, and only 5 were obese. In clinical practice, it is likely more common to measure weight among patients who report a change in weight, and therefore, this study may have missed patients who had stable weight or who did not report a change in weight. As such, we consider missing data to be a minor problem, but we cannot rule out an impact of selection bias. In addition, although weights at the time of initiation of therapy were known, we do not know whether this weight was a normal baseline weight or whether patients had lost weight during their flare or before starting therapy. Another limitation is that this study only included patients on one type of biologic, and our results specifically refer to only infliximab. There were so few patients on other biologics, including patients prescribed other anti-TNF-α agents, vedolizumab and ustekinumab, that only patients on infliximab were included in this study. Although weight gain or lack thereof is believed to be a class effect, it is unknown whether patients prescribed adalimumab, certolizumab pegol, or golimumab had different responses to those medications. Second, disease activity was only based on clinical scores and CRP. Fecal calprotectin was not routinely collected in Denmark during the study period. Because we did not have access to clinic notes, colonoscopy, or pathology reports, we were unable to assess whether patients achieved endoscopic or histologic remission on infliximab. Although laboratory data and scores of IBD-specific clinical indices, such as HBI and SCCAI, were available, accurate assessment of disease activity was limited. Additional data, such as albumin and cholesterol, were not included. Cholesterol is not routinely measured and not available for all patients before and after initiation of infliximab. The database does not consistently include repeated measures of clinical disease activity indices. As a result, we did not use repeated measures of CRP or clinical disease activity index scores as a continuous outcome. Infliximab levels were not registered in the national databases or the Bio-IBD database and therefore response to medication based on the therapeutic level could not be assessed, and it is unknown whether dose optimization occurred.
Weight is often a sensitive topic for individuals, and people are often bothered by unintentional weight gain. Although not statistically significant, weight gain of a few kilograms may be clinically significant for individuals. Our results show that weight gain at 9 months after starting therapy in general was less than 2 kg for all men and women with IBD with the exception of women with CD with a starting BMI of <18.5. The etiology of weight gain among this cohort is likely multifactorial because they were initiated on infliximab when they were underweight and likely gained weight in the setting of decreased inflammation, improved caloric intake and nutrient absorption, and improved overall well-being.
To the best of our knowledge, this is the largest study evaluating the change in weight after initiation of infliximab therapy among patients with IBD. Among women and men who were not underweight, there was no observed significant weight gain at 9 months after initiation of therapy. Although previous studies report unintended weight gain as a concerning problem with anti-TNF-α medications, we cannot confirm these results. Although additional studies with larger cohorts of patients are recommended, including those with patients on other anti-TNF-α medications and other biologics, these data are reassuring that anti-TNF-α therapies do not cause significant weight gain with long-term administration.
CONFLICTS OF INTEREST
Guarantor of the article: Rachel W. Winter, MD, MPH.
Specific author contributions: R.W.W.: conception and design, data collection, data analyses, interpretation of results, and article writing. B.M.N.: funding, conception and design, data collection, assistance with data analyses, interpretation of results, and article writing. S.F.: conception and design, interpretation of results, and article writing. J.N.: data collection, data analyses, interpretation of results, and article writing. M.D.L.: funding, conception and design, data collection, data analyses, interpretation of results, and article writing. J.K.: conception and design, interpretation of results, and article writing. All authors have approved the final draft submitted.
Financial support: None to report.
Potential competing interests: R.W.W. is a consultant for AbbVie Pharmaceuticals and receives research funding from Takeda Pharmaceuticals. All other authors have no conflicts of interest.
WHAT IS KNOWN
- ✓ Patients with inflammatory bowel disease may lose weight during flares and gain weight with healing of inflammation.
- ✓ The association between antitumor necrosis factor-alpha (anti-TNF-α) and unintended weight gain has not previously been clarified.
WHAT IS NEW HERE
- ✓ Most of the patients initiated on anti-TNF therapy gained <2 kg during the 270-day study period.
- ✓ Weight gain among patients treated with anti-TNF-α therapies is unlikely to be due to a medication effect.
1. Mijac DD, Jankovic GL, Jorga J, et al. Nutritional status in patients with active inflammatory bowel disease: Prevalence of malnutrition and methods for routine nutritional assessment. Eur J Intern Med 2010;21:315–9.
2. Rocha R, Santana GO, Almeida N, et al. Analysis of fat and muscle mass in patients with inflammatory bowel disease during remission and active phase. Br J Nutr 2009;101:676–9.
3. Patsalos O, Dalton B, Leppanen J, et al. Impact of TNF-α inhibitors on body weight and BMI: A systematic review and meta-analysis. Front Pharmacol 2020;11:481.
4. Wiese D, Lashner B, Seidner D. Measurement of nutrition status in Crohn's disease patients receiving infliximab therapy. Nutr Clin Pract 2008;23:551–6.
5. Csontos AA, Molnar A, Piri Z, et al. The effect of anti-TNFalpha induction therapy on the nutritional status and dietary intake in inflammatory bowel disease. J Gastrointestin Liver Dis 2016;25:49–56.
6. Christian KE, Russman KM, Rajan DP, et al. Gender differences and other factors associated with weight gain following initiation of infliximab: A post hoc analysis of clinical trials. Inflamm Bowel Dis 2020;26:125–31.
7. Sfriso P, Caso F, Filardo GS, et al. Impact of 24 months of anti-TNF therapy versus methotrexate on body weight in patients with rheumatoid arthritis: A prospective observational study. Clin Rheumatol 2016;35:1615–8.
8. Florin V, Cottencin AC, Delaporte E, et al. Body weight increment in patients treated with infliximab for plaque psoriasis. J Eur Acad Dermatol Venereol 2013;27:e186–90.
9. Christiansen T. Organization and financing of the Danish Health Care System. Health Policy 2002;59:107–18.
10. Schmidt M, Schmidt SA, Sandegaard JL, et al. The Danish National Patient Registry: A review of content, data quality, and research potential. Clin Epidemiol 2015;7:449–90.
11. Larsen L, Jensen MD, Larsen MD, et al. The Danish National Registry for biological therapy in inflammatory bowel disease. Clin Epidemiol 2016;8:607–12.
12. Pottegard A, Schmidt SAJ, Wallach-Kildemoes H, et al. Data resource profile: The Danish National Prescription Registry. Int J Epidemiol 2017;46:798–f.
13. Arendt JFH, Hansen AT, Ladefoged SA, et al. Existing data sources in clinical epidemiology: Laboratory information system databases in Denmark. Clin Epidemiol 2020;12:469–75.
14. Pedersen CB. The Danish Civil Registration System. Scand J Public Health 2011;39:22–5.
15. Schmidt M, Pedersen L, Sorensen HT. The Danish Civil Registration System as a tool in epidemiology. Eur J Epidemiol 2014;29:541–9.
16. Higgins PD, Schwartz M, Mapili J, et al. Patient defined dichotomous end points for remission and clinical improvement in ulcerative colitis. Gut 2005;54:782–8.
17. Howe LD, Tilling K, Matijasevich A, et al. Linear spline multilevel models for summarising childhood growth trajectories: A guide to their application using examples from five birth cohorts. Stat Methods Med Res 2016;25:1854–74.
18. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. J Clin Epidemiol 2008;61:344–9.
19. Borren NZ, Tan W, Jess AT, et al. Assessment of body weight changes in patients with inflammatory bowel diseases initiating biologic therapy: A prospective cohort study. Dig Dis Sci 2020;65:3672–8.
20. Serelis J, Kontogianni MD, Katsiougiannis S, et al. Effect of anti-TNF treatment on body composition and serum adiponectin levels of women with rheumatoid arthritis. Clin Rheumatol 2008;27:795–7.
21. Engvall IL, Tengstrand B, Brismar K, et al. Infliximab therapy increases body fat mass in early rheumatoid arthritis independently of changes in disease activity and levels of leptin and adiponectin: A randomised study over 21 months. Arthritis Res Ther 2010;12:R197.
22. Renzo LD, Saraceno R, Schipani C, et al. Prospective assessment of body weight and body composition changes in patients with psoriasis receiving anti-TNF-alpha treatment. Dermatol Ther 2011;24:446–51.