Abstract
Introduction
IgA nephropathy is the most common primary kidney disease in the world and has a highly variable clinical presentation. While studies have indicated a link between glomerular disease and infections, large-scale studies on IgA nephropathy are missing.
Methods
In our study, IgA nephropathy was defined as having a kidney biopsy record 1997–2011 in Sweden. Each IgA nephropathy patient was matched with five reference individuals based on age, sex, calendar year, and county of residence. We excluded individuals with earlier organ transplants, HIV, immunodeficiency, or end-stage kidney disease. Linear and Cox regressions, adjusted for age, sex, education, and diabetes, were performed to analyze total infections and antimicrobial treatments in both patients and reference individuals. Sibling analyses were also performed.
Results
The linear regression analysis revealed a significant association between IgA nephropathy and the overall frequency of infections compared to the general population (β = 0.44; 95% CI: 0.35–0.53) and siblings (β = 0.36; 95% CI: 0.23–0.49). Similarly, antimicrobial prescriptions, especially antibiotics, were more common in IgA nephropathy compared to the general population and to siblings. Cox regression showed an elevated risk of any infection (adjusted hazard ratio [aHR] = 2.00; 95% CI: 1.84–2.18) and sepsis (aHR = 3.18; 95% CI: 2.17–4.65) corresponding to one extra case of sepsis per 63 patients followed for 10 years. The strongest associations were seen for urinary tract infections; ear, nose, and throat infections; and musculoskeletal and gastrointestinal infections.
Conclusion
Conclusively, our study demonstrates an increased prevalence of infections and antibiotic prescriptions in IgA nephropathy patients. The increased risk of sepsis warrants clinical awareness and prevention.
Keywords: Cohort study, IgA nephropathy, Infections, Sepsis
Introduction
IgA nephropathy (IgAN) is the most common primary kidney disease in the world [1]. It occurs in both sexes and all ages but is most often diagnosed in young and middle-aged adults. The clinical presentation is highly variable, from complete asymptomatic disease detectable only through urinary dipstick test, to terminal kidney failure with consequent use of dialysis treatment or kidney transplantation in 15–50% of the cases [2–4].
There are no biomarkers for IgAN in serum or urine; hence, diagnosing IgAN requires a kidney biopsy. The immunohistological examination is dominated by IgA deposits in the glomerular mesangium, and a central finding is that the deposited IgA is exclusively of the IgA1 subclass and its defect galactosylation in the antibody´s hinge region [4, 5].
Since IgA antibodies are mainly produced by the mucosal associated lymphoid tissue, the relationship between the mucosal immune system and IgAN has been considered since the disease was first identified by Berger and Hinglais [6]. The IgAN pathogenesis is still not fully understood, but research supports multifactorial genetic and environmental factors to be involved, and the role of mucosal infections as a potential contributor in the development of the aberrant IgA1 antibody production has been recognized [7, 8].
The nasopharynx-associated lymphoid tissue and the gut-associated lymphoid tissue are the first line of defense from foreign antigens entering our body, and in some IgAN patients, gastrointestinal and/or upper respiratory infections can aggravate the disease with episodic macrohematuria and/or increased proteinuria [2]. These two kinds of mucosal infections have also shown a higher frequency in patients with selective IgA deficiency [9–11]. Previous studies have indicated a link between glomerular disease and infections [12], but to our knowledge, there are no studies on the risk of infections for patients with IgAN.
Our overall aim was to investigate if IgAN diagnosis confers an increased risk of infections and antimicrobial prescriptions. The secondary aim was to classify the most common subtypes of infections in IgAN.
Methods
Exposure; Definition of IgAN
We defined IgAN as having a computerized biopsy record of IgAN (diagnosed in 1997–2011) at any of the four pathology departments that evaluated all kidney biopsy specimens in Sweden (Stockholm, Gothenburg, Linköping, and Malmö/Lund) at that time. Patients were reviewed manually or ascertained by the local IT department that searched biopsy records for the IgAN Systematized Nomenclature of Medicine Clinical Term (SnoMed) code D67300 [13]. The IgAN patients in our study were first identified by Welander et al. [14], and subsequently a random subset of the IgAN patients were manually validated against patient charts and biopsy reports [15]. The study found that of the 127 patients diagnosed with IgAN after undergoing the kidney biopsy, 121 also had a clinically confirmed diagnosis with typical symptoms such as hematuria, hypertension, or impaired kidney function: positive predictive value of 95% (95% confidence interval [CI]: 92–99%).
We selected patients from 1997 onward because that year marked the inclusion of day surgery data in the National Patient Register (NPR), and in 2001 outpatient care data were added [16]. Additionally, ICD-10 coding was implemented in 2007, improving diagnostic consistency and comparability between subtypes of infections. The last year of available biopsy data was 2011.
Matched Reference Individuals
The government agency Statistics Sweden identified up to five reference individuals for each patient with IgAN, matched for age, sex, calendar year, and county of residence at the time of kidney biopsy. The reference individuals were selected from the Total Population Register (TPR) [17]. The TPR contains data on the birth, death, family, marital status, and migration of all Swedish citizens. Further, we also identified all siblings of IgAN patients through the TPR [17] for sensitivity analyses.
Outcome: Infections and Prescribed Antibiotics
Through the unique personal identity number (PIN) assigned to all Swedish permanent residents [18], data can be linked to medical records and other registers. By linking the studied objects’ PINs, we received our outcomes from the established registers, The National Patient Register (NPR) and the National Prescribed Drug Register. The NPR was established in 1964 and is maintained by the National Board of Health and Welfare in Sweden. The register contains data and diagnoses from all specialist medical care (hospitalizations or non-primary care outpatient visits) without any Swedish resident being lost to follow-up [16]. The Prescribed Drug Register was established in 2005 and is also maintained by the National Board of Health and Welfare. It contains data on all prescribed drugs (including prescriptions from all primary care units) dispensed at pharmacies in Sweden. Drugs administered at hospital settings or nursing homes are not included in the register nor are vaccines and over-the-counter medicines. The register contains information about patients (e.g., sex, age, place of residence), products (e.g., ATC codes, strength, pack sizes), prescription details, and prescriber and cost [19].
We defined infections according to ICD-10 codes (complete list of definitions and codes in Supplement 1 of online suppl. material; for all online suppl. material, see https://doi.org/10.1159/000544753) as registered in the Swedish inpatient, outpatient, or surgical outpatient registries during years 1997–2015. For the prescribed drug registry, we identified the following ATC codes ever prescribed 2006–2015: J01 – antibacterial treatment, J02 – antimycotic treatment, J04 – antimycobacterial treatment, J05 – antiviral treatment, and P01 and P02 – antiprotozoal treatment (all codes listed in Supplement 2 in online suppl. material).
Other Variables
Data on diabetes mellitus diagnosis on both the IgAN patients and reference individuals were retrieved from the NPR. From the Swedish Longitudinal Integrated Database for Health Insurance and Labor Market Studies (LISA), we obtained information on educational levels [20] and categorized data according to the length of education (≤9 years’ compulsory school, 10–12 years, ≥13 years/university, and unknown). Education level is a proxy for socioeconomic status and is acknowledged as a risk factor for a number of diseases [21]. We used the highest level available during the follow-up, and for children, we used their parents’ educational levels.
Steroid Treatment
Since treatment with systematic corticosteroids is a well-known risk factor for infections [22–24], we produced data on prescribed systemic corticosteroids (ATC code H02AB) from the Prescribed Drug Register on both IgA patients and our reference individuals.
Exclusion Criteria
There were no exclusion criteria pertaining to the control sampling except that the reference individuals were not previously diagnosed with biopsy-verified IgAN at the time of matching/kidney biopsy of the index patient. Thus, our risk-set sampled reference individuals represent a random selection of the Swedish population including individuals with and without other chronic diseases. IgAN patients and reference individuals with a record of organ transplant, human immunodeficiency virus, or immunodeficiency before study start (definitions in appendix) were excluded from the study and individuals who developed any of the above conditions during the study were censored at first occurrence in the Swedish Patient Registry. Since end-stage renal disease (ESRD) leads to an impaired immune system with a high prevalence of infections, regardless of the underlying kidney disorder [25], patients who had ESRD at the time of the biopsy or developed ERSD during follow-up were also censored at the first occurrence of the diagnosis.
Follow-Up
Follow-up began on the day of the kidney biopsy/IgAN diagnosis and the corresponding date in the controls. Follow-up ended with death, emigration, first occurrence of above-mentioned exclusion criteria, or December 31, 2015, whichever occurred first.
Statistical Analyses
We performed linear regressions adjusted for age, sex, educational attainment, and diabetes with a total number of infections as the outcome. We also performed cox regression in strata for each patient and his/her 5 reference individuals and further adjusted for diabetes and educational attainment. Sibling analyses were not run in strata but were adjusted for baseline age, sex, educational attainment, and diabetes. We performed the same set of analyses for both outcomes: (1) infections registered in the patient registry and (2) antimicrobial treatments registered in the prescribed drug registry. A subgroup analysis on the risk of different subtypes of infections excluding those with prior corticosteroid prescription and further censoring at the first corticosteroid prescription during follow-up was also added. Further, we also performed linear regressions of previously registered infection types (as registered in the patient registry from January 1, 1997 until date of IgAN/corresponding date in matched controls), and these analyses were similarly adjusted for age, sex, education, and diabetes.
Results
Background Data
In total, our study cohort consisted of 2406 patients with IgAN and 11,609 controls. Of the IgAN patients, 739 (31%) were women. Two-thirds of the patients were between 18 and 59 years old at diagnosis, with approximately equally many individuals either aged <18 years or ≥60 years. Approximately 50% of participants were included during the last study period 2006–2011, and the median follow-up was 9.3 years. The majority (78%) had an educational level beyond compulsory school (≥10 years) (online suppl. material Table 1S).
Incidence and Absolute Excess Risk
The incidence rate of any infection in IgAN patients was 112 per 1,000 person-years compared to 52 for the reference individuals. This leads to a ratio of 2.2:1. This corresponds to an absolute excess risk of 60 more infections episodes in 100 IgAN patients followed for 10 years (Table 1), or one extra infection each year in 17 IgAN patients. During follow-up, there were 60 cases of sepsis in IgAN patients and 143 in the reference individuals. This corresponded to incidence rates of 2.9 versus 1.3 per 1 000 person-years, respectively. Hence, the absolute excess risk was 1.6 cases of sepsis per 100 patients followed for 10 years or equal to one extra case per 63 patients followed for 10 years (online suppl. material Table 2S).
Table 1.
Risk of infection
| Subgroups | IgAN | Controls | Any infection IgAN | Any infection controls | IR IgAN | IR controls | Coefficient (95% CI) |
|---|---|---|---|---|---|---|---|
| All | 2,406 | 11,609 | 2,330 | 5,960 | 112.3 (107.7–116.8) | 52.4 (51.1–53.8) | 0.44 (0.35–0.53) |
| Sex | |||||||
| Women | 739 (30.7%) | 3,561 (30.7%) | 704 | 2,132 | 107.1 (99.2–115.0) | 62.0 (59.3–64.6) | 0.34 (0.18–0.50) |
| Men | 1,667 (69.3%) | 8,048 (69.3%) | 1,626 | 3,828 | 118.3 (112.5–124.0) | 48.4 (46.9–49.9) | 0.48 (0.37–0.60) |
| Age | |||||||
| <18 years | 256 (10.6%) | 1,271 (10.9%) | 254 | 476 | 101.2 (88.7–113.6) | 36.4 (33.1–39.6) | 0.62 (0.46–0.78) |
| 18–39 years | 935 (38.9%) | 4,529 (39.0%) | 832 | 1,898 | 97.4 (90.7–104.0) | 41.7 (39.8–43.6) | 0.47 (0.34–0.60) |
| 40–59 years | 846 (35.2%) | 4,021 (34.6%) | 769 | 2,251 | 114.5 (106.5–122.4) | 56.1 (53.8–58.5) | 0.34 (0.18–0.50) |
| ≥60 years | 369 (15.3%) | 1,788 (15.4%) | 448 | 1,335 | 193.6 (175.7–211.6) | 90.5 (85.7–95.4) | 0.45 (0.14–0.76) |
| Calendar year | |||||||
| 1997–2000 | 524 (21.8%) | 2,537 (21.9%) | 631 | 1,867 | 92.7 (85.5–99.9) | 47.3 (45.1–49.4) | 0.46 (0.25–0.67) |
| 2001–2005 | 742 (30.8%) | 3,587 (30.9%) | 912 | 2,380 | 122.6 (114.7–130.6) | 56.5 (54.3–58.8) | 0.53 (0.32–0.74) |
| 2006–2011 | 1,140 (47.4%) | 5,485 (47.2%) | 787 | 1,713 | 129.5 (120.4–138.5) | 53.7 (51.2–56.3) | 0.37 (0.27–0.46) |
| Education | |||||||
| Compulsory school, ≤9 years | 474 (19.7%) | 2,249 (19.4%) | 552 | 1,518 | 155.6 (142.6–168.5) | 69.1 (65.6–72.6) | 0.46 (0.22–0.70) |
| Upper secondary school, 1–3 years | 1,080 (44.9%) | 5,258 (45.3%) | 1,070 | 2,769 | 115.9 (108.9–122.8) | 53.0 (51.0–55.0) | 0.45 (0.31–0.58) |
| University level | 796 (33.1%) | 3,810 (32.8%) | 627 | 1,583 | 86.6 (79.8–93.3) | 42.0 (39.9–44.1) | 0.37 (0.23–0.51) |
| Unknown | 56 (2.3%) | 292 (2.5%) | 81 | 90 | 272.2 (213.0–331.5) | 57.9 (45.9–69.8) | 1.20 (0.72–1.67) |
| Subgroups | IgAN | Siblings | Any infection IgAN | Any_infection siblings | IR IgAN | IR siblings | Coefficient (95% CI) |
|---|---|---|---|---|---|---|---|
| All | 1,904 | 4,049 | 1,729 | 2,329 | 104.6 (99.6–109.5) | 57.1 (54.7–59.4) | 0.36 (0.23–0.49) |
| Sex | |||||||
| Women | 566 (29.7%) | 2,009 (49.6%) | 493 | 1,311 | 95.4 (86.9–103.8) | 64.6 (61.1–68.1) | 0.23 (0.05–0.41) |
| Men | 1,338 (70.3%) | 2,040 (50.4%) | 1,236 | 1,018 | 108.8 (102.7–114.8) | 49.6 (46.6–52.7) | 0.42 (0.24–0.60) |
| Age | |||||||
| <18 years | 240 (12.6%) | 662 (16.3%) | 238 | 516 | 101.4 (88.5–114.3) | 72.4 (66.2–78.7) | 0.38 (−0.01–0.77) |
| 18–39 | 824 (43.3%) | 1,491 (36.8%) | 679 | 716 | 88.3 (81.6–94.9) | 47.4 (43.9–50.9) | 0.37 (0.17–0.56) |
| 40–59 | 650 (34.1%) | 1,451 (35.8%) | 585 | 849 | 108.9 (100.0–117.7) | 56.6 (52.7–60.4) | 0.31 (0.10–0.53) |
| ≥60 years | 190 (10.0%) | 445 (11.0%) | 227 | 248 | 201.9 (175.6–228.1) | 69.2 (60.6–77.8) | 0.62 (0.19–1.05) |
| Calendar year | |||||||
| 1997–2000 | 405 (21.3%) | 872 (21.5%) | 419 | 646 | 76.2 (68.9–83.4) | 45.5 (42.0–49.1) | 0.36 (0.11–0.60) |
| 2001–2005 | 592 (31.1%) | 1,276 (31.5%) | 740 | 922 | 120.4 (111.8–129.1) | 60.0 (56.1–63.8) | 0.55 (0.26–0.84) |
| 2006–2011 | 907 (47.6%) | 1,901 (46.9%) | 570 | 761 | 116.6 (107.0–126.2) | 67.6 (62.8–72.4) | 0.25 (0.10–0.40) |
| Education | |||||||
| Compulsory school, ≤9 years | 306 (16.1%) | 732 (18.1%) | 336 | 394 | 139.8 (124.8–154.7) | 54.6 (49.2–60.0) | 0.59 (0.37–0.81) |
| Upper secondary school, 1–3 years | 894 (47.0%) | 1,895 (46.8%) | 845 | 1,116 | 107.8 (100.6–115.1) | 56.8 (53.5–60.1) | 0.36 (0.17–0.54) |
| University level | 664 (34.9%) | 1,232 (30.4%) | 485 | 610 | 79.8 (72.7–86.9) | 48.6 (44.7–52.4) | 0.25 (0.04–0.46) |
| Unknown | 40 (2.1%) | 190 (4.7%) | 63 | 209 | 294.1 (221.4–366.7) | 148.2 (128.1–168.2) | 0.48 (−1.16–2.08) |
IgAN, IgA nephropathy; IR, incidence rate; CI, confidence interval.
Risk of Future Infections: Linear Regression
The linear regression analysis showed a significant association between IgAN and the overall frequency of infections compared to the reference individuals β = 0.44 (95% CI: 0.35–0.53). This interprets as each IgAN patient having on average 0.44 more episodes of infections during the follow-up. For women, the corresponding numbers were β = 0.34 (95% CI: 0.18–0.50) and for men β = 0.48 (95% CI: 0.37–0.60). The highest coefficient was seen in the youngest subgroup (<18 years, β = 0.62; 95% CI: 0.46–0.78) and lowest in the 40–59 years’ subgroup (β = 0.34; 95% CI: 0.18–0.50) (Table 1). A comparison between IgAN patients and their siblings also showed a statistically significant association with future infections, but with a lower coefficient (β = 0.36; 95% CI: 0.23–0.49), for women (β = 0.23; 95% CI: 0.05–0.41) and men (β = 0.42; 95% CI: 0.24–0.60) (Table 1; Fig. 1).
Fig. 1.
Main results: β-coefficients and 95% CI (linear regression). IgAN, IgA nephropathy; CI, confidence interval.
Subtypes of Infections
The infectious subtypes with the strongest association with IgAN were infections in the urinary tract (UTI) (β = 0.09; 95% CI: 0.07–0.11); musculoskeletal and connective tissue (β = 0.08; 95% CI: 0.05–0.12); ear, nose; and throat (ENT) (β = 0.07; 95% CI: 0.02–0.11); and gastrointestinal tract (β = 0.06; 95% CI: 0.03–0.09). Infections in the lower respiratory tract, skin, and mycosis were also statistically significant; however, no associations were seen for infections in the central nervous system or infections caused by helminths and protozoans (online suppl. material Table 2S). In the sibling analyses, the same subtypes with strongest associations with IgAN were again UTI (β = 0.10; 95% CI: 0.06–0.13), ENT (β = 0.07; 95% CI: 0.01–0.12), muscular (β = 0.07; 95% CI: 0.01–0.14), and gastrointestinal tract infections (β = 0.04; 95% CI: 0.00–0.08) (online suppl. material Table 2S; Fig. 2).
Fig. 2.
Subtypes of infections: β-coefficients and 95% CI (linear regression). ENT, ear, nose, and throat; GI tract, gastrointestinal tract; CI, confidence interval.
Increased Antimicrobial Prescription
A significant association was found for the frequency of prescribed antimicrobial agents compared to the reference individuals (β = 0.65; 95% CI: 0.49–0.80), women (β = 0.89; 95% CI: 0.52–1.25), and men (β = 0.54; 95% CI: 0.38–0.70). Similar estimates were seen in the sibling analyses (β = 0.47; 95% CI: 0.25–0.69) (Table 2). The highest coefficient was seen for antibiotic prescriptions (β = 0.62; 95% CI: 0.47–0.76) (online suppl. material Table 3S), and corresponding estimate when compared to siblings was β = 0.40; 95% CI: 0.19–0.61 (online suppl. material Table 3S). Cox regressions showed an adjusted hazard ratio (aHR) of 1.49 (95% CI: 1.40–1.58) for overall antimicrobials prescription and 1.48 (95% 1.39–1.58) for antibiotics compared to the reference individuals, and 1.27 (95% CI: 1.17–1.38) for overall prescription and an aHR of 1.26 (95% CI: 1.16–1.37) for antibiotics compared to siblings (online suppl. material Table 3S).
Table 2.
Prescriptions
| Subgroups | IgAN | Controls | Follow-up years IgAN | Follow-up years controls | Prescriptions IgAN | Prescriptions controls | IR IgAN | IR controls | Coefficient (95% CI) |
|---|---|---|---|---|---|---|---|---|---|
| All | 2,406 | 11,609 | 20,735 | 113,479 | 7,132 | 26,472 | 343.7 (335.7–351.6) | 232.9 (230.0–235.7) | 0.64 (0.49–0.80) |
| Sex | |||||||||
| Women | 739 (30.7%) | 3,561 (30.7%) | 6,574 | 34,412 | 3,019 | 11,290 | 459.2 (442.8–474.6) | 328.1 (322.0–334.1) | 0.89 (0.52–1.25) |
| Men | 1,667 (69.3%) | 8,048 (69.3%) | 13,748 | 79,066 | 4,113 | 15,182 | 299.2 (290.0–308.3) | 192.0 (189.0–195.1) | 0.54 (0.38–0.70) |
| Age | |||||||||
| <18 years | 256 (10.6%) | 1,271 (10.9%) | 2,510 | 13,092 | 786 | 2,545 | 313.1 (291.2–335.0) | 194.4 (186.8–201.9) | 1.08 (0.71–1.45) |
| 18–39 years | 935 (38.9%) | 4,529 (39.0%) | 8,545 | 45,535 | 2,674 | 9,277 | 312.9 (301.1–324.8) | 203.7 (199.6–207.9) | 0.80 (0.57–1.02) |
| 40–59 years | 846 (35.2%) | 4,021 (34.6%) | 6,953 | 40,100 | 2,407 | 9,622 | 346.1 (332.1–360.0) | 239.9 (235.2–244.7) | 0.37 (0.09–0.66) |
| ≥60 years | 369 (15.3%) | 1,788 (15.4%) | 2,313 | 14,749 | 1,265 | 5,028 | 546.8 (516.7–576.9) | 340.9 (331.5–350.3) | 0.62 (0.12–1.11) |
| Calendar year | |||||||||
| 1997–2000 | 524 (21.8%) | 2,537 (21.9%) | 6,806 | 39,489 | 1,715 | 7,175 | 252.0 (240.1–263.9) | 181.7 (177.5–185.9) | 0.44 (0.05–0.82) |
| 2001–2005 | 742 (30.8%) | 3,587 (30.9%) | 7,437 | 42,103 | 2,752 | 10,576 | 370.0 (356.2–383.8) | 251.2 (246.4–256.0) | 0.68 (0.36–0.99) |
| 2006–2011 | 1,140 (47.4%) | 5,485 (47.2%) | 6,079 | 31,885 | 2,665 | 8,721 | 438.4 (421.7–455.0) | 273.5 (267.8–279.2) | 0.71 (0.53–0.88) |
| Education | |||||||||
| Compulsory school, ≤9 years | 474 (19.7%) | 2,249 (19.4%) | 3,548 | 21,965 | 1,492 | 5,378 | 420.5 (399.1–441.8) | 244.8 (238.3–251.4) | 0.67 (0.30–1.04) |
| Upper secondary school, 1–3 years | 1,080 (44.9%) | 5,258 (45.3%) | 9,232 | 52,265 | 3,107 | 12,107 | 336.5 (324.7–348.4) | 231.6 (227.5–235.8) | 0.53 (0.29–0.76) |
| University level | 796 (33.1%) | 3,810 (32.8%) | 7,244 | 37,693 | 2,404 | 8,628 | 331.9 (318.6–345.1) | 228.9 (224.1–233.7) | 0.78 (0.50–1.05) |
| Unknown | 56 (2.3%) | 292 (2.5%) | 297 | 1,554 | 129 | 359 | 433.6 (358.7–508.4) | 230.9 (207.0–254.8) | 1.09 (0.31–1.86) |
| Subgroups | IgAN | Siblings | Follow-up years IgAN | Follow-up years siblings | Prescriptions_IgAN | Prescriptions_siblings | IR IgAN | IR siblings | Coefficient (95% CI) |
|---|---|---|---|---|---|---|---|---|---|
| All | 1,904 | 4,049 | 16,904 | 40,896 | 5,631 | 11,079 | 333.1 (324.4–341.8) | 279.9 (265.9–275.9) | 0.47 (0.25–0.69) |
| Sex | |||||||||
| Women | 566 (29.7%) | 2,009 (49.6%) | 5,169 | 20,305 | 2,374 | 6,756 | 459.2 (440.7–477.7) | 332.7 (324.8–340.7) | 0.81 (0.37–1.24) |
| Men | 1,338 (70.3%) | 2,040 (50.4%) | 11,365 | 20,516 | 3,257 | 4,323 | 286.6 (276.7–296.4) | 210.7 (204.4–217.0) | 0.27 (0.04–0.50) |
| Age | |||||||||
| <18 years | 240 (12.6%) | 662 (16.3%) | 2,347 | 7,122 | 749 | 1,862 | 319.1 (296.3–342.0) | 261.4 (249.6–273.3) | 0.65 (0.08–1.22) |
| 18–39 | 824 (43.3%) | 1,491 (36.8%) | 7,691 | 15,101 | 2,430 | 3,618 | 315.9 (303.4–328.5) | 239.6 (231.8–247.4) | 0.78 (0.47–1.10) |
| 40–59 | 650 (34.1%) | 1,451 (35.8%) | 5,371 | 15,012 | 1,783 | 4,338 | 331.9 (316.5–347.3) | 289.0 (280.4–297.6) | −0.04 (−0.44–0.35) |
| ≥60 years | 190 (10.0%) | 445 (11.0%) | 1,124 | 3,585 | 669 | 1,261 | 595.0 (549.9–640.1) | 351.7 (332.2–371.1) | 0.93 (0.12–1.74) |
| Calendar year | |||||||||
| 1997–2000 | 405 (21.3%) | 872 (21.5%) | 5,502 | 14,182 | 1,325 | 2,972 | 240.8 (227.9–253.8) | 209.6 (202.0–217.1) | 0.24 (−0.28–0.76) |
| 2001–2005 | 592 (31.1%) | 1,276 (31.5%) | 6,144 | 15,379 | 2,211 | 4,544 | 359.9 (344.9–374.9) | 295.5 (286.9–304.1) | 0.51 (0.07–0.95) |
| 2006–2011 | 907 (47.6%) | 1,901 (46.9%) | 4,888 | 11,260 | 2,095 | 3,563 | 428.5 (410.2–446.9) | 316.4 (306.0–326.8) | 0.61 (0.35–0.87) |
| Education | |||||||||
| Compulsory school, ≤9 years | 306 (16.1%) | 732 (18.1%) | 2,404 | 7,214 | 986 | 1,913 | 410.1 (384.5–435.7) | 265.2 (253.3–277.0) | 0.83 (0.29–1.37) |
| Upper secondary school, 1–3 years | 894 (47.0%) | 1,895 (46.8%) | 7,836 | 19,643 | 2,592 | 5,287 | 330.8 (318.0–343.5) | 269.1 (261.9–276.4) | 0.35 (0.03–0.66) |
| University level | 664 (34.9%) | 1,232 (30.4%) | 6,080 | 12,553 | 1,955 | 3,440 | 321.5 (307.3–335.8) | 274.0 (264.9–283.2) | 0.47 (0.08–0.85) |
| Unknown | 40 (2.1%) | 190 (4.7%) | 214 | 1,410 | 98 | 439 | 457.4 (366.9–548.0) | 311.2 (282.1–340.3) | 0.48 (−1.09–2.05) |
IgAN, IgA nephropathy; IR, incidence rate; CI, confidence interval.
Risk of Future Infections: Cox Regression
Cox regressions showed an elevated risk of any infection with an aHR of 2.02 (1.85–2.20). The highest aHR was for sepsis aHR 3.18 (2.17–4.65), but all subtypes except infections in the central nervous system, helminth, and protozoal infections were significantly increased (online suppl. material Tables 4–5S). Compared to siblings, the aHR was 1.88 (1.67–2.12) for any infection and 2.36 (1.09–5.12) for sepsis. Compared to siblings, the highest aHRs were found for UTI (aHR 2.67; 95% CI: 2.06–3.45) and for mycosis (aHR 2.73; 95% CI: 1.83–4.08) (online suppl. material Tables 4–5S).
Risk of Infections before IgAN
The linear regressions of previously registered infection also showed a significant association (β = 0.53; 95% CI: 0.47–0.59), supporting an increased risk also before the time of diagnosis. The coefficient was especially increased in IgAN patients diagnosed in the latest calendar period (2006–2011), hence with a longer follow-up (9–14 years) in the analysis, suggesting a continuously increased risk. The most common subtype of infection was UTI (β = 0.25; 95% CI: 0.23–0.27) (online suppl. material Table 6S).
Prescription of Corticosteroids
During follow-up, 793 (33%) of 2,406 IgAN patients were prescribed systemic corticosteroids and corresponding numbers for the reference individuals were 1 932 (17%) of 11,609. A subgroup analysis was performed excluding those with a prescription of corticosteroids prior to study entry and censoring patients at the first registered corticosteroid prescription, showing similar estimates with an aHR 1.93 (95% CI: 1.74–2.14) for any infection in the nonsteroid group compared to 2.02 (1.85–2.20) in the main analysis. Changes in aHRs were minor for all subtypes of infections (online suppl. material Table 5S).
Only 2 (3%) of the 60 patients with a record of sepsis collected a prescription of corticosteroids within 180 days before the sepsis diagnosis, with corresponding numbers for the reference individuals being 3 (2%) out of 143. In the subgroup analysis excluding individuals at the first corticosteroid prescription, the aHR for risk of sepsis was 2.90 (95% CI: 1.78–4.70) as compared to 3.18 (95% CI: 2.17–4.65) in the main analysis.
Discussion
Principal Findings
Our study demonstrates an increased prevalence of infections and antibiotic prescriptions in IgAN patients compared to the general population and their siblings, both before and after IgAN diagnosis. Since our outcome was only based on hospital-based infections, our risk estimates may primarily represent the risk of severe infection, but the concomitantly increased level of antimicrobial prescriptions likely reflects a truly increased frequency of infections as the prescription registry also contains all primary care prescriptions. Another important finding is the increased risk of sepsis, which has, to our knowledge, not been reported in IgAN previously, and warrants increased clinical awareness.
Comparison to Other Studies
We are not aware of any other studies addressing risk of infections in IgAN patients with their acknowledged aberrant IgA1 antibody production; however, there are studies on IgA deficiency. Although generally better tolerated than other antibody deficiencies, patients with IgA deficiency are at an increased risk of infections requiring hospitalization [27]. As aforementioned, infections are most commonly observed in the respiratory and gastrointestinal tracts [9–11], but also ENT infections, UTI, and sepsis [11], which are the same subtypes of infection as indicated in our study. Of note, in the linear regression analysis on the risk of infections before IgAN diagnosis, UTI was the primary subtype, and this increase may be due to misclassification from misinterpreted urinary dipsticks. Nevertheless, these types of infections occurring in mucosal surfaces could implicate a potential mechanism involving aberrant mucosal antibodies.
Not only chronic kidney disease with impaired kidney function but also proteinuria with subsequent loss of gamma globulin and complement factors is an acknowledged risk factor for infections [28]. Previous studies have shown that proteinuria is a risk factor for infections independent of kidney function level [29, 30]. Results from a recent Taiwanese study support proteinuria as a separate risk factor, although foremost in older patients with a weakened association in younger CKD patients [31]. Additionally, a study from the USA on the risk of infections in both children and adult patients with glomerular diseases also showed an independent risk increase associated with severe proteinuria [12]. In our study, we do not have data on proteinuria levels, but this must be considered a potential contributing factor to our results.
Lastly, treatment with corticosteroids is a strong and well-known risk factor for adverse effects, primarily infections, in patients with IgAN [22–24, 32]. Unsurprisingly, corticosteroids were another independent risk factor in the above mentioned study by Glenn et al. [12] on infections in patients with glomerular kidney disease. In our IgAN cohort, 33% were prescribed corticosteroids during follow-up, but only 3% had collected such prescriptions within 180 days before a record of sepsis diagnosis, compared to 2% of the reference individuals diagnosed with sepsis, and the aHR for the risk of sepsis of 2.90 (95% CI: 1.78–4.70) in patients with no corticosteroid prescription compared to 3.18 (95% CI: 2.17–4.65) in those who had been prescribed corticosteroids also supports other, speculatively, internal risk factors in the IgAN patients. Hence, we do not believe that an increased use of corticosteroids fully explains the increased risk of sepsis in IgAN patients.
Strengths and Limitations
Our results are based on nationwide and high-quality health care registries, which grants high external validity to countries with similar health care systems. Our results are particularly robust as we performed both linear and cox regressions with both matched population controls and sibling controls. As infections may recur, the total infectious load may be better displayed through the number of infections ever during follow-up than cox regression, particularly as the start of follow-up constitutes a random timepoint for the reference individuals.
Further, most infection types showed similar linear coefficients both before and after IgAN diagnosis with one exception: UTI being far more commonly diagnosed before IgAN diagnosis, which likely reflects misclassification of prediagnosis proteinuria and/or hematuria. We believe that this misclassification often leads to delayed referral to a nephrology department, particularly when kidney function is preserved. The increased risk prior to diagnosis, along with suspected misclassifications, suggests that both abnormal antibodies and undiagnosed/untreated proteinuria could play a role in the underlying explanation.
A major limitation is of course that we did not have data on proteinuria levels as this is shown to be an independent risk factor for infections [12, 28–31]. These data are not available from the NPR. We reported and adjusted only for the use of prescribed corticosteroids, as other immunosuppressive therapies are generally administered in combination with or following corticosteroid treatment. Therefore, the number of patients receiving additional, more potent cytostatic agents is likely smaller than the already low number of corticosteroid-treated patients observed before the onset of serious infections. Nor have we adjusted for other medication such as antihypertensives. However, previous studies showed no elevated risk of influenza or pneumonia when treated with RAS inhibitors [33], and no difference in the risk of UTI when treated with RAS inhibitors, calcium channel blockers, beta blockers, or diuretics [34].
Additionally, adjustments were only made for diabetes as a comorbidity, as we believe it to be the most significant contributor among potential comorbidities. Another limitation is the inability to stratify by CKD stages, except for the exclusion of patients with ESRD, again due to insufficient data. Finally, our study does not capture infections managed by primary care providers, or infections that did not need medical attention at all. However, data on antimicrobial prescriptions serve as a proxy for infections diagnosed outside of hospital settings, as they include all primary care treatment prescriptions.
Implication of the Findings
Infections and antimicrobial prescriptions, especially antibiotics, were more common in IgAN patients than in matched reference individuals and siblings. We found an absolute excess risk corresponding to one extra case of sepsis per 63 IgAN patients followed for 10 years. The increased risk of sepsis mandates clinical awareness to facilitate prevention with early antimicrobial treatment. The role of proteinuria in the risk development should be considered and treated, and vaccination status for all patients should be discussed.
Conclusion
Our study demonstrates an increased prevalence of infections and antibiotic prescriptions in IgAN patients compared to the general population and their siblings. The increased risk of sepsis, corresponding to 1 extra case in 63 patients followed for 10 years, warrants clinical knowledge and active prevention. Further studies on the potential impact of IgAN and its immunological aberrations on overall infection susceptibility are warranted.
Statement of Ethics
The study was approved by the Stockholm Ethics Review Board (January 22, 2014, Approval No. 2013/2095-31/2). Because this was a strictly register-based study, the requirement of informed consent was waived by the Board [26].
Conflict of Interest Statement
The authors have no conflicts of interest to declare concerning this project. Dr. Rehnberg has obtained reimbursements for lectures from AstraZeneca and Bayer. Dr. Ludvigsson coordinates a study on behalf of the Swedish IBD quality register (SWIBREG) and that study has received funding from Janssen corporation. Dr. Ludvigsson has also received financial support from MSD for developing a paper reviewing national health care registers in China. Dr. Ludvigsson has an ongoing research collaboration with Takeda.
Funding Sources
Dr. Rehnberg has funding from County Council of Värmland, Sweden and County Council of Örebro Län, Sweden. The sponsors had no role in the study design, data collection, data analysis, and interpretation of data; writing of the report; and the decision to submit the paper for publication. The coauthors have not received any funding for this project.
Author Contributions
J.R., J.F.L., and L.E. designed the study. J.R. and L.E. analyzed the data. J.R. made the tables and figures and drafted the paper, with inputs from L.E. and J.F.L. L.E. supervised the study. All the authors approved the final version of the manuscript. L.E. had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Funding Statement
Dr. Rehnberg has funding from County Council of Värmland, Sweden and County Council of Örebro Län, Sweden. The sponsors had no role in the study design, data collection, data analysis, and interpretation of data; writing of the report; and the decision to submit the paper for publication. The coauthors have not received any funding for this project.
Data Availability Statement
The data supporting this article cannot be made accessible in any publicly available repository according to the decision of the Stockholm Ethics Review Board (January 22, 2014, Approval No. 2013/2095-31/2). The kidney biopsy datasets analyzed for the current study were obtained from 4 different Swedish hospital records, regulated by patient secrecy, and hence cannot be made publicly available. The biopsy records were combined with national nationwide Swedish registries. Data from Swedish national registries are available to researchers worldwide after having obtained appropriate ethical approval. Further inquiries can be directed to the corresponding author.
Supplementary Material.
References
- 1. D’Amico G. The commonest glomerulonephritis in the world: IgA nephropathy. Q J Med. 1987;64(245):709–27. [PubMed] [Google Scholar]
- 2. Donadio JV, Grande JP. IgA nephropathy. N Engl J Med. 2002;347(10):738–48. [DOI] [PubMed] [Google Scholar]
- 3. D’Amico G. Natural history of idiopathic IgA nephropathy and factors predictive of disease outcome. Semin Nephrol. 2004;24(3):179–96. [DOI] [PubMed] [Google Scholar]
- 4. Wyatt RJ, Julian BA. MEdical progress: IgA nephropathy. N Engl J Med Overseas Ed. 2013;368(25):2402–14. [DOI] [PubMed] [Google Scholar]
- 5. Novak J, Moldoveanu Z, Julian BA, Raska M, Wyatt RJ, Suzuki Y, et al. Aberrant glycosylation of iga1 and anti-glycan antibodies in iga nephropathy: role of mucosal immune system. Adv Otorhinolaryngology. 2011. [DOI] [PubMed] [Google Scholar]
- 6. Berger J, Hinglais N. Intercapillary deposits of IgA-IgG. J Urol Nephrol. 1968;74(9):694–5. [PubMed] [Google Scholar]
- 7. Rollino C, Vischini G, Coppo R. IgA nephropathy and infections. J Nephrol. 2016;29(4):463–8. [DOI] [PubMed] [Google Scholar]
- 8. He JW, Zhou XJ, Lv JC, Zhang H. Perspectives on how mucosal immune responses, infections and gut microbiome shape iga nephropathy and future therapies. Theranostics. 2020;10(25):11462–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Yel L. Selective IgA deficiency. J Clin Immunol. 2010;30(1):10–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Rawla P, Killeen RB, Joseph NI. Selective IgA deficiency: Treasure Island (FL); 2024. [Google Scholar]
- 11. Ludvigsson JF, Neovius M, Hammarström L. Risk of infections among 2100 individuals with IgA deficiency: a nationwide cohort study. J Clin Immunol. 2016;36(2):134–40. [DOI] [PubMed] [Google Scholar]
- 12. Glenn DA, Henderson CD, O’Shaughnessy M, Hu Y, Bomback A, Gibson K, et al. Infection-related acute care events among patients with glomerular disease. Clin J Am Soc Nephrol. 2020;15(12):1749–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. SNOMED International SNOMED CT . Swedish Edition v1.37.3. 2018. Available from: http://browser.ihtsdotools.org/index-ie.html?perspective=full&conceptId1=404684003&edition=se-edition&release=v20180531&server=https://prod-browser-exten.ihtsdotools.org/api/s
- 14. Welander A, Sundelin B, Fored M, Ludvigsson JF. Increased risk of IgA nephropathy among individuals with celiac disease. J Clin Gastroenterol. 2013;47(8):678–83. [DOI] [PubMed] [Google Scholar]
- 15. Jarrick S, Lundberg S, Welander A, Fored CM, Ludvigsson JF. Clinical validation of immunoglobulin A nephropathy diagnosis in Swedish biopsy registers. Clin Epidemiol. 2017;9:67–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim JL, Reuterwall C, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011;11:450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Ludvigsson JF, Almqvist C, Bonamy AKE, Ljung R, Michaëlsson K, Neovius M, et al. Registers of the Swedish total population and their use in medical research. Eur J Epidemiol. 2016;31(2):125–36. [DOI] [PubMed] [Google Scholar]
- 18. Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, Ekbom A. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol. 2009;24(11):659–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Available from: https://www.socialstyrelsen.se/en/statistics-and-data/registers/national-prescribed-drug-register/
- 20. Ludvigsson JF, Svedberg P, Olén O, Bruze G, Neovius M. The longitudinal integrated database for health insurance and labour market studies (LISA) and its use in medical research. Eur J Epidemiol. 2019;34(4):423–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Olén O, Bihagen E, Rasmussen F, Ludvigsson JF. Socioeconomic position and education in patients with coeliac disease. Dig Liver Dis. 2012;44(6):471–6. [DOI] [PubMed] [Google Scholar]
- 22. Wong MG, Lv J, Hladunewich MA, Jha V, Hooi LS, Monaghan H, et al. The therapeutic evaluation of steroids in IgA nephropathy global (TESTING) study: trial design and baseline characteristics. Am J Nephrol. 2021;52(10–11):827–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Rauen T, Eitner F, Fitzner C, Sommerer C, Zeier M, Otte B, et al. Intensive supportive care plus immunosuppression in IgA nephropathy. N Engl J Med. 2015;373(23):2225–36. [DOI] [PubMed] [Google Scholar]
- 24. Rivedal M, Haaskjold YL, Eikrem Ø, Bjørneklett R, Marti HP, Knoop T. Use of corticosteroids in Norwegian patients with immunoglobulin a nephropathy progressing to end-stage kidney disease: a retrospective cohort study. BMC Nephrol. 2024;25(1):42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Syed-ahmed M, Narayanan M. Immune dysfunction and risk of infection in chronic kidney disease. Adv Chronic Kidney Dis. 2019;26(1):8–15. [DOI] [PubMed] [Google Scholar]
- 26. Ludvigsson JF, Håberg SE, Knudsen GP, Lafolie P, Zoega H, Sarkkola C, et al. Ethical aspects of registry-based research in the Nordic countries. Clin Epidemiol. 2015;7:491–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Furst DE. Serum immunoglobulins and risk of infection: how low can you go? Semin Arthritis Rheum. 2009;39(1):18–29. [DOI] [PubMed] [Google Scholar]
- 28. Claudio P, Gabriella M. Nephrotic syndrome: pathophysiology and consequences. J Nephrol. 2023;36(8):2179–90. [DOI] [PubMed] [Google Scholar]
- 29. Ishigami J, Grams ME, Chang AR, Carrero JJ, Coresh J, Matsushita K. CKD and risk for hospitalization with infection: the atherosclerosis risk in communities (ARIC) study. Am J Kidney Dis. 2017;69(6):752–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. McDonald HI, Thomas SL, Millett ERC, Nitsch D. CKD and the risk of acute, community-acquired infections among older people with diabetes mellitus: a retrospective cohort study using electronic health records. Am J Kidney Dis. 2015;66(1):60–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Yang WS, Chang YC, Hsieh ML, Wang JL, Wu LC, Chang CH. Stratified risks of infection-related hospitalization in patients with chronic kidney disease: a prospective cohort study. Sci Rep. 2020;10(1):4475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Zhang Y, Hu YT, Lv JC, Zhang H. Corticosteroids in the treatment of IgA nephropathy: lessons from the TESTING trial. Pediatr Nephrol. 2023;38(10):3211–20. [DOI] [PubMed] [Google Scholar]
- 33. Christiansen CF, Heide-Jørgensen U, Rasmussen TB, Bodilsen J, Søgaard OS, Maeng M, et al. Renin-angiotensin system blockers and adverse outcomes of influenza and pneumonia: a Danish cohort study. J Am Heart Assoc. 2020;9(19):e017297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Gremke N, Kostev K, Kalder M. Association between antihypertensive medication and the risk of urinary tract infection (UTI) of outpatients: a retrospective cohort study. Infection. 2023;51(2):417–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data supporting this article cannot be made accessible in any publicly available repository according to the decision of the Stockholm Ethics Review Board (January 22, 2014, Approval No. 2013/2095-31/2). The kidney biopsy datasets analyzed for the current study were obtained from 4 different Swedish hospital records, regulated by patient secrecy, and hence cannot be made publicly available. The biopsy records were combined with national nationwide Swedish registries. Data from Swedish national registries are available to researchers worldwide after having obtained appropriate ethical approval. Further inquiries can be directed to the corresponding author.
