Incidence rate patterns, cumulative incidence, and time trends for moderate and severe albuminuria in individuals diagnosed with type 1 diabetes aged 0-14 years: a population-based retrospective cohort study
Fanny Jansson Sigfrids,Per-Henrik Groop,Valma Harjutsalo
DOI: https://doi.org/10.1016/S2213-8587(22)00099-7
Abstract:Background: The incidence and temporal trends of moderate and severe albuminuria during recent decades are poorly described in type 1 diabetes. We aimed to assess diabetes duration-specific incidence rates, cumulative incidence, and secular trends of albuminuria in type 1 diabetes in Finland. Methods: We conducted a population-based, retrospective cohort study of a stratified random sample (n=1500) of all individuals diagnosed with type 1 diabetes before age 15 years during 1970-99 in Finland. The sampling frame was the database of the Finnish Institute for Health and Welfare. Individuals with an atypical clinical course, presentation of non-diabetic kidney disease, insufficient albumin excretion rate measurements, or unavailable medical records were excluded (final sample n=1430). Study participants were followed up until death, the event of interest (moderate or severe albuminuria or kidney failure), or the most recent event-free date. Medical records retrieved up to Dec 31, 2020 were systematically reviewed for albuminuria determinations. Moderate and severe albuminuria were categorised on the basis of international reference limits (two of three consecutive urine samples). Kidney failure was defined as dialysis treatment or kidney transplant. Cohorts defined by calendar year of diabetes diagnosis (1970-79, 1980-89, and 1990-99) were assessed. Patterns of duration-specific incidences were evaluated by fitting generalised additive models to the data, which were split into multiple observations of half-year duration. Cumulative incidences were calculated with Kaplan-Meier analysis. In analyses with kidney failure as the endpoint, competing risk for mortality was incorporated. Findings: In our stratified random sample, 462 individuals were diagnosed with diabetes in 1970-79, 481 were diagnosed in 1980-89, and 487 were diagnosed in 1990-99. The incidence rate pattern of severe albuminuria changed over time; a peak at 15-19 years since diabetes onset in the 1970-79 cohort was not replicated in those diagnosed later. In the combined 1980-99 diagnosis-year cohorts, the incidence rate rose during the first 14 years after diabetes onset, after which it levelled off to a plateau. Between the 1970-79 and 1980-89 diabetes diagnosis cohorts, the cumulative incidence of severe albuminuria had approximately halved (hazard ratio [HR] 0·55 [95% CI 0·42-0·72] with the 1970-79 cohort as reference, p<0·0001), whereas, between the 1980-89 and 1990-99 cohorts, no further decrease was observed (HR 0·83 [0·54-1·26] with the 1980-89 cohort as reference, p=0·38). The 25-year cumulative incidence for severe albuminuria was 26·8% (22·6-30·8) in the 1970-79 diagnosis cohort, 12·0% (9·0-15·0) in the 1980-89 cohort, and 10·8% (6·7-14·6) in the 1990-99 cohort. 15 years after onset of severe albuminuria, cumulative progression rate from severe albuminuria to kidney failure was 35·2% (27·4-43·0) in the 1970-79 cohort and 35·6% (24·3-47·0) in the 1980-99 cohorts combined (Gray's test p=0·37). In the cohorts with data on moderate albuminuria (1980-89 and 1990-99), cumulative incidence of moderate albuminuria showed no calendar effect between the earlier and later cohorts (HR 0·99 [0·78-1·28] with the 1980-89 cohort as reference, p=0·97). The incidence rate of moderate albuminuria increased until 10 years after diabetes onset, then remained mostly stable until starting to decrease at around 25 years after diabetes onset. Interpretation: Our analyses show that the cumulative incidence of severe albuminuria has decreased between 1970-79 and 1980-99; however, whether this decrease solely denotes a delay in albuminuria, or also a true prevention of albuminuria, needs to be investigated further. Nevertheless, diabetic kidney disease remains a significant complication of type 1 diabetes. Due to the robust association of diabetic kidney disease with premature mortality, novel therapies to improve prognosis are needed. Funding: Folkhälsan Research Foundation, Medical Society of Finland, Wilhelm and Else Stockmann Foundation, Finnish Diabetes Research Foundation, Waldemar von Frenckell Foundation, Liv och Hälsa Society, Academy of Finland, and Novo Nordisk Foundation. Translations: For the Finnish and Swedish translations of the abstract see Supplementary Materials section.