FLOW Safety Across CKD Severity

A prespecified secondary analysis of FLOW found no unexpected safety pattern for once-weekly semaglutide across the baseline kidney-function range represented in the trial. Serious adverse events were numerically less frequent with semaglutide than placebo in the reported estimated glomerular filtration rate, or eGFR, subgroups. Adverse events leading to treatment discontinuation became more common as baseline eGFR declined in both groups, while discontinuations attributed to gastrointestinal events were more frequent with semaglutide^[1].

This is not evidence that risk was absent or identical at every level of kidney function. The authors described event frequencies rather than reporting a formal test that proved equivalence between eGFR subgroups. FLOW also enrolled a defined population with type 2 diabetes, chronic kidney disease and albuminuria. Our semaglutide research overview provides broader context; this readout focuses on what the subgroup safety report measured and what it leaves unresolved.

How FLOW was designed

FLOW was a phase 3, randomized, double-blind, placebo-controlled trial. It assigned 3,533 adults with type 2 diabetes and chronic kidney disease to subcutaneous semaglutide or placebo alongside standard care. Participants in the active group escalated to a maintenance dose of 1.0 mg once weekly. The main trial followed participants for a median 3.4 years before early stopping was recommended at a prespecified interim analysis^[2][3].

Eligibility required an eGFR from 50 to 75 mL/min/1.73 m² with a urine albumin-to-creatinine ratio above 300 and below 5,000 mg/g, or an eGFR from 25 to below 50 with a ratio above 100 and below 5,000. Participants generally also had to be receiving a stable, tolerated renin-angiotensin system blocker unless it was contraindicated or not tolerated^[2][3]. These criteria selected people with substantial albuminuric CKD. They do not represent all causes or severities of kidney disease.

What the safety analysis counted

The 2026 report was prespecified, but its assessment was descriptive. Safety data were collected selectively and included serious adverse events, adverse events that led participants to stop trial treatment, and prespecified adverse events of special interest. The researchers then reported these events numerically by baseline eGFR^[1]. Selective collection means the analysis should not be read as a complete inventory of every symptom experienced during follow-up.

A serious adverse event is a regulatory reporting category based on outcomes such as death, life-threatening illness, hospitalization or major disability. It does not automatically mean the study drug caused the event. Participants in FLOW had diabetes and chronic kidney disease, so many medical events could arise from underlying disease, concurrent treatment or other causes. Randomization supports treatment-group comparison, but attribution still depends on the event and clinical assessment.

Serious events across eGFR groups

In the subgroup with baseline eGFR of at least 60 mL/min/1.73 m², serious adverse events were reported in 45.1% of participants assigned semaglutide and 50.4% assigned placebo. In the subgroup below 30 mL/min/1.73 m², the corresponding frequencies were 56.4% and 60.4%^[1]. The higher percentages in the lower-eGFR subgroup appeared in both treatment arms, which is consistent with a population carrying more advanced kidney impairment rather than a semaglutide-specific increase.

The report says serious events involving infections, cardiac disorders, respiratory or thoracic disorders, and metabolism or nutrition disorders were generally less frequent with semaglutide than placebo regardless of baseline eGFR. It also reports similar frequencies of prespecified adverse events of special interest across treatment groups within the eGFR subgroups^[1]. The word generally matters. The abstract does not provide event counts for every category or uncertainty intervals for each subgroup comparison.

Discontinuation rose as kidney function fell

Adverse events led to treatment discontinuation more often at lower baseline eGFR in both randomized groups. Among participants starting at 60 mL/min/1.73 m² or higher, 11.2% stopped semaglutide because of an adverse event and 11.3% stopped placebo. Below 30 mL/min/1.73 m², the figures were 18.8% and 18.7%^[1]. The near-matching treatment-group percentages at each reported extreme do not prove that the reasons for stopping were identical.

Gastrointestinal adverse events produced a different treatment pattern. Discontinuations specifically attributed to gastrointestinal events were reported in 4.5% with semaglutide and 1.1% with placebo, with the higher semaglutide frequency seen regardless of baseline eGFR^[1]. This is compatible with the established gastrointestinal tolerability burden of GLP-1 receptor agonists, but the subgroup abstract does not list individual gastrointestinal diagnoses, timing or severity.

How this fits the main FLOW result

The main FLOW publication reported a primary kidney-outcome event in 331 semaglutide participants and 410 placebo participants, corresponding to a hazard ratio of 0.76. Serious adverse events occurred in 49.6% and 53.8%, respectively^[2]. The later analysis adds a kidney-function lens to that overall safety summary. It does not replace the main efficacy analysis, and the efficacy result should not be used to dismiss gastrointestinal intolerance or individual treatment constraints.

FLOW was sponsored by Novo Nordisk, which also funded the main trial^[2][3]. Industry sponsorship does not negate randomization or masking. It remains relevant context when evaluating study design, selective safety collection and reporting, especially for a secondary analysis focused on tolerability.

What we do not yet know

The accessible abstract provides percentages for the highest and lowest eGFR categories but not the denominators, confidence intervals or every intermediate subgroup. It describes comparisons as numerical and descriptive^[1]. Without formal interaction testing in the available record, the analysis cannot establish that semaglutide has exactly the same safety profile at all kidney-function levels.

Rare events may remain difficult to assess even in a trial of this size, particularly after division into kidney-function subgroups. The results also do not directly cover dialysis, most non-albuminuric CKD, type 1 diabetes, hereditary or autoimmune kidney disease, or people outside the trial's treatment and monitoring conditions. A clinician must interpret current prescribing information, kidney function, other medicines, hydration risk and symptoms for an individual patient.

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.