diff --git a/psycop/common/feature_generation/loaders/raw/load_structured_sfi.py b/psycop/common/feature_generation/loaders/raw/load_structured_sfi.py
index ff3fde427..5475d0a8e 100644
--- a/psycop/common/feature_generation/loaders/raw/load_structured_sfi.py
+++ b/psycop/common/feature_generation/loaders/raw/load_structured_sfi.py
@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Literal
 
 import polars as pl
 
@@ -295,21 +295,34 @@ def smoking_categorical(mapping: dict[str, int] | None = None) -> pd.DataFrame:
     ).to_pandas()
 
 
-def systolic_blood_pressure() -> pd.DataFrame:
-    df = pl.from_pandas(
-        sql_load(
-            query="SELECT * FROM [fct].[FOR_SFI_Blodtyk_Puls_psyk_somatik_inkl_2021]",
-        ),
+def _get_blood_pressure_pulse(
+    subtype: Literal["Systolisk", "Diastolisk", "Pulsslag / min"]
+) -> pl.LazyFrame:
+    df = (
+        pl.from_pandas(
+            sql_load(
+                query="SELECT * FROM [fct].[FOR_SFI_Blodtyk_Puls_psyk_somatik_inkl_2021]",
+            ),
+        )
+        .lazy()
+        .rename(
+            {"datotid_senest_aendret_i_sfien": "timestamp", "numelementvaerdi": "value"}
+        )
     )
 
-    df_pl_subset = df.select(
-        [
-            "dw_ek_borger",
-            "datotid_senest_aendret_i_sfien",
-            "numelementvaerdi",
-        ],
-    ).filter(pl.col("numelementvaerdi").is_not_null())
+    subset = df.select(["dw_ek_borger", "timestamp", "value", "elementledetekst"])
+    return df.filter(pl.col("elementledetekst") == pl.lit(subtype)).select(
+        ["dw_ek_borger", "timestamp", "value"]
+    )
 
-    return df_pl_subset.rename(
-        {"datotid_senest_aendret_i_sfien": "timestamp", "numelementvaerdi": "value"},
-    ).to_pandas()
+
+def systolic_blood_pressure() -> pd.DataFrame:
+    df = _get_blood_pressure_pulse(subtype="Systolisk")
+
+    return df.collect().to_pandas()
+
+
+def diastolic_blood_pressure() -> pd.DataFrame:
+    df = _get_blood_pressure_pulse(subtype="Diastolisk")
+
+    return df.collect().to_pandas()