diff --git a/AppImage/components/health-status-modal.tsx b/AppImage/components/health-status-modal.tsx
index af219ff4..556f86dd 100644
--- a/AppImage/components/health-status-modal.tsx
+++ b/AppImage/components/health-status-modal.tsx
@@ -105,6 +105,8 @@ export function HealthStatusModal({ open, onOpenChange, getApiUrl }: HealthStatu
     setError(null)
 
     try {
+      let newOverallStatus = "OK"
+      
       // Use the new combined endpoint for fewer round-trips
       const response = await fetch(getApiUrl("/api/health/full"))
       if (!response.ok) {
@@ -114,14 +116,17 @@ export function HealthStatusModal({ open, onOpenChange, getApiUrl }: HealthStatu
         const data = await legacyResponse.json()
         setHealthData(data)
         setDismissedItems([])
+        newOverallStatus = data?.overall || "OK"
       } else {
         const fullData: FullHealthData = await response.json()
         setHealthData(fullData.health)
         setDismissedItems(fullData.dismissed || [])
+        newOverallStatus = fullData.health?.overall || "OK"
       }
 
+      // Emit event with the FRESH data from the response, not the stale state
       const event = new CustomEvent("healthStatusUpdated", {
-        detail: { status: healthData?.overall || "OK" },
+        detail: { status: newOverallStatus },
       })
       window.dispatchEvent(event)
     } catch (err) {
@@ -129,7 +134,7 @@ export function HealthStatusModal({ open, onOpenChange, getApiUrl }: HealthStatu
     } finally {
       setLoading(false)
     }
-  }, [getApiUrl, healthData?.overall])
+  }, [getApiUrl])
 
   useEffect(() => {
     if (open) {
diff --git a/AppImage/scripts/flask_health_routes.py b/AppImage/scripts/flask_health_routes.py
index c4528fab..34f1021b 100644
--- a/AppImage/scripts/flask_health_routes.py
+++ b/AppImage/scripts/flask_health_routes.py
@@ -83,6 +83,12 @@ def acknowledge_error():
                 health_monitor.last_check_times.pop(cache_key, None)
                 health_monitor.cached_results.pop(cache_key, None)
             
+            # Also invalidate overall status caches so header updates immediately
+            health_monitor.last_check_times.pop('_bg_overall', None)
+            health_monitor.cached_results.pop('_bg_overall', None)
+            health_monitor.last_check_times.pop('overall_health', None)
+            health_monitor.cached_results.pop('overall_health', None)
+            
             # Determine suppression period for the response
             category = result.get('category', '')
             if category == 'updates':
diff --git a/AppImage/scripts/flask_server.py b/AppImage/scripts/flask_server.py
index 71466c7f..85c58b4a 100644
--- a/AppImage/scripts/flask_server.py
+++ b/AppImage/scripts/flask_server.py
@@ -569,6 +569,34 @@ def _temperature_collector_loop():
         time.sleep(60)
 
 
+def _health_collector_loop():
+    """Background thread: run full health checks every 5 minutes.
+    Keeps the health cache always fresh and records events/errors in the DB
+    so the future notification service can consume them."""
+    from health_monitor import health_monitor
+    
+    # Wait 30s after startup to let other services initialize
+    time.sleep(30)
+    
+    while True:
+        try:
+            # Run full health check (results get cached internally + recorded in DB)
+            result = health_monitor.get_detailed_status()
+            
+            # Update the quick-status cache so the header stays fresh without extra work
+            overall = result.get('overall', 'OK')
+            summary = result.get('summary', 'All systems operational')
+            health_monitor.cached_results['_bg_overall'] = {
+                'status': overall,
+                'summary': summary
+            }
+            health_monitor.last_check_times['_bg_overall'] = time.time()
+        except Exception as e:
+            print(f"[ProxMenux] Health collector error: {e}")
+        
+        time.sleep(300)  # Every 5 minutes
+
+
 def get_uptime():
     """Get system uptime in a human-readable format."""
     try:
@@ -7006,6 +7034,15 @@ if __name__ == '__main__':
     else:
         print("[ProxMenux] Temperature history disabled (DB init failed)")
 
+    # ── Background Health Monitor ──
+    # Run full health checks every 5 min, keeping cache fresh and recording events for notifications
+    try:
+        health_thread = threading.Thread(target=_health_collector_loop, daemon=True)
+        health_thread.start()
+        print("[ProxMenux] Background health monitor started (5 min interval)")
+    except Exception as e:
+        print(f"[ProxMenux] Background health monitor failed to start: {e}")
+
     # Check for SSL configuration
     ssl_ctx = None
     try:
diff --git a/AppImage/scripts/health_monitor.py b/AppImage/scripts/health_monitor.py
index e79af507..0334cdd9 100644
--- a/AppImage/scripts/health_monitor.py
+++ b/AppImage/scripts/health_monitor.py
@@ -201,17 +201,25 @@ class HealthMonitor:
     def get_cached_health_status(self) -> Dict[str, str]:
         """
         Get cached health status without running expensive checks.
-        Returns the last calculated status or triggers a check if too old.
+        The background health collector keeps '_bg_overall' always fresh (every 5 min).
+        Falls back to calculating on demand if background data is stale or unavailable.
         """
-        cache_key = 'overall_health'
         current_time = time.time()
         
-        # If cache exists and is less than 60 seconds old, return it
+        # 1. Check background collector cache (updated every 5 min by _health_collector_loop)
+        bg_key = '_bg_overall'
+        if bg_key in self.last_check_times:
+            age = current_time - self.last_check_times[bg_key]
+            if age < 360:  # 6 min (5 min interval + 1 min tolerance)
+                return self.cached_results.get(bg_key, {'status': 'OK', 'summary': 'System operational'})
+        
+        # 2. Check regular cache (updated by modal fetches or on-demand)
+        cache_key = 'overall_health'
         if cache_key in self.last_check_times:
             if current_time - self.last_check_times[cache_key] < 60:
                 return self.cached_results.get(cache_key, {'status': 'OK', 'summary': 'System operational'})
         
-        # Otherwise, calculate and cache
+        # 3. No fresh cache - calculate on demand (happens only on first load before bg thread runs)
         status = self.get_overall_status()
         self.cached_results[cache_key] = {
             'status': status['status'],