---

The Cosmic Hunger Strike: Why Supermassive Black Holes Are Growing So Slowly

What if the most powerful objects in the universe — monsters weighing millions to billions of times the mass of our Sun — are quietly going on a diet? And what if the reason is that the universe itself ran low on their favorite meal?

Welcome to FreeAstroScience, where we break down complex scientific principles into language that makes sense. We’re glad you’re here. Whether you’re an astronomy enthusiast, a curious student, or someone scrolling through your feed at midnight looking for something that sparks wonder, this one’s for you.

A groundbreaking study — published in The Astrophysical Journal in December 2025 and covered by Universe Today on April 2, 2026 — analyzed over 1.3 million galaxies and 8,000 actively growing supermassive black holes across billions of years of cosmic history. The findings? These gravitational giants are eating far more slowly than they did 10 billion years ago. And the main culprit isn’t what many scientists expected.

Stick with us until the end. We’ll walk you through the science, the data, and the human story behind one of extragalactic astronomy’s most important discoveries. No jargon walls. No gatekeeping. Just the universe, explained.

---

📑 Table of Contents

1. 1. What Exactly Is a Supermassive Black Hole?
2. 2. What Was the Cosmic Noon — And Why Does It Matter?
3. 3. What Is AGN Downsizing?
4. 4. How Did Astronomers Tackle This Mystery?
5. 5. What Are the Three Possible Explanations?
6. 6. The Eddington Ratio: What’s Driving the Slowdown?
7. 7. Where Did All the Cold Gas Go?
8. 8. How Do Black Holes and Galaxies Grow Together?
9. 9. The Math Behind the Mystery
10. 10. Key Numbers at a Glance
11. 11. What Comes Next for Black Hole Research?
12. 12. Final Thoughts

---

What Exactly Is a Supermassive Black Hole?

Before we talk about why these giants are slowing down, let’s make sure we’re all on the same page.

A supermassive black hole (SMBH) sits at the center of most large galaxies — including our own Milky Way. We’re not talking about the small, stellar-mass black holes born from dying stars. These are behemoths. They carry millions to billions of solar masses packed into a region smaller than our solar system.

Our galaxy’s central black hole, Sagittarius A*, weighs roughly 4 million solar masses. That sounds enormous. But it’s modest compared to others. Some SMBHs in distant galaxies tip the scales at billions of solar masses.

Here’s what we’ve known for a while: there are tight connections between a SMBH’s mass and the properties of its host galaxy — the bulge stellar mass, the velocity dispersion, and even the star formation rate . This tells us something profound. Black holes and their galaxies don’t grow independently. They evolve together, hand in hand, across cosmic time.

“These relations indicate that SMBHs and host galaxies evolve in a coordinated manner,” write Yu and colleagues in their 2025 paper . Tracing how SMBHs grow, then, is really tracing how galaxies themselves came to be.

---

What Was the Cosmic Noon — And Why Does It Matter?

Imagine the universe as a living city. Early on, construction crews are everywhere. Cranes fill the skyline. Buildings rise fast. That’s the Cosmic Noon.

The Cosmic Noon spanned roughly from 2 billion to 4 billion years after the Big Bang — or about 10 billion years ago . During this era, both star formation and SMBH growth hit their peak. Galaxies were assembling at a furious pace. Cold gas — the raw material for stars and the preferred fuel of supermassive black holes — was plentiful.

Then something changed.

After the Cosmic Noon, the rate of star formation slowed. And so did the growth of supermassive black holes. The cosmic SMBH accretion-rate density (called ρ_BHAR in technical language) peaked at a redshift of about z ≈ 1.5–2 and then dropped by roughly 1 to 1.5 orders of magnitude (a factor of 10 to 30) toward the present day .

In everyday terms: black holes went from feasting to fasting.

What Does Redshift Mean Here?

Quick primer. Astronomers use redshift (z) as a proxy for time. Higher redshift = farther away = earlier in the universe. A redshift of z ≈ 2 corresponds to about 10 billion years ago, when the universe was only 3–4 billion years old. A redshift of z ≈ 0 is today .

So when we say SMBH growth peaked at z ≈ 1.5–2, we’re talking about a time when the universe was young, vibrant, and overflowing with gas.

---

What Is AGN Downsizing?

When a SMBH actively swallows material, it becomes what astronomers call an Active Galactic Nucleus (AGN). The infalling material forms a spinning accretion disk that heats up to millions of degrees and blasts out radiation — especially X-rays .

Here’s the twist: the brightest AGN (those with the most luminous accretion) peaked earlier in cosmic history, while dimmer AGN peaked later . Scientists call this pattern “AGN downsizing.”

At first glance, this seems backward. In the standard picture of cosmic structure formation, smaller things assemble first and bigger things come later. So why would the most powerful black hole engines turn off before the weaker ones?

The answer, as the new study shows, isn’t that massive black holes formed before smaller ones. It’s that the rate of feeding — the intensity of accretion — has dropped everywhere, and it dropped fastest for the most voracious eaters .

AGN downsizing isn’t about creation order. It’s about appetite. And that appetite has been shrinking for 10 billion years.

---

How Did Astronomers Tackle This Mystery?

This brings us to the study itself. The paper, titled “The Drivers of the Decline in Supermassive Black Hole Growth at z < 2,” was led by Zhibo Yu, a graduate student in the Department of Astronomy and Astrophysics at Pennsylvania State University .

Yu and his team — including co-authors W. N. Brandt (Penn State), Fan Zou (University of Michigan), Bin Luo (Nanjing University), Qingling Ni (Max Planck Institute for Extraterrestrial Physics), D. P. Schneider (Penn State), and Fabio Vito (INAF Bologna) — analyzed data from nine well-characterized extragalactic survey fields .

The Wedding-Cake Design

The researchers used a layered approach they call a “wedding-cake design” . Picture a tiered wedding cake:

• The bottom tier (broad and shallow): Wide-area surveys from eROSITA and XMM-Newton covering up to 60 square degrees of sky, but at shallower depths.
• The middle tier: Medium-depth X-ray surveys from the LSST Deep-Drilling Fields, spanning about 13 square degrees with sensitive coverage from XMM-Newton and Chandra.
• The top tier (narrow and deep): Ultra-deep Chandra observations of the CANDELS fields — pencil-beam surveys reaching megasecond exposures, detecting the faintest and most distant AGN .

“By combining these data from different X-ray telescopes, we can construct a better picture of how these black holes are growing than any one telescope could do alone,” said co-author Fan Zou .

This design let the team sample AGN across a huge range of X-ray luminosities (log L_X from about 40 to 45 erg/s) and redshifts (z = 0 to 4), capturing the full story of SMBH growth over most of cosmic history .

---

What Are the Three Possible Explanations?

The researchers started with a straightforward question: why is SMBH growth slowing down?

“A longstanding mystery has been the cause of this big slowdown,” said lead author Yu in a press release .

They considered three potential explanations:

1. Slower accretion rates: Modern SMBHs are simply eating more slowly — their feeding intensity has dropped.
2. Smaller black hole masses: The “typical” SMBH contributing most of today’s accretion activity is less massive than it was in the past.
3. Fewer active black holes: There are simply fewer SMBHs that are actively accreting right now .

Each explanation would leave a different fingerprint on the data. Telling them apart required careful analysis of how accretion rates, black hole masses, and AGN number densities evolved across cosmic time.

---

The Eddington Ratio: What’s Driving the Slowdown?

To distinguish between those three explanations, we need to understand a concept called the Eddington ratio (λ_Edd).

What Is the Eddington Ratio?

Every black hole has a theoretical maximum feeding rate. Push too much material toward it, and the radiation pressure from the infalling matter actually blows the gas away. That limit is called the Eddington luminosity (L_Edd). The Eddington ratio is the fraction of that maximum:

Eddington Ratio

λ_Edd = L_bol / L_Edd

Where L_bol is the actual (bolometric) luminosity of the AGN, and L_Edd is its maximum possible luminosity.

A ratio near 1 means the black hole is feeding near its maximum. A ratio of 0.01 means it’s barely nibbling.

The Verdict

Yu et al. found that the Eddington ratio is the dominant driver of the decline in SMBH growth .

The typical Eddington ratio dropped by approximately 1.35 dex (about a factor of 22) from z ≈ 1.5–2 down to z ≈ 0.2 . In comparison, the typical black hole mass decreased by only 0.21 dex (about a factor of 1.6) over the same period .

That’s a difference of more than 13 times in impact. The decline in feeding rate completely overshadows the small shift in black hole mass.

And the effective AGN number density? It actually stayed roughly constant or even slightly increased — by about 0.29 dex — because as the typical Eddington ratio fell, the main contributors to total accretion shifted toward the low-luminosity AGN that were always more numerous .

So the answer is clear: black holes haven’t gotten smaller. They haven’t disappeared. They’ve just stopped eating as fast.

---

Where Did All the Cold Gas Go?

If the feeding rate is the problem, the next question is obvious: why?

The answer points to the universe’s pantry running low.

Supermassive black holes mostly feed on cold gas — the same material that stars form from . During the Cosmic Noon, this cold gas was abundant. Galaxies were rich with it. Mergers between galaxies were frequent, funneling vast streams of gas toward galactic centers, fueling both star formation and SMBH accretion .

But as the universe expanded and aged, several things happened:

• Gas was consumed. Generations of stars and black hole feeding events used up the available cold gas supply.
• Mergers became less common. Galaxy-scale gas inflows and major mergers — known drivers of both star formation and SMBH fueling — declined in frequency .
• Gas reservoirs weren’t replenished fast enough. The rate of gas cooling and infall from the cosmic web slowed.

“It appears that black holes’ consumption of material has greatly slowed down as the universe has aged,” said co-author Niel Brandt of Penn State. “This is probably because the amount of cold gas available for them to ingest has decreased since cosmic noon.”

Think of it like a pond drying up in a long summer. The fish don’t vanish. They just can’t eat as much because there’s less water and fewer nutrients to go around. The black holes are still there. They’re just hungry.

---

How Do Black Holes and Galaxies Grow Together?

One of the most remarkable findings of modern astrophysics is that black holes and their host galaxies grow in lockstep. This isn’t coincidence — it’s coevolution.

The long-term average black hole accretion rate correlates with the host galaxy’s total stellar mass and the star formation rate in its bulge . When star formation peaks, SMBH growth peaks. When star formation declines, SMBH growth declines.

The Yu et al. study reinforces this picture. At a fixed redshift, galaxies with more stellar mass (M★) tend to host more luminous AGN . The main reason? More massive galaxies host more massive black holes, and those bigger engines naturally produce more radiation when they feed.

Does Galaxy Mass Affect Feeding Rate or Feeding Frequency?

The researchers asked a clever follow-up: Does the mass of a galaxy mainly determine how brightly its black hole shines during active episodes (outburst luminosity), or how often those episodes occur (duty cycle)?

The answer: outburst luminosity dominates .

At z ≲ 1, the decline in average BHAR from massive to less-massive galaxies is driven almost entirely by differences in the typical brightness of active episodes — not by how often the black hole turns on. The AGN duty cycle (f_AGN) shows weaker dependence on stellar mass, especially at lower redshifts .

At higher redshifts (z ≳ 1), the duty cycle does show some mass dependence, likely because large amounts of available gas preferentially trigger luminous AGN episodes in massive galaxies with deeper gravitational wells .

---

The Math Behind the Mystery

For those who enjoy the quantitative side, let’s look at how the researchers decomposed SMBH growth. The cosmic SMBH accretion-rate density can be factored into three redshift-dependent components :

Decomposition of Cosmic SMBH Accretion-Rate Density

ρ_BHAR = n^eff_AGN · ⟨λ_Edd⟩ · ⟨M_BH⟩ · (1 − ε) × 1.26 × 10³⁸ / (ε c²)

Where:
ρ_BHAR = cosmic SMBH accretion-rate density
n^eff_AGN = effective AGN number density
⟨λ_Edd⟩ = typical Eddington ratio of the AGN contributing most to ρ_BHAR
⟨M_BH⟩ = typical black hole mass of those AGN
ε = radiative efficiency (assumed 0.1)
c = speed of light

The change in ρ_BHAR from the z = 1.5–2 bin to z = 0.2–0.5 breaks down logarithmically as :

Logarithmic Decomposition of the Decline

Δ log ρ_BHAR = Δ log n^eff_AGN + Δ log ⟨λ_Edd⟩ + Δ log ⟨M_BH⟩

Component	Change (dex)	Direction
Δ log ρBHAR	−1.28	↓ Decline
Δ log ⟨λEdd⟩	−1.35	↓ Dominant driver
Δ log ⟨MBH⟩	−0.21	↓ Minor contribution
Δ log neffAGN	+0.29	↑ Slight increase

Source: Yu et al. 2025, Table 1 (The Astrophysical Journal, 995:205)

The numbers speak for themselves. The Eddington ratio drop of −1.35 dex accounts for essentially the entire observed decline. The small decrease in black hole mass (−0.21 dex) barely registers. And the effective number density of AGN actually went up slightly, because the “typical” accretion shifted to the more numerous, low-luminosity population .

---

Key Numbers at a Glance

Let’s consolidate the critical numbers from this study:

Parameter	Value
Galaxies studied	~1,313,000
X-ray AGN studied	~7,268 (z = 0–4)
Peak SMBH growth epoch	z ≈ 1.5–2 (~10 billion years ago)
Total ρBHAR decline (z ≈ 2 → z ≈ 0.2)	−1.28 ± 0.08 dex
Typical Eddington ratio decline	−1.35 dex (factor ~22)
Typical MBH decline	−0.21 dex (factor ~1.6)
Effective AGN number density change	+0.29 dex (roughly constant)
Adopted MBH–M★ relation	MBH = 0.002 × M★
Adopted radiative efficiency (ε)	0.1

Data from Yu et al. 2025, The Astrophysical Journal, 995:205

---

What Comes Next for Black Hole Research?

Science doesn’t stop at answers — it stops at better questions. And this study opens up several exciting paths forward.

New Telescopes, New Data

The team notes that upcoming missions and surveys will dramatically sharpen our view :

• LSST (Vera C. Rubin Observatory) will provide unprecedented photometric coverage of the southern sky.
• Euclid and Roman space telescopes will deliver deep infrared surveys.
• NewAthena, AXIS, and Lynx — future X-ray observatories — will probe AGN populations hidden by heavy obscuration.
• JWST is already revealing AGN that X-ray surveys missed, especially at high redshifts.

These new datasets will let researchers extend the wedding-cake design to much larger areas and deeper depths. They’ll help answer questions the current study couldn’t fully address — like what’s happening at z > 2, where obscured accretion is harder to measure and JWST has begun finding “X-ray-weak” AGN that current surveys may have missed .

Unanswered Questions

The study focused on z < 2 — the decline phase. The rise of SMBH growth from z ≈ 4 to z ≈ 2 is trickier to characterize because obscured AGN fractions increase at earlier epochs . There may be hidden accretion at high redshift that we’re only now beginning to detect.

And there’s the question of Compton-thick AGN — those so heavily shrouded in gas and dust that even X-rays struggle to escape. The missing accretion from these sources represents a systematic uncertainty of about 0.2 dex in the measurements . Future missions with greater penetrating power should clarify this.

---

Final Thoughts

Let’s take a step back.

Ten billion years ago, the universe was alive with growth. Galaxies were forming stars at record rates. Supermassive black holes at their centers were gorging on rivers of cold gas, shining so brightly that we can detect them across the observable universe.

Then the pantry emptied. The cold gas — that fuel shared by both stars and black holes — became scarce. Mergers slowed. The feeding frenzy ended.

Today, supermassive black holes sit at the hearts of their galaxies, still massive, still gravitationally dominant, but barely eating. Their Eddington ratios have plummeted by a factor of 22 since the Cosmic Noon . It’s not that they’ve shrunk. It’s not that they’ve vanished. They’re simply running on empty.

What strikes us most about this discovery is the poetry of it. The universe and its black holes aged together. As the cosmos cooled and expanded, its greatest engines wound down in parallel. There’s a kind of quiet dignity in that — a cosmic retirement, earned after billions of years of furious activity.

This research, led by Zhibo Yu and colleagues, gives us the most compelling evidence yet that the decline in SMBH growth is driven by the falling Eddington ratio — the feeding intensity — rather than by shifts to smaller black holes or fewer active ones . It’s a clean, elegant answer to a question that has puzzled astrophysicists for decades.

At FreeAstroScience, we believe that understanding the universe isn’t a luxury. It’s a necessity. As the great Goya etching reminds us: “El sueño de la razón produce monstruos” — the sleep of reason breeds monsters. We write these articles so you never have to stop thinking, never have to stop asking, never have to stop wondering.

Come back to FreeAstroScience.com anytime. We’ll be here, translating the language of the cosmos into something all of us can share.

---

📚 References & Sources

1. Yu, Z., Brandt, W. N., Zou, F., Luo, B., Ni, Q., Schneider, D. P., & Vito, F. (2025). “The Drivers of the Decline in Supermassive Black Hole Growth at z < 2.” The Astrophysical Journal, 995, 205. doi:10.3847/1538-4357/ae173d
2. Gough, E. (2026, April 2). “Why Are Supermassive Black Holes Growing So Slowly?” Universe Today. universetoday.com
3. Zou, F. et al. (2024). Best current measurements of sample-averaged SMBH accretion rates from nine extragalactic survey fields.
4. Madau, P. & Dickinson, M. (2014). “Cosmic Star-Formation History.” Annual Review of Astronomy and Astrophysics, 52, 415–486.
5. Ueda, Y. et al. (2014). “Toward the Cosmic X-Ray Background: A Combined Analysis.” The Astrophysical Journal, 786, 104.
6. Weaver, J. R. et al. (2023). Galaxy stellar mass functions used for convolution in this study.
7. NASA/CXC — Chandra X-ray Observatory data and imaging. chandra.harvard.edu