Retrograde vs anterograde amnesia: which way the memory loss runs, and why neither is interference

Retrograde amnesia is the loss of memories that a person formed before an injury or illness. Anterograde amnesia is the loss of the ability to form new long-term memories after the injury or illness. A person who cannot remember the year before a head injury has retrograde amnesia. A person who cannot remember the events of each day after the injury has anterograde amnesia. The two types usually occur together.

The definitions are short. The names cause the problem. "Antero" sounds like "before", and both words sound like retroactive and proactive interference, which are a different subject. This article explains each term, the origin of the two words, the findings from patients, and a method to sort exam questions.

• Both names give a direction in time, and you count it from the event that caused the amnesia. Retrograde amnesia runs backward from the event. Anterograde amnesia runs forward.
• Anterograde amnesia affects new facts and events. The patient H.M. had severe anterograde amnesia, but he learned a new motor skill.
• Retrograde amnesia usually affects recent memories more than old memories. This pattern is called Ribot's law, and not all researchers accept it.
• The two types usually occur together. In a 2013 study of 11 patients, more severe anterograde amnesia went with more extensive retrograde amnesia.
• Amnesia follows an injury, an illness or a psychological trauma. Interference occurs in healthy memory.

The difference in one table

The two definitions come from the MSD Manual's page on amnesias, a medical reference. The APA Dictionary of Psychology says the same in other words. It calls anterograde amnesia the "inability to learn new information". It calls retrograde amnesia the "inability to recall previously learned information or past events".

Both patients in the last row had both types of amnesia, which is the usual case.

Where the two words come from

Both words end in "-grade", from the Latin verb gradi, "to go". The American Heritage Dictionary traces retrograde to the Latin retrogradi, "to go back". Retrograde amnesia goes back in time from the injury.

"Antero-" is the difficult half. It belongs with the word anterior. The first meaning of anterior in the same dictionary is "Placed before or in front". Use that meaning. Anterograde amnesia goes to the front, which is forward in time from the injury.

The dictionary also gives a second meaning of anterior: "Occurring before in time; earlier". This second meaning causes the error. A student who reads "antero" as "earlier" puts anterograde amnesia before the injury, and that is the wrong side.

The rule is the same for both words. Start at the event that caused the amnesia. Retrograde amnesia covers the time behind that point. Anterograde amnesia covers the time in front of it.

Anterograde amnesia: new memories do not form

Anterograde amnesia is "an impaired capacity for new learning", in the words of a 2013 study by Christine Smith and colleagues. Old memories can stay. New facts and events do not go into long-term memory.

The best-known case is H.M. He had severe epileptic seizures. In 1953, at the age of 27, he had an experimental operation that removed tissue from the inner side of both temporal lobes. The operation controlled the seizures.

But the researcher Brenda Milner found a severe memory impairment. H.M. "forgot daily events nearly as fast as they occurred", as Larry Squire wrote in a 2009 review of the case. H.M. described his state as "like waking from a dream ... every day is alone in itself". William Scoville, his surgeon, and Milner published the first report in 1957. H.M. died in 2008 at the age of 82.

The parts of memory that stayed intact are as important as the parts that failed. Squire's review describes two:

• H.M. could repeat a string of six or seven digits. He could hold a three-digit number in mind for 15 minutes if he rehearsed it without a pause. When his attention moved to a different topic, he forgot the number and the task.
• He learned a new motor skill. Milner asked him to trace a star while he saw his hand only in a mirror. He improved during ten trials and kept the skill for three days. At the end, he did not remember that he had done the task before.

So anterograde amnesia is not a failure of all learning. It affects conscious memory for facts and events, which researchers call declarative memory. Skills are a different kind of memory. Our article on episodic and semantic memory shows where these types sit.

Squire also notes that the impairment of H.M. "was not absolute". He was able to acquire significant new knowledge.

One caution applies to this case. The 1957 report is sometimes cited as evidence for the role of the hippocampus. Squire notes that this is incorrect, because the operation also removed the amygdala and the adjacent parahippocampal gyrus. Damage limited to the hippocampus causes "considerably less impairment than in H.M."

Retrograde amnesia: recent memories are usually lost first

Retrograde amnesia is the "loss of information that was acquired before the onset of memory impairment". That definition is from a 2007 review by Larry Squire and Peter Bayley. The loss can cover a few years or several decades.

Théodule Ribot described the usual pattern in 1881. Elizabeth Warrington quotes his rule in a 1996 review: "The dissolution of memory is inversely related to the recency of the event." The rule is now called Ribot's law. In plain words, the most recent memories are the most vulnerable, and old memories are the safest.

H.M. fits the rule. Squire writes that the 1957 report described "a partial loss of memory (retrograde amnesia) for the 3 years leading up to his surgery". The report called his earlier memories "seemingly normal".

The size of the loss depends on the size of the damage. Squire and Bayley give the first figure below, and Smith and colleagues give the other two:

• In six patients with damage limited to the hippocampus, retrograde amnesia for news events "extended no more than a few years".
• Patients with larger lesions had a mean retrograde amnesia of 35 years.
• The patient E.P. had the most severe loss. It covered as much as 50 years. His scores reached normal levels for events from the time when he was 20 to 24 years old.

The usual explanation is memory consolidation. In Squire's account, the structures removed in H.M. form a memory and maintain it "for a period of time after learning". Later, other parts of the cortex can support the memory without them. An old memory is thus safe from damage that destroys a new memory.

This explanation is the majority view, not a settled fact. Squire and Bayley write that the literature "does not reflect an easy consensus". Warrington reported patients whose retrograde loss was lifelong, with no sparing of the oldest memories. She argues that the gradient in other studies is "an artifact of item selection".

A third position, multiple trace theory, holds that detailed personal memories depend on the same structures for as long as the memories exist. The estimates for H.M. also changed. A 1984 report put his retrograde loss at 11 years, and Squire advises caution about the later data.

Why the two usually occur together

Smith and colleagues write that the two types "tend to occur together in the same patients". They also write that the severity of one "is usually correlated" with the severity of the other.

Their own study measured both types in 11 patients with damage to the inner temporal lobes on both sides. The result was an orderly relation. Patients with more severe anterograde amnesia also had more extensive retrograde amnesia. Retrograde amnesia became measurable only when anterograde amnesia was already substantial.

The same paper reports older figures from a 1946 study of head injuries by Russell and Nathan. Of 503 patients whose anterograde amnesia lasted one day or less, 32 had no retrograde amnesia at all. Of 185 patients with gunshot wounds to the head, 65 had a definite period of anterograde amnesia and no retrograde amnesia.

Smith and colleagues draw one conclusion from both sets of data: "anterograde memory is easier to disrupt than retrograde memory." New learning fails first. Stored memories are harder to damage.

There are exceptions in both directions. Anterograde amnesia can occur alone, as the gunshot cases show. Researchers have also reported a retrograde loss that is much larger than the anterograde loss. These cases are associated with damage to the outer temporal cortex. The 2013 study has a limit too. It tested memory for public events and did not test memory for personal events.

What causes each pattern

The MSD Manual lists traumatic brain injury, degeneration, metabolic disorders, seizures and psychological disturbances as causes of amnesia. The causes do not divide cleanly into retrograde causes and anterograde causes. Several causes produce both types, as the table shows.

The first row is from the MSD Manual's page on amnesias. Rows two, three and five are from its pages on transient global amnesia, Korsakoff psychosis and dissociative amnesia. The fourth row describes H.M. and E.P.

Two details help with exam questions. Transient global amnesia occurs mostly in people aged 50 to 70, and an episode lasts a mean of 6 hours. Dissociative amnesia is "usually caused by trauma or stress". The APA Dictionary places it with the psychological causes of amnesia, not with injury or disease.

This table explains terms. It cannot identify the cause of a person's memory loss. The Cleveland Clinic states that sudden memory loss is always a sign to get medical attention. Memory loss after a head injury needs medical attention immediately. The NHS advises you to see a GP if memory problems affect your day-to-day life.

What films get wrong

In many films, a blow to the head removes the past and the identity of the hero, but the hero learns new things without difficulty. A second blow then repairs the memory. The neuropsychologist Sallie Baxendale compared film amnesia with the medical literature in a 2004 review in the BMJ. She concludes: "Most amnesic conditions in films bear little relation to reality."

The real pattern is close to the opposite. In amnesia with a neurological cause, Baxendale writes, intelligence and attention span are normal. The patient has "severe and permanent difficulties in taking in new information". Then she adds: "Personality and identity are unaffected."

She also names accurate films. In Memento, from 2000, the main character has severe anterograde amnesia. He keeps his identity and has little retrograde amnesia. In Finding Nemo, from 2003, the fish Dory cannot learn or keep new information. Baxendale calls Dory one of the most accurate portrayals in cinema.

Amnesia is not interference

Four terms cause confusion here: retrograde amnesia, anterograde amnesia, retroactive interference and proactive interference. They are two different pairs.

Amnesia is a loss of memory after an injury, an illness or a trauma. Interference occurs in healthy memory, when two sets of learning compete. OpenStax Psychology 2e defines proactive interference as the case "when old information hinders the recall of newly learned information". Retroactive interference occurs "when information learned more recently hinders the recall of older information".

Two questions sort each example:

1. 1.Is there an injury, an illness or a trauma that divides time into before and after? If yes, the term is an amnesia. If two sets of ordinary learning compete, the term is an interference.
2. 2.Which memories are affected? If the older memories are affected, the term starts with "retro". If the newer memories are affected, the term is the other one of the pair.

Our article on proactive and retroactive interference gives the history and the evidence for the second pair.

Classifying exam scenarios

The scenarios below are typical examples. Apply the two questions to each one.

Exam questions usually ask for one term. Real patients usually have both types of amnesia, so read the scenario for the time that it stresses.

How to keep the terms apart

Three memory aids help:

• Start at the event. Retrograde goes back from it. Anterograde goes forward from it.
• Think of a retro style, which looks back to the past. Think of anterior, which means in front.
• Use the letter A. Anterograde amnesia is amnesia for the time after the event.

For the four-term grid, use one more aid. Amnesia needs an injury. Interference needs two sets of learning.

A memory aid helps you only if you practise with it. Our guide to memorising psychology terminology explains the two study methods with the strongest evidence. The article on spaced repetition explains why short sessions on different days are better than one long session.

Psychology World is built for this kind of practice. You look up a term and save it to one of your own lists. Then you drill the list as flashcards, by typing the answer, or as multiple-choice questions once the list has four or more terms. The four terms in the grid make a ready list. The four steps of the learning loop and what each study mode does are on the home page.