Dreams during pregnancy, grief and illness: when the content shifts

If dream content tracks waking preoccupation — the continuity hypothesis, the field's most robust general principle — then periods of intense preoccupation should reshape dreams. They do, and in ways specific enough to have been studied.

Pregnancy

Dream recall increases during pregnancy, particularly in the third trimester. Part of the explanation is mechanical: sleep becomes fragmented by discomfort, nocturia, heartburn and fetal movement, and more awakenings mean more opportunities to encode dream content. Hormonal changes and altered sleep architecture likely contribute.

Content shifts predictably. Studies find increased dreams about the baby, about labour, about the partner, and about the dreamer's own mother — the last being one of the more consistent findings, and an interesting one.

Anxiety dreams are common and frequently distressing: dreams of losing the baby, of forgetting it, of the baby being deformed or unwell, of being unable to protect it. Tore Nielsen and Tonya Paquette's work found that pregnant women reported more infant-related nightmares than non-pregnant controls.

The reassuring finding is that this appears to be normal processing rather than a warning sign. Some studies have found that women who report more anxiety dreams during pregnancy have shorter labours and fewer complications — consistent with the idea that dream rehearsal of feared outcomes is doing something useful. That finding is not robust enough to be stated as fact, but it is the opposite of alarming.

Grief

Dreams of the dead are reported by a large majority of the bereaved — survey estimates commonly range from 50 to 80 per cent — and they follow a rough trajectory.

Early dreams often feature the deceased as ill, dying, or in distress, and frequently reproduce the circumstances of death. Over months, content tends to shift: the person appears well, younger, at peace, sometimes explicitly communicating that they are all right. Later still, they may simply appear as a presence in ordinary scenes.

Joshua Black's research on grief dreams found that most bereaved people describe them as comforting rather than distressing, and that they are commonly experienced as a form of continued connection. A substantial minority find them painful — particularly dreams in which the person is alive and the dreamer wakes to lose them again.

Rosalind Cartwright's longitudinal work on people going through divorce found a related pattern with practical implications: those whose dreams featured the former spouse, and whose dream narratives changed over time, had better psychological outcomes a year later than those who did not dream of them. Dreaming about it, and the dream evolving, tracked adaptation.

Whether the dreaming causes the adaptation or reflects it is not established. But it suggests that dreams of the dead are part of the work of grief rather than an interruption to it.

Serious and terminal illness

The most studied area here is end-of-life dreams and visions, particularly work by Christopher Kerr's group at a hospice in Buffalo.

In a prospective study of hospice patients, the great majority reported dreams or visions in their final weeks, and most rated them as comforting. Content clustered strongly around a few themes: deceased relatives and friends appearing, often waiting; preparing to travel or go somewhere; reliving meaningful past experiences; and reuniting with those already dead.

Notably, dreams featuring the deceased were rated as the most comforting, while those featuring living people were less so. And the frequency of such dreams increased as death approached.

These are distinguished from delirium, which is common in the dying and is characterised by confusion, fluctuating consciousness and distress. End-of-life dreams occur in clear consciousness, the person can describe them coherently afterwards, and they are typically reassuring rather than frightening. Confusing the two matters clinically, because delirium requires treatment and these do not.

What the pattern suggests

Across all three situations, the same structure appears. Dreams concentrate on the central preoccupation. They frequently rehearse feared outcomes. And, over time, the narrative tends to move toward resolution — the baby is safe, the dead are well, the journey is prepared for.

This is what an emotion-processing account of REM sleep would predict, and it is among the better observational support for it. It does not prove the mechanism, and these are self-reported data from people in extremis, which brings obvious limitations.

But it does suggest that in the periods of life when dreaming most obviously changes, it is doing something recognisable — and that the frightening dreams during those periods are more likely part of the process than a sign that something has gone wrong.