Romance and Fertility Treatment: What Actually Changes

On the measured evidence, usually not. ESHRE's 2015 guideline on routine psychosocial care states at evidence level B that "patients starting first-line or ART treatments do not have worse marital and sexual relationships than the general population", and that "relational satisfaction of patients does not change from before they start an IVF/ICSI cycle to after the pregnancy test".

What does change is sex. At the same evidence level the guideline records that "women experience lower sexual satisfaction after the pregnancy test than before the start of an IVF/ICSI cycle". Its own summary of the evidence puts both halves in one sentence: marital satisfaction remains stable across one cycle of treatment, regardless of its outcome, but sexual satisfaction decreases for all women.

That distinction is worth holding onto, because it is the opposite of how this subject is usually written about. The thing under strain is generally the sex, not the partnership. If you are reading this at eleven at night because it feels like the relationship is failing, the measured finding is that the part which actually tends to change is narrower than it feels.

ESHRE's graded statements map onto the cycle closely enough to be useful as a forecast.

• Intimacy goes up, not down. Women report more intimacy with their partner during an IVF or ICSI cycle than during a normal menstrual cycle, in particular on the retrieval and transfer days. Level B.
• Support from other people goes down. Women report lower social support from significant others in the period between the oocyte retrieval and the embryo transfer than during the equivalent period in a normal menstrual cycle. Level B.
• Sexual satisfaction falls after the pregnancy test, compared with before the cycle started — and the guideline notes this happened for women whose treatment succeeded as well as for those whose treatment did not. Level B.
• Work takes a visible hit. During an IVF or ICSI cycle, 6 in 10 patients report treatment-related absences from work, missing 23 hours on average. Level C.

Over a longer horizon than a single cycle, the picture is less comfortable. A 12-month randomised trial that followed 450 people actively trying to conceive found that depression rose and sexual functioning declined across the whole study, in every arm, including the arm given no intervention at all. In women, every dimension measured fell — desire, satisfaction, arousal, pain, orgasm and lubrication. In men, orgasm, intercourse satisfaction and erectile function changed significantly, while desire and overall sexual satisfaction did not. Stress dipped at six months and was back to baseline at a year. The detail of what happens to sex specifically is covered in why sex often gets harder after an unsuccessful cycle.

This is the one piece of standard advice on this topic that has actually been put in a randomised trial, and it did not survive.

The trial randomised 450 childless people actively trying to conceive into three arms: 127 were taught to have intercourse every other day, 135 to monitor the fertile window, and 134 got no intervention. They were followed for 12 months. There were no significant differences between the arms in stress, anxiety, depression or sexual functioning at any point. Revealed pregnancy rates were 16%, 30% and 20% respectively and did not differ significantly between the randomised arms. The authors' conclusion is blunt: stress does not arise from feeling pressured on the fertile period.

The 2023 Cochrane review of timed intercourse went further and treated harm as an outcome to be measured rather than assumed. Across 7 randomised trials and 2,464 women or couples, it looked at stress, anxiety, depression and erectile dysfunction. For fertility-awareness timing against no ovulation prediction, the differences were small and the confidence intervals crossed zero throughout — erectile dysfunction mean difference 1.2 (95% CI -0.38 to 2.78), stress -1.10 (95% CI -3.88 to 1.68) — all graded very low quality.

So "stop tracking, just be spontaneous, it will take the pressure off" is a reasonable-sounding suggestion that has been tested and did not reduce stress. If timing is making you miserable, that is a good enough reason to change it. But it is worth knowing that the misery is unlikely to be coming from the schedule.

Two large cohorts have looked, and they do not fully agree with each other. Both are worth knowing precisely because they disagree.

The Danish study followed 47,515 women referred for fertility problems between 1990 and 2006, tracked through national registries. After up to 12 years of follow-up, nearly 27% of the cohort were no longer living with the person they had lived with at the time of the fertility evaluation. Women who did not have a child after the evaluation had significantly higher odds of the relationship ending — with odds ratios up to 3.13 (95% CI 2.88 to 3.41) — than women who did, regardless of whether they already had a child beforehand.

The Swedish follow-up disagrees on the key point. It followed 660 participants for up to 10 years after fertility treatment. Among heterosexual couples, 11% to 17% of participants reported having separated, and the authors report that neither background factors nor treatment success protected against separation. Where the Danish data suggest having a child is protective, the Swedish data do not find that.

Neither study compared its couples with a matched general-population cohort. So neither one establishes that fertility treatment raises the chance of separation above the background rate for couples of the same age over the same number of years. That is a real gap, and it is the gap that would answer the question most people are actually asking.

Nobody has shown that it does, and the reason is worth understanding rather than glossing.

The 2016 Cochrane review of psychological and educational interventions for subfertile men and women covered 39 studies and 4,925 participants undergoing assisted reproduction. The reviewers did not produce a pooled effect, because they judged that pooling these studies would not give a clinically meaningful estimate of a treatment effect. Every included study was at high risk of bias in at least one domain. Only two of the 39 reported live birth, both with substantial drop-out. Their conclusion was that the effects on mental health and on live birth or ongoing pregnancy are uncertain because the quality of the evidence is very low.

Note what that does and does not say. It does not say counselling fails. It says the trials were too poorly built to tell, which is a different and more frustrating finding. ESHRE still recommends routine psychosocial care, on the basis that patients need it, not on the basis of a demonstrated effect on pregnancy.

There is one measured reason to take relationship strain seriously as a practical matter rather than an emotional one: relational problems appear among the most-cited reasons people give for stopping treatment altogether. If the relationship is the thing that stalls, the treatment stalls with it. What distress during treatment actually looks like, with the prevalence figures, is in which points in an IVF cycle are hardest and how common distress in infertility actually is.

This section is deliberately free of citations, because there is nothing honest to attach to it.

The standard suggestions — a weekly date night, a small ritual like coffee together before the day starts, non-sexual physical affection, protecting one conversation a week from the subject of treatment, keeping a hobby that has nothing to do with any of it — have not been tested in a trial against relationship outcomes or against pregnancy. There is no randomised trial of a date night. Any article that attaches a reference to that advice is attaching it to something else.

That is not an argument against doing any of it. Reasonable and untested is a perfectly normal category, and it is different from harmful. It just means the honest framing is "couples commonly find this helps" and not "this is shown to protect your relationship".

Two things in this area do have evidence behind them, both at ESHRE's weakest grade, level C, and both point somewhere less obvious than a restaurant booking. First, coping style: meaning-based coping, such as thinking about the fertility problem in a positive light or finding other goals in life, is associated with lower fertility-specific marital and social distress, while avoidance coping, such as avoiding being among pregnant women, is associated with higher distress. Second, and more usefully: couples who have different views on the importance of parenthood and on social concerns may show lower relationship satisfaction than those who have similar views.

That second one suggests the conversation with the most evidence behind it is not about romance at all. It is about how much becoming parents actually matters to each of you, and whether your answers match — including the answer about when you would stop. Also at level C: the way one partner reacts to the diagnosis is associated with how the other reacts, so neither of you is processing this alone even when it feels that way.

On a subject this easy to write comfortingly about, the limits matter more than the findings.

• That treatment raises separation risk above background. Neither the Danish nor the Swedish cohort had a matched general-population comparison group, so the honest reading of both is that a proportion of these relationships ended over a decade, not that treatment ended them.
• Whether having a child protects the relationship. The two cohorts directly disagree: the Danish registry data found not having a child strongly associated with separation, and the Swedish follow-up reported that treatment success did not protect against it.
• That anything protects the relationship. No trial has tested any specific relationship intervention during fertility treatment. Cochrane could not even pool the counselling trials that exist.
• That sexual decline and relationship decline travel together. A one-year study of 66 couples in treatment found the two domains moved independently: men's sexual functioning declined significantly while relationship adjustment stayed stable for both partners, and neither predicted the other. That is a small study and should be read as a caution against assuming, not as a finding.
• Anything about Indian couples. ESHRE states directly that these needs may differ according to patients' cultural, ethnic and religious background but that evidence about this is non-existent. Every study on this page is Danish, Dutch, Portuguese, Swedish, Swiss or American. Nothing was retrieved on Indian couples, on joint-family living arrangements, or on marriages where the pressure to conceive comes substantially from outside the couple — which for many readers here is the actual problem.
• Most of the ESHRE evidence covers one cycle. Treatment often runs for years, and the guideline's grade B findings about stability were measured from before a cycle to after its pregnancy test.

The most reassuring thing on this page is also the most solid: across a treatment cycle the evidence says your relationship satisfaction probably holds, even though your sex life probably does not, and those two facts are easy to confuse when you are living inside them.