Evidence-based definition of hypoprolactinemia in European men aged 40-86 years: the European male ageing study

Abstract

Empirical evidence for a low normal or reference interval for serum prolactin (PRL) is lacking for men, while the implications of very low PRL levels for human health have never been studied. A clinical state of "PRL deficiency" has not been defined except in relation to lactation. Using data from the European Male Ageing Study (EMAS), we analyzed the distribution of PRL in 3,369 community-dwelling European men, aged 40-80 years at phase-1 and free from acute illnesses. In total, 2,948 and 2,644 PRL samples were collected during phase-1 and phase-2 (3 to 5.7 years later). All samples were analysed in the same centre with the same assay. After excluding individuals with known pituitary diseases, PRL ≥ 35 ng/ml, and PRL-altering drugs including antipsychotic agents, selective serotonin reuptake inhibitors, or dopamine agonists, 5,086 data points (2,845 in phase-1 and 2,241 in phase-2) were available for analysis. The results showed that PRL declined minimally with age (slope = -0.02) and did not correlate with BMI. The positively skewed PRL distribution was log-transformed to a symmetrical distribution (skewness reduced from 13.3 to 0.015). Using two-sigma empirical rule (2[]SD about the mean), a threshold at 2.5% of the lower end of the distribution was shown to correspond to a PRL value of 2.98ng/ml. With reference to individuals with PRL levels of 5-34.9 ng/ml (event rate = 6.3%), the adjusted risk of developing type 2 diabetes increased progressively in those with PRL levels of 3-4.9 ng/ml: event rate = 9.3%, OR (95% CI) 1.59 (0.93-2.71), and more so with PRL levels of 0.3-2.9 ng/ml: event rate = 22.7%, OR 5.45 (1.78-16.62). There was also an increasing trend in prediabetes and diabetes based on fasting blood glucose levels was observed with lower categories of PRL. However, PRL levels were not associated with cancer, cardiovascular diseases, depressive symptoms or mortality. Our findings suggest that a PRL level below 3 ng/ml (64 mlU/l) significantly identifies European men with a clinically-important outcome (of type 2 diabetes), offering a lower reference-value for research and clinical practice.

Keywords: Health outcomes; Hypoprolactinemia; Pituitary function; Reference values; Type 2 diabetes.

Fig. 1

Scatter plots showing PRL levels against age: ranging from 40 to 86 years (up to 80 years in phase-1)

Fig. 2

Distribution of PRL in men before (A) and after log-transformation. Arrow indicates value correspond to mean? 2SD (0.853? 1.96 []0.139 = 0.475) (B).

Fig. 3

Incidence of type 2 diabetes developed by the time of phase-2 study according to phase-1 PRL levels

Fig. 4

Distribution of fasting blood glucose levels according to different categories of phase-1 PRL levels amongst men without a previous history of diabetes