In his momentous study L'origine des systèmes familiaux, Emmanuel Todd notes that, to the best of a rather sparse archival knowledge, first, the status of women in Eurasia in the 5000BC-1500CE interval seems to historically follow a lowering trajectory, second, this lowering trajectory seems to proceed in a top/down fashion and, third, the adoption of anti-women social mores in the higher classes is linked with the militarization of society and military expansion.
To be more precise, pre-civilization societies seem to display a uniformly rather high status for women. It is usually first in the ruling classes that the status of women is lowered (the low status then trickles down through society) and the adoption of anti-feminine social arrangements frequently temporally coincides with changes in the military organization of the society. Todd offers the tentative hypothesis that the idea of restructuring social mores in more military efficient but less pro-feminine ways might have appeared originally in very few places in Eurasia (among them the steppe people of northern China) and then, later on, has spread through military conquests and imitation.
Todd offers the following evidence for his thesis. First, the vast majority of hunter/gatherer people that modern anthropologists have described ascribes a high status to woman. Second, a map of Eurasia describing the gradient of woman status circa 1500CE displays the usual appearance of a diffusion process map, with peripheral zones (both in the sense of geography and intensity of cultural contacts) typically exhibiting the conservation of archaic forms and central (ditto) zones exhibiting the newer anti-feminine forms. Third, the archival record displays a clear duality in which royal and higher classes families gradually adopt more and more anti-feminine familial and social arrangements whereas lower classes censuses betray more egalitarian arrangements up to a much later period (though they eventually catch up).
I don't think this hypothesis can be deemed uncontroversial, but I do recommend skimming the reference for a sample of the evidence in support of it (which is both quite impressive and quite sparse, as it is of course extremely hard to probe the nature of gender relations in past societies, if only because contemporary documents take them for granted).
As an aside, I'll note that seen under this perspective, the relatively rather high status ascribed to women in late medieval, early modern period in England (see for instance the works of Peter Laslett) derives from the peripheral position of the British Isles at the very end of Eurasia. As the attitudes and conceptions of the English speaking world became the de facto international norm because of the disproportionate geopolitical role the United Kingdom and the United States came to play in the XIXth and XXth century, we may attribute the rather high status given to women in the contemporary developed world to the fortuitous later spread of an archaic form originally preserved because of geographical isolation. In a more pessimistic note, the thesis also suggests a factor explaining why historical centers of eurasian civilization (China, northern India, the middle-east, the Arab world) remained largely and very significantly underdeveloped in the XIXth and XXth century: as is well-known, the spread of literacy is highly correlated with a high status of women, so historical centers of civilizations, which had adopted the most "advanced" (we would anachronistically call them backwards) forms of anti-feminine mores, dragged (and still drag, to a large extent) behind peripheral zones in the adoption of widespread literacy.