There are many variables that can affect both a childs attachment style and well-being later in life, whether an unfortunate event, depression in the caregiver, or socioeconomic status. However, the variable does not even have to be so general, it could be something as simple as a single bad experience, like a child getting into trouble. Say, for example, a boy likes this girl. In his way of getting her attention, he mischievously puts gum in her hair (he is typically good, but this day he emulates his friends behavior). Assuming he gets caught, he would get chastised by his teacher, parents, classmates, and would be ignored by this girl. This isolated circumstance could lead to a different attachment style and state of well-being later in life. It is a bit extreme, but say this boy is now shunned by his classmates then he is not able to socialize and develop strong interpersonal skills. He could be scarred by this experience and may be insecure around all girls for the rest of his life. This would certainly lead to an anxious/avoidant attachment and a significant change in the well-being later in life. This alternate variable may be extreme, but the point is that really any circumstance that alters the childs internal working model would lead to a different attachment style and sense of well-being later in life.
I agree that life has too many variables that can affect them. Uncontrollable events make it very difficult for studies to show causality. Controlling a situation perfectly is necessary to show causality. Only one variable can be changing. In life there is always more than thing changing. I agree that Isolated incidents can change the long term outcome. This is a good alternative variable because there are always new incidents that can happen to change the orientation of a persons attachment, such as being conned by someone who you thought was a close friend.
